{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9763241396143519,
  "global_step": 100000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "accuracy": 16.1133,
      "active_queue_size": 16384.0,
      "cl_loss": 152.6867,
      "doc_norm": 8.4295,
      "encoder_q-embeddings": 51007.5977,
      "encoder_q-layer.0": 53569.707,
      "encoder_q-layer.1": 41899.4609,
      "encoder_q-layer.10": 135516.8594,
      "encoder_q-layer.11": 77749.5703,
      "encoder_q-layer.2": 47946.8906,
      "encoder_q-layer.3": 54582.1797,
      "encoder_q-layer.4": 62467.4023,
      "encoder_q-layer.5": 71678.1875,
      "encoder_q-layer.6": 95023.6719,
      "encoder_q-layer.7": 113375.4844,
      "encoder_q-layer.8": 144191.7656,
      "encoder_q-layer.9": 117985.0703,
      "epoch": 0.0,
      "inbatch_neg_score": 40.2664,
      "inbatch_pos_score": 48.8438,
      "learning_rate": 5.000000000000001e-07,
      "loss": 152.6867,
      "norm_diff": 0.4408,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 120451.7487,
      "preclip_grad_norm_avg": 0.0011,
      "q@queue_neg_score": 40.25,
      "query_norm": 7.9887,
      "queue_k_norm": 8.42,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7838,
      "sent_len_1": 66.8485,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.14,
      "stdk": 0.1804,
      "stdq": 0.1822,
      "stdqueue_k": 0.1803,
      "stdqueue_q": 0.0,
      "step": 100
    },
    {
      "accuracy": 23.4375,
      "active_queue_size": 16384.0,
      "cl_loss": 98.4148,
      "doc_norm": 8.3214,
      "encoder_q-embeddings": 6186.1929,
      "encoder_q-layer.0": 5516.9307,
      "encoder_q-layer.1": 5860.103,
      "encoder_q-layer.10": 16453.1895,
      "encoder_q-layer.11": 17405.5723,
      "encoder_q-layer.2": 6710.937,
      "encoder_q-layer.3": 7751.21,
      "encoder_q-layer.4": 8745.4561,
      "encoder_q-layer.5": 9579.3057,
      "encoder_q-layer.6": 11831.2695,
      "encoder_q-layer.7": 12676.0449,
      "encoder_q-layer.8": 15957.1455,
      "encoder_q-layer.9": 12641.7002,
      "epoch": 0.0,
      "inbatch_neg_score": 36.1859,
      "inbatch_pos_score": 41.8125,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 98.4148,
      "norm_diff": 1.1728,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 15678.1734,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 36.2188,
      "query_norm": 7.1486,
      "queue_k_norm": 8.3464,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1228,
      "sent_len_1": 66.9379,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8713,
      "stdk": 0.1769,
      "stdq": 0.1352,
      "stdqueue_k": 0.178,
      "stdqueue_q": 0.0,
      "step": 200
    },
    {
      "accuracy": 21.9727,
      "active_queue_size": 16384.0,
      "cl_loss": 61.9136,
      "doc_norm": 8.2003,
      "encoder_q-embeddings": 3729.0554,
      "encoder_q-layer.0": 3289.6116,
      "encoder_q-layer.1": 3431.8345,
      "encoder_q-layer.10": 9481.2656,
      "encoder_q-layer.11": 14086.2002,
      "encoder_q-layer.2": 3918.9822,
      "encoder_q-layer.3": 4536.834,
      "encoder_q-layer.4": 5001.6821,
      "encoder_q-layer.5": 5331.4565,
      "encoder_q-layer.6": 6225.4551,
      "encoder_q-layer.7": 6548.1802,
      "encoder_q-layer.8": 7332.2334,
      "encoder_q-layer.9": 6760.0791,
      "epoch": 0.0,
      "inbatch_neg_score": 33.6667,
      "inbatch_pos_score": 37.3438,
      "learning_rate": 1.5e-06,
      "loss": 61.9136,
      "norm_diff": 1.3912,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 10066.434,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 33.6875,
      "query_norm": 6.8091,
      "queue_k_norm": 8.22,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7772,
      "sent_len_1": 66.5307,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1687,
      "stdk": 0.1733,
      "stdq": 0.1107,
      "stdqueue_k": 0.174,
      "stdqueue_q": 0.0,
      "step": 300
    },
    {
      "accuracy": 23.1445,
      "active_queue_size": 16384.0,
      "cl_loss": 45.1854,
      "doc_norm": 8.0457,
      "encoder_q-embeddings": 2858.24,
      "encoder_q-layer.0": 2590.9626,
      "encoder_q-layer.1": 3073.8232,
      "encoder_q-layer.10": 6466.6074,
      "encoder_q-layer.11": 12099.9307,
      "encoder_q-layer.2": 3444.9409,
      "encoder_q-layer.3": 3504.2925,
      "encoder_q-layer.4": 3945.6741,
      "encoder_q-layer.5": 4097.5342,
      "encoder_q-layer.6": 4063.8943,
      "encoder_q-layer.7": 4379.2822,
      "encoder_q-layer.8": 5131.9609,
      "encoder_q-layer.9": 4168.5591,
      "epoch": 0.0,
      "inbatch_neg_score": 31.5259,
      "inbatch_pos_score": 34.3438,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 45.1854,
      "norm_diff": 1.3234,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8146.2371,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 31.4531,
      "query_norm": 6.7223,
      "queue_k_norm": 8.0574,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0017,
      "sent_len_1": 66.9184,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4412,
      "stdk": 0.1667,
      "stdq": 0.0987,
      "stdqueue_k": 0.168,
      "stdqueue_q": 0.0,
      "step": 400
    },
    {
      "accuracy": 21.7773,
      "active_queue_size": 16384.0,
      "cl_loss": 34.5652,
      "doc_norm": 7.8753,
      "encoder_q-embeddings": 3824.6829,
      "encoder_q-layer.0": 3238.5513,
      "encoder_q-layer.1": 3905.2439,
      "encoder_q-layer.10": 5740.6982,
      "encoder_q-layer.11": 8291.0645,
      "encoder_q-layer.2": 4116.2266,
      "encoder_q-layer.3": 4481.3774,
      "encoder_q-layer.4": 4755.6123,
      "encoder_q-layer.5": 4885.0317,
      "encoder_q-layer.6": 4665.1968,
      "encoder_q-layer.7": 4779.8628,
      "encoder_q-layer.8": 5177.8521,
      "encoder_q-layer.9": 3587.8091,
      "epoch": 0.0,
      "inbatch_neg_score": 28.7854,
      "inbatch_pos_score": 30.8594,
      "learning_rate": 2.5e-06,
      "loss": 34.5652,
      "norm_diff": 1.1632,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7096.5572,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 28.7188,
      "query_norm": 6.7122,
      "queue_k_norm": 7.899,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8715,
      "sent_len_1": 66.8038,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5425,
      "stdk": 0.1616,
      "stdq": 0.0973,
      "stdqueue_k": 0.1634,
      "stdqueue_q": 0.0,
      "step": 500
    },
    {
      "accuracy": 25.0977,
      "active_queue_size": 16384.0,
      "cl_loss": 28.0766,
      "doc_norm": 7.7053,
      "encoder_q-embeddings": 2919.553,
      "encoder_q-layer.0": 2174.5518,
      "encoder_q-layer.1": 2527.2869,
      "encoder_q-layer.10": 4905.1699,
      "encoder_q-layer.11": 8246.8965,
      "encoder_q-layer.2": 2868.7869,
      "encoder_q-layer.3": 3287.3162,
      "encoder_q-layer.4": 3584.0942,
      "encoder_q-layer.5": 3656.9592,
      "encoder_q-layer.6": 4182.854,
      "encoder_q-layer.7": 4374.104,
      "encoder_q-layer.8": 4863.1099,
      "encoder_q-layer.9": 3183.1306,
      "epoch": 0.01,
      "inbatch_neg_score": 26.7571,
      "inbatch_pos_score": 28.7031,
      "learning_rate": 3e-06,
      "loss": 28.0766,
      "norm_diff": 1.2085,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6247.8511,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 26.7188,
      "query_norm": 6.4967,
      "queue_k_norm": 7.7237,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0611,
      "sent_len_1": 66.6908,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3963,
      "stdk": 0.1564,
      "stdq": 0.0912,
      "stdqueue_k": 0.1572,
      "stdqueue_q": 0.0,
      "step": 600
    },
    {
      "accuracy": 20.3125,
      "active_queue_size": 16384.0,
      "cl_loss": 24.4455,
      "doc_norm": 7.5581,
      "encoder_q-embeddings": 4678.5566,
      "encoder_q-layer.0": 3961.03,
      "encoder_q-layer.1": 4597.9473,
      "encoder_q-layer.10": 6290.4248,
      "encoder_q-layer.11": 9845.4521,
      "encoder_q-layer.2": 5590.4956,
      "encoder_q-layer.3": 6382.0537,
      "encoder_q-layer.4": 7086.5435,
      "encoder_q-layer.5": 7008.1816,
      "encoder_q-layer.6": 6839.5986,
      "encoder_q-layer.7": 5888.8345,
      "encoder_q-layer.8": 5331.6733,
      "encoder_q-layer.9": 3622.8784,
      "epoch": 0.01,
      "inbatch_neg_score": 22.6203,
      "inbatch_pos_score": 24.0625,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 24.4455,
      "norm_diff": 1.7144,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8988.4622,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 22.6406,
      "query_norm": 5.8437,
      "queue_k_norm": 7.5491,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9345,
      "sent_len_1": 66.8035,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.655,
      "stdk": 0.1527,
      "stdq": 0.0848,
      "stdqueue_k": 0.1516,
      "stdqueue_q": 0.0,
      "step": 700
    },
    {
      "accuracy": 23.3398,
      "active_queue_size": 16384.0,
      "cl_loss": 20.7209,
      "doc_norm": 7.3859,
      "encoder_q-embeddings": 5386.4146,
      "encoder_q-layer.0": 4290.5127,
      "encoder_q-layer.1": 5047.5718,
      "encoder_q-layer.10": 4962.1499,
      "encoder_q-layer.11": 7966.4873,
      "encoder_q-layer.2": 5975.3179,
      "encoder_q-layer.3": 6707.395,
      "encoder_q-layer.4": 7883.54,
      "encoder_q-layer.5": 8294.8926,
      "encoder_q-layer.6": 7347.8965,
      "encoder_q-layer.7": 4995.4824,
      "encoder_q-layer.8": 4064.4773,
      "encoder_q-layer.9": 2562.7864,
      "epoch": 0.01,
      "inbatch_neg_score": 16.8464,
      "inbatch_pos_score": 18.0938,
      "learning_rate": 4.000000000000001e-06,
      "loss": 20.7209,
      "norm_diff": 2.4878,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8796.3367,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 16.8125,
      "query_norm": 4.8981,
      "queue_k_norm": 7.3864,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0355,
      "sent_len_1": 66.9774,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8975,
      "stdk": 0.1448,
      "stdq": 0.079,
      "stdqueue_k": 0.1456,
      "stdqueue_q": 0.0,
      "step": 800
    },
    {
      "accuracy": 21.7773,
      "active_queue_size": 16384.0,
      "cl_loss": 17.2291,
      "doc_norm": 7.2554,
      "encoder_q-embeddings": 3836.0134,
      "encoder_q-layer.0": 3397.3567,
      "encoder_q-layer.1": 4210.4028,
      "encoder_q-layer.10": 3917.2729,
      "encoder_q-layer.11": 6850.8428,
      "encoder_q-layer.2": 4756.7373,
      "encoder_q-layer.3": 5011.4355,
      "encoder_q-layer.4": 5103.0269,
      "encoder_q-layer.5": 5417.605,
      "encoder_q-layer.6": 4845.3281,
      "encoder_q-layer.7": 4115.0869,
      "encoder_q-layer.8": 4466.5796,
      "encoder_q-layer.9": 2171.3955,
      "epoch": 0.01,
      "inbatch_neg_score": 12.5544,
      "inbatch_pos_score": 13.5703,
      "learning_rate": 4.5e-06,
      "loss": 17.2291,
      "norm_diff": 3.2823,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6738.4929,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 12.5547,
      "query_norm": 3.9731,
      "queue_k_norm": 7.2469,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9054,
      "sent_len_1": 66.6723,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.71,
      "stdk": 0.1418,
      "stdq": 0.0715,
      "stdqueue_k": 0.1412,
      "stdqueue_q": 0.0,
      "step": 900
    },
    {
      "accuracy": 24.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 14.7158,
      "doc_norm": 7.0913,
      "encoder_q-embeddings": 7003.1187,
      "encoder_q-layer.0": 6053.4297,
      "encoder_q-layer.1": 7565.7158,
      "encoder_q-layer.10": 6889.4858,
      "encoder_q-layer.11": 9275.0742,
      "encoder_q-layer.2": 9215.9004,
      "encoder_q-layer.3": 10496.5645,
      "encoder_q-layer.4": 10570.2217,
      "encoder_q-layer.5": 12979.5234,
      "encoder_q-layer.6": 10683.4268,
      "encoder_q-layer.7": 6766.2407,
      "encoder_q-layer.8": 5287.9434,
      "encoder_q-layer.9": 3001.6709,
      "epoch": 0.01,
      "inbatch_neg_score": 6.2447,
      "inbatch_pos_score": 7.0664,
      "learning_rate": 5e-06,
      "loss": 14.7158,
      "norm_diff": 4.1098,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12373.619,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 6.2305,
      "query_norm": 2.9816,
      "queue_k_norm": 7.1011,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7538,
      "sent_len_1": 66.7028,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.835,
      "stdk": 0.1346,
      "stdq": 0.0659,
      "stdqueue_k": 0.1356,
      "stdqueue_q": 0.0,
      "step": 1000
    },
    {
      "accuracy": 22.8516,
      "active_queue_size": 16384.0,
      "cl_loss": 13.0161,
      "doc_norm": 6.9623,
      "encoder_q-embeddings": 5020.8496,
      "encoder_q-layer.0": 4367.4409,
      "encoder_q-layer.1": 5435.6465,
      "encoder_q-layer.10": 6886.0049,
      "encoder_q-layer.11": 9586.541,
      "encoder_q-layer.2": 6104.6611,
      "encoder_q-layer.3": 6454.103,
      "encoder_q-layer.4": 6910.0054,
      "encoder_q-layer.5": 7788.3604,
      "encoder_q-layer.6": 6524.2666,
      "encoder_q-layer.7": 4797.209,
      "encoder_q-layer.8": 4587.3198,
      "encoder_q-layer.9": 2720.4368,
      "epoch": 0.01,
      "inbatch_neg_score": 5.1627,
      "inbatch_pos_score": 5.9102,
      "learning_rate": 5.500000000000001e-06,
      "loss": 13.0161,
      "norm_diff": 4.4445,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8832.3905,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 5.1523,
      "query_norm": 2.5179,
      "queue_k_norm": 6.9724,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9937,
      "sent_len_1": 66.8561,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8713,
      "stdk": 0.1295,
      "stdq": 0.0613,
      "stdqueue_k": 0.1306,
      "stdqueue_q": 0.0,
      "step": 1100
    },
    {
      "accuracy": 21.4844,
      "active_queue_size": 16384.0,
      "cl_loss": 11.8585,
      "doc_norm": 6.8405,
      "encoder_q-embeddings": 6105.896,
      "encoder_q-layer.0": 5575.6924,
      "encoder_q-layer.1": 5750.4629,
      "encoder_q-layer.10": 3034.7285,
      "encoder_q-layer.11": 5181.7393,
      "encoder_q-layer.2": 5588.8657,
      "encoder_q-layer.3": 6181.9927,
      "encoder_q-layer.4": 6501.4424,
      "encoder_q-layer.5": 6340.4478,
      "encoder_q-layer.6": 5119.0654,
      "encoder_q-layer.7": 3790.7144,
      "encoder_q-layer.8": 3076.1892,
      "encoder_q-layer.9": 1429.9585,
      "epoch": 0.01,
      "inbatch_neg_score": 4.361,
      "inbatch_pos_score": 5.0312,
      "learning_rate": 6e-06,
      "loss": 11.8585,
      "norm_diff": 4.5677,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7643.8992,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 4.3516,
      "query_norm": 2.2727,
      "queue_k_norm": 6.8585,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9739,
      "sent_len_1": 66.9489,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.79,
      "stdk": 0.1242,
      "stdq": 0.059,
      "stdqueue_k": 0.1257,
      "stdqueue_q": 0.0,
      "step": 1200
    },
    {
      "accuracy": 22.3633,
      "active_queue_size": 16384.0,
      "cl_loss": 11.12,
      "doc_norm": 6.7301,
      "encoder_q-embeddings": 5076.1084,
      "encoder_q-layer.0": 4198.1504,
      "encoder_q-layer.1": 5093.6304,
      "encoder_q-layer.10": 4892.8271,
      "encoder_q-layer.11": 7971.5298,
      "encoder_q-layer.2": 5878.6816,
      "encoder_q-layer.3": 6171.4009,
      "encoder_q-layer.4": 6603.8149,
      "encoder_q-layer.5": 6605.2729,
      "encoder_q-layer.6": 5236.2656,
      "encoder_q-layer.7": 4029.437,
      "encoder_q-layer.8": 3778.9209,
      "encoder_q-layer.9": 2266.604,
      "epoch": 0.01,
      "inbatch_neg_score": 3.8684,
      "inbatch_pos_score": 4.5391,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 11.12,
      "norm_diff": 4.58,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7828.1873,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 3.873,
      "query_norm": 2.1502,
      "queue_k_norm": 6.7462,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9514,
      "sent_len_1": 66.4993,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7375,
      "stdk": 0.1205,
      "stdq": 0.0578,
      "stdqueue_k": 0.1208,
      "stdqueue_q": 0.0,
      "step": 1300
    },
    {
      "accuracy": 22.7539,
      "active_queue_size": 16384.0,
      "cl_loss": 10.6478,
      "doc_norm": 6.6385,
      "encoder_q-embeddings": 6853.3169,
      "encoder_q-layer.0": 5723.9565,
      "encoder_q-layer.1": 6950.7607,
      "encoder_q-layer.10": 5074.8027,
      "encoder_q-layer.11": 7628.8955,
      "encoder_q-layer.2": 7920.2319,
      "encoder_q-layer.3": 9578.335,
      "encoder_q-layer.4": 11384.2676,
      "encoder_q-layer.5": 10430.2451,
      "encoder_q-layer.6": 9264.7822,
      "encoder_q-layer.7": 8680.9775,
      "encoder_q-layer.8": 7842.1167,
      "encoder_q-layer.9": 2144.6252,
      "epoch": 0.01,
      "inbatch_neg_score": 0.4837,
      "inbatch_pos_score": 1.0859,
      "learning_rate": 7.000000000000001e-06,
      "loss": 10.6478,
      "norm_diff": 4.5605,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 11596.4019,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4954,
      "query_norm": 2.078,
      "queue_k_norm": 6.6406,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9964,
      "sent_len_1": 66.6952,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 193.69,
      "stdk": 0.1167,
      "stdq": 0.0576,
      "stdqueue_k": 0.1161,
      "stdqueue_q": 0.0,
      "step": 1400
    },
    {
      "accuracy": 29.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 10.2913,
      "doc_norm": 6.5185,
      "encoder_q-embeddings": 2317.3347,
      "encoder_q-layer.0": 2040.4631,
      "encoder_q-layer.1": 2488.2002,
      "encoder_q-layer.10": 3016.5315,
      "encoder_q-layer.11": 5266.8667,
      "encoder_q-layer.2": 2849.0801,
      "encoder_q-layer.3": 2999.0083,
      "encoder_q-layer.4": 3306.5488,
      "encoder_q-layer.5": 3838.7483,
      "encoder_q-layer.6": 3354.4009,
      "encoder_q-layer.7": 2812.9854,
      "encoder_q-layer.8": 3181.0479,
      "encoder_q-layer.9": 1624.0332,
      "epoch": 0.01,
      "inbatch_neg_score": 2.8672,
      "inbatch_pos_score": 3.5391,
      "learning_rate": 7.5e-06,
      "loss": 10.2913,
      "norm_diff": 4.4826,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4384.8854,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 2.8535,
      "query_norm": 2.036,
      "queue_k_norm": 6.5362,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1342,
      "sent_len_1": 66.8171,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4212,
      "stdk": 0.1108,
      "stdq": 0.0551,
      "stdqueue_k": 0.1117,
      "stdqueue_q": 0.0,
      "step": 1500
    },
    {
      "accuracy": 25.5859,
      "active_queue_size": 16384.0,
      "cl_loss": 10.1659,
      "doc_norm": 6.4187,
      "encoder_q-embeddings": 5117.6982,
      "encoder_q-layer.0": 4360.0142,
      "encoder_q-layer.1": 5421.3486,
      "encoder_q-layer.10": 3502.4888,
      "encoder_q-layer.11": 6449.5957,
      "encoder_q-layer.2": 6348.2896,
      "encoder_q-layer.3": 7336.668,
      "encoder_q-layer.4": 9089.9873,
      "encoder_q-layer.5": 11757.1758,
      "encoder_q-layer.6": 11675.5254,
      "encoder_q-layer.7": 11565.0283,
      "encoder_q-layer.8": 14050.2793,
      "encoder_q-layer.9": 2745.9832,
      "epoch": 0.02,
      "inbatch_neg_score": 1.4739,
      "inbatch_pos_score": 2.1094,
      "learning_rate": 8.000000000000001e-06,
      "loss": 10.1659,
      "norm_diff": 4.3007,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 11956.2675,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 1.4619,
      "query_norm": 2.1179,
      "queue_k_norm": 6.4355,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.967,
      "sent_len_1": 66.7065,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5288,
      "stdk": 0.1059,
      "stdq": 0.0571,
      "stdqueue_k": 0.1074,
      "stdqueue_q": 0.0,
      "step": 1600
    },
    {
      "accuracy": 26.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 9.8474,
      "doc_norm": 6.2998,
      "encoder_q-embeddings": 3390.6692,
      "encoder_q-layer.0": 2550.1501,
      "encoder_q-layer.1": 3091.7336,
      "encoder_q-layer.10": 4054.4353,
      "encoder_q-layer.11": 6081.4326,
      "encoder_q-layer.2": 3583.7302,
      "encoder_q-layer.3": 4008.1538,
      "encoder_q-layer.4": 4194.8525,
      "encoder_q-layer.5": 4100.749,
      "encoder_q-layer.6": 3310.9434,
      "encoder_q-layer.7": 2567.8972,
      "encoder_q-layer.8": 2661.2268,
      "encoder_q-layer.9": 2427.8179,
      "epoch": 0.02,
      "inbatch_neg_score": 1.2413,
      "inbatch_pos_score": 1.8281,
      "learning_rate": 8.500000000000002e-06,
      "loss": 9.8474,
      "norm_diff": 4.2423,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5235.0911,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 1.2354,
      "query_norm": 2.0575,
      "queue_k_norm": 6.328,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.096,
      "sent_len_1": 66.8993,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8638,
      "stdk": 0.102,
      "stdq": 0.0569,
      "stdqueue_k": 0.1031,
      "stdqueue_q": 0.0,
      "step": 1700
    },
    {
      "accuracy": 25.9766,
      "active_queue_size": 16384.0,
      "cl_loss": 9.5674,
      "doc_norm": 6.1876,
      "encoder_q-embeddings": 12693.8818,
      "encoder_q-layer.0": 11755.1133,
      "encoder_q-layer.1": 13685.4785,
      "encoder_q-layer.10": 3898.5986,
      "encoder_q-layer.11": 5853.0601,
      "encoder_q-layer.2": 14593.3398,
      "encoder_q-layer.3": 15675.1875,
      "encoder_q-layer.4": 17667.0547,
      "encoder_q-layer.5": 15614.4443,
      "encoder_q-layer.6": 12648.0723,
      "encoder_q-layer.7": 9721.9795,
      "encoder_q-layer.8": 8022.1592,
      "encoder_q-layer.9": 2939.8848,
      "epoch": 0.02,
      "inbatch_neg_score": 1.8673,
      "inbatch_pos_score": 2.4785,
      "learning_rate": 9e-06,
      "loss": 9.5674,
      "norm_diff": 4.1094,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 17854.8773,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 1.8643,
      "query_norm": 2.0782,
      "queue_k_norm": 6.2058,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8858,
      "sent_len_1": 66.7078,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2962,
      "stdk": 0.0986,
      "stdq": 0.0577,
      "stdqueue_k": 0.0991,
      "stdqueue_q": 0.0,
      "step": 1800
    },
    {
      "accuracy": 24.1211,
      "active_queue_size": 16384.0,
      "cl_loss": 9.4188,
      "doc_norm": 6.0618,
      "encoder_q-embeddings": 7496.1006,
      "encoder_q-layer.0": 6368.9683,
      "encoder_q-layer.1": 7031.6719,
      "encoder_q-layer.10": 4900.3135,
      "encoder_q-layer.11": 5984.6592,
      "encoder_q-layer.2": 6769.1797,
      "encoder_q-layer.3": 7260.5469,
      "encoder_q-layer.4": 7191.2959,
      "encoder_q-layer.5": 6851.2905,
      "encoder_q-layer.6": 4693.9624,
      "encoder_q-layer.7": 3475.1055,
      "encoder_q-layer.8": 3344.3804,
      "encoder_q-layer.9": 2331.8652,
      "epoch": 0.02,
      "inbatch_neg_score": 0.3877,
      "inbatch_pos_score": 0.9648,
      "learning_rate": 9.5e-06,
      "loss": 9.4188,
      "norm_diff": 3.8818,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8762.1053,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3818,
      "query_norm": 2.18,
      "queue_k_norm": 6.0767,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.038,
      "sent_len_1": 66.8197,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7525,
      "stdk": 0.0939,
      "stdq": 0.0609,
      "stdqueue_k": 0.0952,
      "stdqueue_q": 0.0,
      "step": 1900
    },
    {
      "accuracy": 24.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 9.2028,
      "doc_norm": 5.9258,
      "encoder_q-embeddings": 6337.502,
      "encoder_q-layer.0": 5136.2886,
      "encoder_q-layer.1": 6217.7012,
      "encoder_q-layer.10": 5915.4165,
      "encoder_q-layer.11": 6690.9316,
      "encoder_q-layer.2": 7487.5703,
      "encoder_q-layer.3": 7591.8301,
      "encoder_q-layer.4": 7863.6704,
      "encoder_q-layer.5": 7124.5464,
      "encoder_q-layer.6": 5649.3862,
      "encoder_q-layer.7": 5210.644,
      "encoder_q-layer.8": 4757.3521,
      "encoder_q-layer.9": 3741.1741,
      "epoch": 0.02,
      "inbatch_neg_score": 1.2622,
      "inbatch_pos_score": 1.8281,
      "learning_rate": 1e-05,
      "loss": 9.2028,
      "norm_diff": 3.749,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9048.9027,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 1.2598,
      "query_norm": 2.1768,
      "queue_k_norm": 5.9313,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.757,
      "sent_len_1": 66.4313,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.725,
      "stdk": 0.092,
      "stdq": 0.0604,
      "stdqueue_k": 0.0918,
      "stdqueue_q": 0.0,
      "step": 2000
    },
    {
      "accuracy": 26.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 8.9643,
      "doc_norm": 5.7567,
      "encoder_q-embeddings": 12919.2881,
      "encoder_q-layer.0": 11313.4482,
      "encoder_q-layer.1": 12379.6309,
      "encoder_q-layer.10": 2239.6572,
      "encoder_q-layer.11": 3857.6262,
      "encoder_q-layer.2": 14157.3076,
      "encoder_q-layer.3": 14984.0381,
      "encoder_q-layer.4": 14584.9121,
      "encoder_q-layer.5": 14634.5342,
      "encoder_q-layer.6": 9965.7676,
      "encoder_q-layer.7": 7566.5913,
      "encoder_q-layer.8": 7879.6206,
      "encoder_q-layer.9": 2312.843,
      "epoch": 0.02,
      "inbatch_neg_score": 1.0038,
      "inbatch_pos_score": 1.5898,
      "learning_rate": 1.05e-05,
      "loss": 8.9643,
      "norm_diff": 3.5766,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 16510.0582,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.999,
      "query_norm": 2.1801,
      "queue_k_norm": 5.7695,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0797,
      "sent_len_1": 66.9304,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2463,
      "stdk": 0.0873,
      "stdq": 0.0609,
      "stdqueue_k": 0.0881,
      "stdqueue_q": 0.0,
      "step": 2100
    },
    {
      "accuracy": 26.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 8.73,
      "doc_norm": 5.5551,
      "encoder_q-embeddings": 11415.8613,
      "encoder_q-layer.0": 9797.6904,
      "encoder_q-layer.1": 11668.1025,
      "encoder_q-layer.10": 9421.3262,
      "encoder_q-layer.11": 10732.5957,
      "encoder_q-layer.2": 13241.7676,
      "encoder_q-layer.3": 13175.9482,
      "encoder_q-layer.4": 13794.2217,
      "encoder_q-layer.5": 14288.5381,
      "encoder_q-layer.6": 14483.4561,
      "encoder_q-layer.7": 15834.4404,
      "encoder_q-layer.8": 15432.7031,
      "encoder_q-layer.9": 6712.8267,
      "epoch": 0.02,
      "inbatch_neg_score": 0.4619,
      "inbatch_pos_score": 1.0166,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 8.73,
      "norm_diff": 3.3905,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 18184.4984,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.4646,
      "query_norm": 2.1646,
      "queue_k_norm": 5.597,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7649,
      "sent_len_1": 66.7646,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2637,
      "stdk": 0.0842,
      "stdq": 0.0591,
      "stdqueue_k": 0.085,
      "stdqueue_q": 0.0,
      "step": 2200
    },
    {
      "accuracy": 25.293,
      "active_queue_size": 16384.0,
      "cl_loss": 8.3553,
      "doc_norm": 5.35,
      "encoder_q-embeddings": 5705.1411,
      "encoder_q-layer.0": 4769.1045,
      "encoder_q-layer.1": 5296.2285,
      "encoder_q-layer.10": 11568.3174,
      "encoder_q-layer.11": 11316.4355,
      "encoder_q-layer.2": 6202.4102,
      "encoder_q-layer.3": 6571.5391,
      "encoder_q-layer.4": 6900.2842,
      "encoder_q-layer.5": 6537.2485,
      "encoder_q-layer.6": 6244.1709,
      "encoder_q-layer.7": 6335.9824,
      "encoder_q-layer.8": 7365.1191,
      "encoder_q-layer.9": 7952.9458,
      "epoch": 0.02,
      "inbatch_neg_score": 1.2201,
      "inbatch_pos_score": 1.7646,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 8.3553,
      "norm_diff": 3.1451,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 10070.4756,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 1.2119,
      "query_norm": 2.2049,
      "queue_k_norm": 5.392,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8673,
      "sent_len_1": 67.0374,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.26,
      "stdk": 0.0814,
      "stdq": 0.062,
      "stdqueue_k": 0.0821,
      "stdqueue_q": 0.0,
      "step": 2300
    },
    {
      "accuracy": 25.5859,
      "active_queue_size": 16384.0,
      "cl_loss": 7.9879,
      "doc_norm": 5.1648,
      "encoder_q-embeddings": 15436.998,
      "encoder_q-layer.0": 13625.4375,
      "encoder_q-layer.1": 12128.8613,
      "encoder_q-layer.10": 24416.2891,
      "encoder_q-layer.11": 19975.4102,
      "encoder_q-layer.2": 12006.7705,
      "encoder_q-layer.3": 11388.1592,
      "encoder_q-layer.4": 10859.7891,
      "encoder_q-layer.5": 11582.8447,
      "encoder_q-layer.6": 11344.1211,
      "encoder_q-layer.7": 13919.2764,
      "encoder_q-layer.8": 16824.6543,
      "encoder_q-layer.9": 18599.9883,
      "epoch": 0.02,
      "inbatch_neg_score": 0.2313,
      "inbatch_pos_score": 0.7441,
      "learning_rate": 1.2e-05,
      "loss": 7.9879,
      "norm_diff": 3.0592,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 21269.5337,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.2189,
      "query_norm": 2.1056,
      "queue_k_norm": 5.1696,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8346,
      "sent_len_1": 66.725,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.605,
      "stdk": 0.0783,
      "stdq": 0.0585,
      "stdqueue_k": 0.0792,
      "stdqueue_q": 0.0,
      "step": 2400
    },
    {
      "accuracy": 30.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 7.7803,
      "doc_norm": 4.9201,
      "encoder_q-embeddings": 3455.4053,
      "encoder_q-layer.0": 2773.4604,
      "encoder_q-layer.1": 3353.9885,
      "encoder_q-layer.10": 28194.9023,
      "encoder_q-layer.11": 22591.1641,
      "encoder_q-layer.2": 4105.6631,
      "encoder_q-layer.3": 4680.0664,
      "encoder_q-layer.4": 6010.9863,
      "encoder_q-layer.5": 7773.3911,
      "encoder_q-layer.6": 9756.8174,
      "encoder_q-layer.7": 12424.1201,
      "encoder_q-layer.8": 15527.6143,
      "encoder_q-layer.9": 18891.5625,
      "epoch": 0.02,
      "inbatch_neg_score": 1.3595,
      "inbatch_pos_score": 1.9268,
      "learning_rate": 1.25e-05,
      "loss": 7.7803,
      "norm_diff": 2.7658,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 17051.0924,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 1.3516,
      "query_norm": 2.1544,
      "queue_k_norm": 4.9519,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0805,
      "sent_len_1": 66.7136,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9263,
      "stdk": 0.0764,
      "stdq": 0.0601,
      "stdqueue_k": 0.0768,
      "stdqueue_q": 0.0,
      "step": 2500
    },
    {
      "accuracy": 31.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 7.5198,
      "doc_norm": 4.7222,
      "encoder_q-embeddings": 3696.8184,
      "encoder_q-layer.0": 2900.1475,
      "encoder_q-layer.1": 3421.0701,
      "encoder_q-layer.10": 30141.0938,
      "encoder_q-layer.11": 22744.1016,
      "encoder_q-layer.2": 4143.0332,
      "encoder_q-layer.3": 4812.7021,
      "encoder_q-layer.4": 6110.4731,
      "encoder_q-layer.5": 7914.8164,
      "encoder_q-layer.6": 9566.9443,
      "encoder_q-layer.7": 12384.1309,
      "encoder_q-layer.8": 14977.874,
      "encoder_q-layer.9": 19830.0273,
      "epoch": 0.03,
      "inbatch_neg_score": 0.6084,
      "inbatch_pos_score": 1.1611,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 7.5198,
      "norm_diff": 2.6502,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 17644.0161,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.603,
      "query_norm": 2.072,
      "queue_k_norm": 4.7377,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0472,
      "sent_len_1": 66.7499,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4512,
      "stdk": 0.0744,
      "stdq": 0.0587,
      "stdqueue_k": 0.0747,
      "stdqueue_q": 0.0,
      "step": 2600
    },
    {
      "accuracy": 30.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 7.4157,
      "doc_norm": 4.5044,
      "encoder_q-embeddings": 3457.7356,
      "encoder_q-layer.0": 2833.4492,
      "encoder_q-layer.1": 3381.8723,
      "encoder_q-layer.10": 13088.1094,
      "encoder_q-layer.11": 11289.1211,
      "encoder_q-layer.2": 3954.1545,
      "encoder_q-layer.3": 4109.5303,
      "encoder_q-layer.4": 4382.0474,
      "encoder_q-layer.5": 4247.3687,
      "encoder_q-layer.6": 3911.4045,
      "encoder_q-layer.7": 4476.8921,
      "encoder_q-layer.8": 6050.8433,
      "encoder_q-layer.9": 8340.6006,
      "epoch": 0.03,
      "inbatch_neg_score": 0.8111,
      "inbatch_pos_score": 1.3613,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 7.4157,
      "norm_diff": 2.3987,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8398.5192,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.8086,
      "query_norm": 2.1057,
      "queue_k_norm": 4.5147,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8759,
      "sent_len_1": 66.7117,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4525,
      "stdk": 0.0726,
      "stdq": 0.0604,
      "stdqueue_k": 0.0726,
      "stdqueue_q": 0.0,
      "step": 2700
    },
    {
      "accuracy": 30.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 7.2024,
      "doc_norm": 4.3026,
      "encoder_q-embeddings": 8673.3047,
      "encoder_q-layer.0": 6639.1504,
      "encoder_q-layer.1": 8210.8076,
      "encoder_q-layer.10": 38858.8438,
      "encoder_q-layer.11": 26500.7461,
      "encoder_q-layer.2": 9227.0078,
      "encoder_q-layer.3": 10849.252,
      "encoder_q-layer.4": 13533.3086,
      "encoder_q-layer.5": 18044.5996,
      "encoder_q-layer.6": 20056.5176,
      "encoder_q-layer.7": 23418.5762,
      "encoder_q-layer.8": 24949.8535,
      "encoder_q-layer.9": 29128.8984,
      "epoch": 0.03,
      "inbatch_neg_score": 0.5704,
      "inbatch_pos_score": 1.1172,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 7.2024,
      "norm_diff": 2.1696,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 27044.227,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.5586,
      "query_norm": 2.1329,
      "queue_k_norm": 4.3121,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7775,
      "sent_len_1": 67.0262,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8275,
      "stdk": 0.0706,
      "stdq": 0.0602,
      "stdqueue_k": 0.071,
      "stdqueue_q": 0.0,
      "step": 2800
    },
    {
      "accuracy": 28.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 7.0893,
      "doc_norm": 4.0893,
      "encoder_q-embeddings": 3281.7393,
      "encoder_q-layer.0": 2514.0898,
      "encoder_q-layer.1": 3148.23,
      "encoder_q-layer.10": 29085.6016,
      "encoder_q-layer.11": 19870.832,
      "encoder_q-layer.2": 3972.3391,
      "encoder_q-layer.3": 4397.252,
      "encoder_q-layer.4": 5673.6533,
      "encoder_q-layer.5": 7040.9106,
      "encoder_q-layer.6": 9687.0752,
      "encoder_q-layer.7": 13598.0859,
      "encoder_q-layer.8": 16965.3262,
      "encoder_q-layer.9": 22972.1738,
      "epoch": 0.03,
      "inbatch_neg_score": 0.5606,
      "inbatch_pos_score": 1.0781,
      "learning_rate": 1.45e-05,
      "loss": 7.0893,
      "norm_diff": 2.0064,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 17726.3485,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.5522,
      "query_norm": 2.0828,
      "queue_k_norm": 4.1192,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9662,
      "sent_len_1": 66.8527,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.245,
      "stdk": 0.0688,
      "stdq": 0.0598,
      "stdqueue_k": 0.0694,
      "stdqueue_q": 0.0,
      "step": 2900
    },
    {
      "accuracy": 30.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 6.9011,
      "doc_norm": 3.8799,
      "encoder_q-embeddings": 2068.6372,
      "encoder_q-layer.0": 1822.0653,
      "encoder_q-layer.1": 1984.5815,
      "encoder_q-layer.10": 3455.0557,
      "encoder_q-layer.11": 4377.0542,
      "encoder_q-layer.2": 2257.0464,
      "encoder_q-layer.3": 2372.0884,
      "encoder_q-layer.4": 2447.9189,
      "encoder_q-layer.5": 2318.8079,
      "encoder_q-layer.6": 2038.561,
      "encoder_q-layer.7": 1868.6631,
      "encoder_q-layer.8": 2329.9836,
      "encoder_q-layer.9": 2601.5042,
      "epoch": 0.03,
      "inbatch_neg_score": 0.7292,
      "inbatch_pos_score": 1.252,
      "learning_rate": 1.5e-05,
      "loss": 6.9011,
      "norm_diff": 1.786,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3556.0123,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7207,
      "query_norm": 2.0939,
      "queue_k_norm": 3.9339,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8866,
      "sent_len_1": 66.6738,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5475,
      "stdk": 0.068,
      "stdq": 0.0589,
      "stdqueue_k": 0.0684,
      "stdqueue_q": 0.0,
      "step": 3000
    },
    {
      "accuracy": 35.0586,
      "active_queue_size": 16384.0,
      "cl_loss": 6.6907,
      "doc_norm": 3.7704,
      "encoder_q-embeddings": 2260.564,
      "encoder_q-layer.0": 1817.0837,
      "encoder_q-layer.1": 2012.3044,
      "encoder_q-layer.10": 3071.0916,
      "encoder_q-layer.11": 4246.3877,
      "encoder_q-layer.2": 2267.8101,
      "encoder_q-layer.3": 2246.3926,
      "encoder_q-layer.4": 2442.2295,
      "encoder_q-layer.5": 2643.29,
      "encoder_q-layer.6": 2247.3586,
      "encoder_q-layer.7": 2094.6328,
      "encoder_q-layer.8": 2405.6636,
      "encoder_q-layer.9": 2338.0364,
      "epoch": 0.03,
      "inbatch_neg_score": 0.6745,
      "inbatch_pos_score": 1.2373,
      "learning_rate": 1.55e-05,
      "loss": 6.6907,
      "norm_diff": 1.7345,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3586.7928,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.667,
      "query_norm": 2.036,
      "queue_k_norm": 3.7682,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0487,
      "sent_len_1": 66.9252,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2887,
      "stdk": 0.0667,
      "stdq": 0.0575,
      "stdqueue_k": 0.0671,
      "stdqueue_q": 0.0,
      "step": 3100
    },
    {
      "accuracy": 32.9102,
      "active_queue_size": 16384.0,
      "cl_loss": 6.597,
      "doc_norm": 3.5943,
      "encoder_q-embeddings": 3144.855,
      "encoder_q-layer.0": 2712.8218,
      "encoder_q-layer.1": 3549.4385,
      "encoder_q-layer.10": 47150.1172,
      "encoder_q-layer.11": 30145.8125,
      "encoder_q-layer.2": 4702.8623,
      "encoder_q-layer.3": 5742.6206,
      "encoder_q-layer.4": 7973.0322,
      "encoder_q-layer.5": 10944.25,
      "encoder_q-layer.6": 14880.5459,
      "encoder_q-layer.7": 21088.8008,
      "encoder_q-layer.8": 26943.6953,
      "encoder_q-layer.9": 36047.6016,
      "epoch": 0.03,
      "inbatch_neg_score": 0.4917,
      "inbatch_pos_score": 1.0557,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 6.597,
      "norm_diff": 1.5658,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 27693.8371,
      "preclip_grad_norm_avg": 0.0003,
      "q@queue_neg_score": 0.4836,
      "query_norm": 2.0285,
      "queue_k_norm": 3.6222,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0023,
      "sent_len_1": 66.74,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5863,
      "stdk": 0.0656,
      "stdq": 0.0577,
      "stdqueue_k": 0.0661,
      "stdqueue_q": 0.0,
      "step": 3200
    },
    {
      "accuracy": 33.3984,
      "active_queue_size": 16384.0,
      "cl_loss": 6.5829,
      "doc_norm": 3.4835,
      "encoder_q-embeddings": 2243.1672,
      "encoder_q-layer.0": 1751.4442,
      "encoder_q-layer.1": 2030.1901,
      "encoder_q-layer.10": 18507.3652,
      "encoder_q-layer.11": 12051.4961,
      "encoder_q-layer.2": 2588.4795,
      "encoder_q-layer.3": 2908.519,
      "encoder_q-layer.4": 3505.5547,
      "encoder_q-layer.5": 4559.5269,
      "encoder_q-layer.6": 5131.4033,
      "encoder_q-layer.7": 6847.791,
      "encoder_q-layer.8": 9132.5527,
      "encoder_q-layer.9": 13863.4229,
      "epoch": 0.03,
      "inbatch_neg_score": 0.7994,
      "inbatch_pos_score": 1.3359,
      "learning_rate": 1.65e-05,
      "loss": 6.5829,
      "norm_diff": 1.411,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 10611.4136,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7944,
      "query_norm": 2.0726,
      "queue_k_norm": 3.4847,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0437,
      "sent_len_1": 66.9894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9712,
      "stdk": 0.0655,
      "stdq": 0.0578,
      "stdqueue_k": 0.0652,
      "stdqueue_q": 0.0,
      "step": 3300
    },
    {
      "accuracy": 27.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 6.4904,
      "doc_norm": 3.3554,
      "encoder_q-embeddings": 2840.5635,
      "encoder_q-layer.0": 2248.0698,
      "encoder_q-layer.1": 2843.1338,
      "encoder_q-layer.10": 59699.25,
      "encoder_q-layer.11": 35836.4102,
      "encoder_q-layer.2": 3724.3623,
      "encoder_q-layer.3": 4592.377,
      "encoder_q-layer.4": 6229.6274,
      "encoder_q-layer.5": 8314.9062,
      "encoder_q-layer.6": 11343.9004,
      "encoder_q-layer.7": 17443.4434,
      "encoder_q-layer.8": 25554.8691,
      "encoder_q-layer.9": 42752.9922,
      "epoch": 0.03,
      "inbatch_neg_score": 0.3627,
      "inbatch_pos_score": 0.8789,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.4904,
      "norm_diff": 1.2817,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 31241.8158,
      "preclip_grad_norm_avg": 0.0003,
      "q@queue_neg_score": 0.355,
      "query_norm": 2.0737,
      "queue_k_norm": 3.353,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0472,
      "sent_len_1": 66.6085,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4275,
      "stdk": 0.0645,
      "stdq": 0.0582,
      "stdqueue_k": 0.0646,
      "stdqueue_q": 0.0,
      "step": 3400
    },
    {
      "accuracy": 36.2305,
      "active_queue_size": 16384.0,
      "cl_loss": 6.3886,
      "doc_norm": 3.2186,
      "encoder_q-embeddings": 2580.5237,
      "encoder_q-layer.0": 2150.9041,
      "encoder_q-layer.1": 2411.2959,
      "encoder_q-layer.10": 36668.8086,
      "encoder_q-layer.11": 21914.4414,
      "encoder_q-layer.2": 3024.4646,
      "encoder_q-layer.3": 3150.8928,
      "encoder_q-layer.4": 4268.6475,
      "encoder_q-layer.5": 5604.1978,
      "encoder_q-layer.6": 7717.9839,
      "encoder_q-layer.7": 11863.9609,
      "encoder_q-layer.8": 16584.2578,
      "encoder_q-layer.9": 26262.9414,
      "epoch": 0.03,
      "inbatch_neg_score": 0.4726,
      "inbatch_pos_score": 1.0312,
      "learning_rate": 1.75e-05,
      "loss": 6.3886,
      "norm_diff": 1.2159,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 19564.7421,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.4609,
      "query_norm": 2.0027,
      "queue_k_norm": 3.2386,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0539,
      "sent_len_1": 66.8938,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.505,
      "stdk": 0.0632,
      "stdq": 0.0555,
      "stdqueue_k": 0.0637,
      "stdqueue_q": 0.0,
      "step": 3500
    },
    {
      "accuracy": 33.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 6.2282,
      "doc_norm": 3.1343,
      "encoder_q-embeddings": 2132.3774,
      "encoder_q-layer.0": 1764.6981,
      "encoder_q-layer.1": 2085.8801,
      "encoder_q-layer.10": 14154.418,
      "encoder_q-layer.11": 9012.9336,
      "encoder_q-layer.2": 2599.1843,
      "encoder_q-layer.3": 2678.5591,
      "encoder_q-layer.4": 3037.2458,
      "encoder_q-layer.5": 3775.6316,
      "encoder_q-layer.6": 4501.959,
      "encoder_q-layer.7": 5784.7246,
      "encoder_q-layer.8": 7148.4619,
      "encoder_q-layer.9": 10115.1875,
      "epoch": 0.04,
      "inbatch_neg_score": 0.5256,
      "inbatch_pos_score": 1.0566,
      "learning_rate": 1.8e-05,
      "loss": 6.2282,
      "norm_diff": 1.1358,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8275.9341,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.52,
      "query_norm": 1.9985,
      "queue_k_norm": 3.1307,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9419,
      "sent_len_1": 66.7182,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4338,
      "stdk": 0.0629,
      "stdq": 0.0568,
      "stdqueue_k": 0.0631,
      "stdqueue_q": 0.0,
      "step": 3600
    },
    {
      "accuracy": 34.668,
      "active_queue_size": 16384.0,
      "cl_loss": 6.0376,
      "doc_norm": 3.0305,
      "encoder_q-embeddings": 1485.4427,
      "encoder_q-layer.0": 1242.4307,
      "encoder_q-layer.1": 1339.1317,
      "encoder_q-layer.10": 10410.7363,
      "encoder_q-layer.11": 7364.2119,
      "encoder_q-layer.2": 1518.7808,
      "encoder_q-layer.3": 1618.0499,
      "encoder_q-layer.4": 1869.7915,
      "encoder_q-layer.5": 2123.1375,
      "encoder_q-layer.6": 2919.5051,
      "encoder_q-layer.7": 4473.189,
      "encoder_q-layer.8": 6227.3149,
      "encoder_q-layer.9": 8197.6689,
      "epoch": 0.04,
      "inbatch_neg_score": 0.5893,
      "inbatch_pos_score": 1.1406,
      "learning_rate": 1.85e-05,
      "loss": 6.0376,
      "norm_diff": 1.0669,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6339.8204,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5869,
      "query_norm": 1.9636,
      "queue_k_norm": 3.0337,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9507,
      "sent_len_1": 66.8867,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5275,
      "stdk": 0.0624,
      "stdq": 0.0556,
      "stdqueue_k": 0.0624,
      "stdqueue_q": 0.0,
      "step": 3700
    },
    {
      "accuracy": 32.3242,
      "active_queue_size": 16384.0,
      "cl_loss": 5.9922,
      "doc_norm": 2.9273,
      "encoder_q-embeddings": 2208.2341,
      "encoder_q-layer.0": 1750.5997,
      "encoder_q-layer.1": 1943.4844,
      "encoder_q-layer.10": 6463.561,
      "encoder_q-layer.11": 5075.4531,
      "encoder_q-layer.2": 2299.377,
      "encoder_q-layer.3": 2167.1616,
      "encoder_q-layer.4": 2086.2922,
      "encoder_q-layer.5": 2002.0435,
      "encoder_q-layer.6": 1960.4974,
      "encoder_q-layer.7": 2167.9165,
      "encoder_q-layer.8": 3067.0,
      "encoder_q-layer.9": 4594.6797,
      "epoch": 0.04,
      "inbatch_neg_score": 0.5818,
      "inbatch_pos_score": 1.1064,
      "learning_rate": 1.9e-05,
      "loss": 5.9922,
      "norm_diff": 1.002,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4362.9719,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5786,
      "query_norm": 1.9254,
      "queue_k_norm": 2.9386,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9592,
      "sent_len_1": 66.7808,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1975,
      "stdk": 0.0622,
      "stdq": 0.0549,
      "stdqueue_k": 0.0619,
      "stdqueue_q": 0.0,
      "step": 3800
    },
    {
      "accuracy": 33.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 5.7114,
      "doc_norm": 2.8333,
      "encoder_q-embeddings": 1736.1433,
      "encoder_q-layer.0": 1374.8561,
      "encoder_q-layer.1": 1592.6689,
      "encoder_q-layer.10": 11937.3896,
      "encoder_q-layer.11": 8270.9404,
      "encoder_q-layer.2": 1960.9795,
      "encoder_q-layer.3": 2147.1021,
      "encoder_q-layer.4": 2500.9417,
      "encoder_q-layer.5": 3594.5098,
      "encoder_q-layer.6": 4240.7881,
      "encoder_q-layer.7": 5510.0625,
      "encoder_q-layer.8": 6776.4927,
      "encoder_q-layer.9": 8980.5752,
      "epoch": 0.04,
      "inbatch_neg_score": 0.5003,
      "inbatch_pos_score": 1.0254,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 5.7114,
      "norm_diff": 0.967,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7373.5008,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4946,
      "query_norm": 1.8663,
      "queue_k_norm": 2.8488,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9755,
      "sent_len_1": 66.6361,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7438,
      "stdk": 0.0615,
      "stdq": 0.0531,
      "stdqueue_k": 0.0614,
      "stdqueue_q": 0.0,
      "step": 3900
    },
    {
      "accuracy": 34.5703,
      "active_queue_size": 16384.0,
      "cl_loss": 5.5284,
      "doc_norm": 2.7466,
      "encoder_q-embeddings": 1181.4463,
      "encoder_q-layer.0": 973.2521,
      "encoder_q-layer.1": 1057.1333,
      "encoder_q-layer.10": 2191.6152,
      "encoder_q-layer.11": 3180.4678,
      "encoder_q-layer.2": 1242.3732,
      "encoder_q-layer.3": 1307.5814,
      "encoder_q-layer.4": 1298.3933,
      "encoder_q-layer.5": 1265.6052,
      "encoder_q-layer.6": 1347.4902,
      "encoder_q-layer.7": 1212.5006,
      "encoder_q-layer.8": 1421.6805,
      "encoder_q-layer.9": 1296.719,
      "epoch": 0.04,
      "inbatch_neg_score": 0.3562,
      "inbatch_pos_score": 0.8589,
      "learning_rate": 2e-05,
      "loss": 5.5284,
      "norm_diff": 0.9317,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2207.2403,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3499,
      "query_norm": 1.8149,
      "queue_k_norm": 2.7519,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1435,
      "sent_len_1": 66.6089,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5037,
      "stdk": 0.0602,
      "stdq": 0.0514,
      "stdqueue_k": 0.0607,
      "stdqueue_q": 0.0,
      "step": 4000
    },
    {
      "accuracy": 35.9375,
      "active_queue_size": 16384.0,
      "cl_loss": 5.3576,
      "doc_norm": 2.6494,
      "encoder_q-embeddings": 4444.9224,
      "encoder_q-layer.0": 3516.6235,
      "encoder_q-layer.1": 3495.1462,
      "encoder_q-layer.10": 9908.4307,
      "encoder_q-layer.11": 8489.3652,
      "encoder_q-layer.2": 3595.5732,
      "encoder_q-layer.3": 3189.418,
      "encoder_q-layer.4": 2606.887,
      "encoder_q-layer.5": 2302.4924,
      "encoder_q-layer.6": 2318.9956,
      "encoder_q-layer.7": 3246.269,
      "encoder_q-layer.8": 4711.1904,
      "encoder_q-layer.9": 7373.3809,
      "epoch": 0.04,
      "inbatch_neg_score": 0.3337,
      "inbatch_pos_score": 0.8481,
      "learning_rate": 2.05e-05,
      "loss": 5.3576,
      "norm_diff": 0.8883,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7133.1759,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3262,
      "query_norm": 1.7611,
      "queue_k_norm": 2.6589,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9338,
      "sent_len_1": 66.8712,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2725,
      "stdk": 0.0597,
      "stdq": 0.0497,
      "stdqueue_k": 0.0602,
      "stdqueue_q": 0.0,
      "step": 4100
    },
    {
      "accuracy": 37.1094,
      "active_queue_size": 16384.0,
      "cl_loss": 5.2551,
      "doc_norm": 2.5558,
      "encoder_q-embeddings": 3487.4387,
      "encoder_q-layer.0": 2825.7168,
      "encoder_q-layer.1": 3025.0967,
      "encoder_q-layer.10": 7373.6807,
      "encoder_q-layer.11": 7375.3452,
      "encoder_q-layer.2": 3811.0713,
      "encoder_q-layer.3": 3825.4785,
      "encoder_q-layer.4": 4018.6855,
      "encoder_q-layer.5": 3618.0486,
      "encoder_q-layer.6": 3317.5989,
      "encoder_q-layer.7": 3892.5146,
      "encoder_q-layer.8": 4671.9717,
      "encoder_q-layer.9": 5972.8599,
      "epoch": 0.04,
      "inbatch_neg_score": 0.3256,
      "inbatch_pos_score": 0.8506,
      "learning_rate": 2.1e-05,
      "loss": 5.2551,
      "norm_diff": 0.7984,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6449.7728,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3188,
      "query_norm": 1.7574,
      "queue_k_norm": 2.5641,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8171,
      "sent_len_1": 66.3805,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.36,
      "stdk": 0.0591,
      "stdq": 0.0492,
      "stdqueue_k": 0.0595,
      "stdqueue_q": 0.0,
      "step": 4200
    },
    {
      "accuracy": 38.6719,
      "active_queue_size": 16384.0,
      "cl_loss": 5.097,
      "doc_norm": 2.4658,
      "encoder_q-embeddings": 3355.0881,
      "encoder_q-layer.0": 2720.1055,
      "encoder_q-layer.1": 2829.0764,
      "encoder_q-layer.10": 9603.5742,
      "encoder_q-layer.11": 9965.4502,
      "encoder_q-layer.2": 3118.2839,
      "encoder_q-layer.3": 3078.145,
      "encoder_q-layer.4": 3051.335,
      "encoder_q-layer.5": 2807.0249,
      "encoder_q-layer.6": 2775.0549,
      "encoder_q-layer.7": 3276.5684,
      "encoder_q-layer.8": 3690.5154,
      "encoder_q-layer.9": 5789.9683,
      "epoch": 0.04,
      "inbatch_neg_score": 0.3683,
      "inbatch_pos_score": 0.9111,
      "learning_rate": 2.15e-05,
      "loss": 5.097,
      "norm_diff": 0.7691,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6811.2713,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.364,
      "query_norm": 1.6967,
      "queue_k_norm": 2.469,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8104,
      "sent_len_1": 66.5864,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.1463,
      "stdk": 0.0586,
      "stdq": 0.0477,
      "stdqueue_k": 0.0587,
      "stdqueue_q": 0.0,
      "step": 4300
    },
    {
      "accuracy": 42.1875,
      "active_queue_size": 16384.0,
      "cl_loss": 5.0303,
      "doc_norm": 2.3764,
      "encoder_q-embeddings": 6556.6504,
      "encoder_q-layer.0": 5855.1436,
      "encoder_q-layer.1": 6170.3691,
      "encoder_q-layer.10": 5702.2886,
      "encoder_q-layer.11": 8014.5146,
      "encoder_q-layer.2": 7092.2261,
      "encoder_q-layer.3": 7035.042,
      "encoder_q-layer.4": 6688.1768,
      "encoder_q-layer.5": 5860.8062,
      "encoder_q-layer.6": 4483.292,
      "encoder_q-layer.7": 3102.7166,
      "encoder_q-layer.8": 2950.2827,
      "encoder_q-layer.9": 3836.657,
      "epoch": 0.04,
      "inbatch_neg_score": 0.302,
      "inbatch_pos_score": 0.8628,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 5.0303,
      "norm_diff": 0.6417,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8785.1679,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2986,
      "query_norm": 1.7347,
      "queue_k_norm": 2.3814,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0837,
      "sent_len_1": 66.7005,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2312,
      "stdk": 0.058,
      "stdq": 0.0487,
      "stdqueue_k": 0.058,
      "stdqueue_q": 0.0,
      "step": 4400
    },
    {
      "accuracy": 37.8906,
      "active_queue_size": 16384.0,
      "cl_loss": 4.9001,
      "doc_norm": 2.2965,
      "encoder_q-embeddings": 3147.5359,
      "encoder_q-layer.0": 2737.2007,
      "encoder_q-layer.1": 2862.8865,
      "encoder_q-layer.10": 5668.4595,
      "encoder_q-layer.11": 6981.2539,
      "encoder_q-layer.2": 3117.4448,
      "encoder_q-layer.3": 3161.2021,
      "encoder_q-layer.4": 3073.3816,
      "encoder_q-layer.5": 3034.4358,
      "encoder_q-layer.6": 3033.4958,
      "encoder_q-layer.7": 3088.0544,
      "encoder_q-layer.8": 3275.9072,
      "encoder_q-layer.9": 4041.3496,
      "epoch": 0.04,
      "inbatch_neg_score": 0.3819,
      "inbatch_pos_score": 0.9116,
      "learning_rate": 2.25e-05,
      "loss": 4.9001,
      "norm_diff": 0.5894,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5419.6375,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3784,
      "query_norm": 1.7072,
      "queue_k_norm": 2.2959,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9556,
      "sent_len_1": 66.6556,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.135,
      "stdk": 0.0573,
      "stdq": 0.0479,
      "stdqueue_k": 0.0572,
      "stdqueue_q": 0.0,
      "step": 4500
    },
    {
      "accuracy": 37.9883,
      "active_queue_size": 16384.0,
      "cl_loss": 4.85,
      "doc_norm": 2.2089,
      "encoder_q-embeddings": 2241.606,
      "encoder_q-layer.0": 1903.6028,
      "encoder_q-layer.1": 1942.4916,
      "encoder_q-layer.10": 7102.4634,
      "encoder_q-layer.11": 7476.9355,
      "encoder_q-layer.2": 2427.8164,
      "encoder_q-layer.3": 2399.8853,
      "encoder_q-layer.4": 2293.1809,
      "encoder_q-layer.5": 2204.4631,
      "encoder_q-layer.6": 2002.8632,
      "encoder_q-layer.7": 2049.6877,
      "encoder_q-layer.8": 2889.448,
      "encoder_q-layer.9": 4982.8184,
      "epoch": 0.04,
      "inbatch_neg_score": 0.3638,
      "inbatch_pos_score": 0.874,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 4.85,
      "norm_diff": 0.5351,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5213.9016,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3577,
      "query_norm": 1.6738,
      "queue_k_norm": 2.2199,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9349,
      "sent_len_1": 66.7083,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6887,
      "stdk": 0.0564,
      "stdq": 0.047,
      "stdqueue_k": 0.0567,
      "stdqueue_q": 0.0,
      "step": 4600
    },
    {
      "accuracy": 41.1133,
      "active_queue_size": 16384.0,
      "cl_loss": 4.7714,
      "doc_norm": 2.1498,
      "encoder_q-embeddings": 6564.1084,
      "encoder_q-layer.0": 6634.3052,
      "encoder_q-layer.1": 6761.9146,
      "encoder_q-layer.10": 9209.4854,
      "encoder_q-layer.11": 9669.5732,
      "encoder_q-layer.2": 6862.7139,
      "encoder_q-layer.3": 6564.3008,
      "encoder_q-layer.4": 6077.4072,
      "encoder_q-layer.5": 6553.1069,
      "encoder_q-layer.6": 6518.9316,
      "encoder_q-layer.7": 6995.1274,
      "encoder_q-layer.8": 8186.9277,
      "encoder_q-layer.9": 8342.873,
      "epoch": 0.05,
      "inbatch_neg_score": 0.3504,
      "inbatch_pos_score": 0.9033,
      "learning_rate": 2.35e-05,
      "loss": 4.7714,
      "norm_diff": 0.489,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 10581.7047,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3491,
      "query_norm": 1.6608,
      "queue_k_norm": 2.1476,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0941,
      "sent_len_1": 66.9886,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6337,
      "stdk": 0.0559,
      "stdq": 0.0467,
      "stdqueue_k": 0.056,
      "stdqueue_q": 0.0,
      "step": 4700
    },
    {
      "accuracy": 42.0898,
      "active_queue_size": 16384.0,
      "cl_loss": 4.7112,
      "doc_norm": 2.0855,
      "encoder_q-embeddings": 2036.2462,
      "encoder_q-layer.0": 1646.6766,
      "encoder_q-layer.1": 1728.7375,
      "encoder_q-layer.10": 5440.0532,
      "encoder_q-layer.11": 6308.624,
      "encoder_q-layer.2": 1917.5059,
      "encoder_q-layer.3": 2126.0156,
      "encoder_q-layer.4": 2144.7031,
      "encoder_q-layer.5": 2089.2493,
      "encoder_q-layer.6": 1843.7428,
      "encoder_q-layer.7": 1918.8691,
      "encoder_q-layer.8": 2671.3538,
      "encoder_q-layer.9": 4141.521,
      "epoch": 0.05,
      "inbatch_neg_score": 0.4167,
      "inbatch_pos_score": 0.9746,
      "learning_rate": 2.4e-05,
      "loss": 4.7112,
      "norm_diff": 0.3937,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4422.2589,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4131,
      "query_norm": 1.6918,
      "queue_k_norm": 2.0877,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9682,
      "sent_len_1": 66.6893,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.665,
      "stdk": 0.0553,
      "stdq": 0.0464,
      "stdqueue_k": 0.0554,
      "stdqueue_q": 0.0,
      "step": 4800
    },
    {
      "accuracy": 39.4531,
      "active_queue_size": 16384.0,
      "cl_loss": 4.6952,
      "doc_norm": 2.0254,
      "encoder_q-embeddings": 2260.6055,
      "encoder_q-layer.0": 1747.516,
      "encoder_q-layer.1": 1915.866,
      "encoder_q-layer.10": 5871.9648,
      "encoder_q-layer.11": 6407.9785,
      "encoder_q-layer.2": 2054.1997,
      "encoder_q-layer.3": 2084.9241,
      "encoder_q-layer.4": 2016.7483,
      "encoder_q-layer.5": 1914.6471,
      "encoder_q-layer.6": 1975.2855,
      "encoder_q-layer.7": 2436.4026,
      "encoder_q-layer.8": 3997.8586,
      "encoder_q-layer.9": 4868.1802,
      "epoch": 0.05,
      "inbatch_neg_score": 0.424,
      "inbatch_pos_score": 0.9463,
      "learning_rate": 2.45e-05,
      "loss": 4.6952,
      "norm_diff": 0.3533,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4766.1374,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4189,
      "query_norm": 1.6721,
      "queue_k_norm": 2.0322,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1227,
      "sent_len_1": 66.7986,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5975,
      "stdk": 0.0548,
      "stdq": 0.0467,
      "stdqueue_k": 0.0549,
      "stdqueue_q": 0.0,
      "step": 4900
    },
    {
      "accuracy": 42.2852,
      "active_queue_size": 16384.0,
      "cl_loss": 4.6598,
      "doc_norm": 1.9804,
      "encoder_q-embeddings": 2474.8286,
      "encoder_q-layer.0": 2025.771,
      "encoder_q-layer.1": 2376.1453,
      "encoder_q-layer.10": 3372.7251,
      "encoder_q-layer.11": 4901.3452,
      "encoder_q-layer.2": 2857.5166,
      "encoder_q-layer.3": 2724.4297,
      "encoder_q-layer.4": 2537.5715,
      "encoder_q-layer.5": 2245.5276,
      "encoder_q-layer.6": 2253.3992,
      "encoder_q-layer.7": 2428.2729,
      "encoder_q-layer.8": 2391.8108,
      "encoder_q-layer.9": 2464.8755,
      "epoch": 0.05,
      "inbatch_neg_score": 0.4555,
      "inbatch_pos_score": 1.0059,
      "learning_rate": 2.5e-05,
      "loss": 4.6598,
      "norm_diff": 0.3069,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4088.4713,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4526,
      "query_norm": 1.6735,
      "queue_k_norm": 1.9827,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9284,
      "sent_len_1": 66.7867,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.915,
      "stdk": 0.0541,
      "stdq": 0.0471,
      "stdqueue_k": 0.0543,
      "stdqueue_q": 0.0,
      "step": 5000
    },
    {
      "accuracy": 40.625,
      "active_queue_size": 16384.0,
      "cl_loss": 4.5802,
      "doc_norm": 1.9378,
      "encoder_q-embeddings": 2976.7021,
      "encoder_q-layer.0": 2806.7544,
      "encoder_q-layer.1": 2830.1917,
      "encoder_q-layer.10": 6243.5229,
      "encoder_q-layer.11": 7323.9062,
      "encoder_q-layer.2": 2987.3574,
      "encoder_q-layer.3": 2995.032,
      "encoder_q-layer.4": 3117.6211,
      "encoder_q-layer.5": 3309.5146,
      "encoder_q-layer.6": 3814.915,
      "encoder_q-layer.7": 4304.9277,
      "encoder_q-layer.8": 5453.3848,
      "encoder_q-layer.9": 5783.5195,
      "epoch": 0.05,
      "inbatch_neg_score": 0.4943,
      "inbatch_pos_score": 1.0176,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 4.5802,
      "norm_diff": 0.306,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6146.6297,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4915,
      "query_norm": 1.6317,
      "queue_k_norm": 1.9406,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0704,
      "sent_len_1": 66.8346,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.9988,
      "stdk": 0.0538,
      "stdq": 0.0453,
      "stdqueue_k": 0.0538,
      "stdqueue_q": 0.0,
      "step": 5100
    },
    {
      "accuracy": 41.6992,
      "active_queue_size": 16384.0,
      "cl_loss": 4.562,
      "doc_norm": 1.8988,
      "encoder_q-embeddings": 1632.1571,
      "encoder_q-layer.0": 1287.5238,
      "encoder_q-layer.1": 1409.1388,
      "encoder_q-layer.10": 2776.8232,
      "encoder_q-layer.11": 4452.8403,
      "encoder_q-layer.2": 1621.2954,
      "encoder_q-layer.3": 1620.3961,
      "encoder_q-layer.4": 1598.0752,
      "encoder_q-layer.5": 1551.0533,
      "encoder_q-layer.6": 1457.5011,
      "encoder_q-layer.7": 1420.3085,
      "encoder_q-layer.8": 1681.2561,
      "encoder_q-layer.9": 2085.948,
      "epoch": 0.05,
      "inbatch_neg_score": 0.5335,
      "inbatch_pos_score": 1.0693,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 4.562,
      "norm_diff": 0.2301,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2954.8142,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5278,
      "query_norm": 1.6687,
      "queue_k_norm": 1.9029,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9071,
      "sent_len_1": 66.5305,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3487,
      "stdk": 0.0532,
      "stdq": 0.0471,
      "stdqueue_k": 0.0533,
      "stdqueue_q": 0.0,
      "step": 5200
    },
    {
      "accuracy": 44.2383,
      "active_queue_size": 16384.0,
      "cl_loss": 4.531,
      "doc_norm": 1.866,
      "encoder_q-embeddings": 2742.1235,
      "encoder_q-layer.0": 2085.7517,
      "encoder_q-layer.1": 2317.1113,
      "encoder_q-layer.10": 3458.9331,
      "encoder_q-layer.11": 4643.252,
      "encoder_q-layer.2": 2724.498,
      "encoder_q-layer.3": 2795.9426,
      "encoder_q-layer.4": 2728.2744,
      "encoder_q-layer.5": 2631.1555,
      "encoder_q-layer.6": 2314.4221,
      "encoder_q-layer.7": 2303.2625,
      "encoder_q-layer.8": 2990.0823,
      "encoder_q-layer.9": 3108.1934,
      "epoch": 0.05,
      "inbatch_neg_score": 0.5583,
      "inbatch_pos_score": 1.125,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 4.531,
      "norm_diff": 0.1851,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4237.8587,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5571,
      "query_norm": 1.681,
      "queue_k_norm": 1.8749,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0979,
      "sent_len_1": 66.8008,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1575,
      "stdk": 0.0526,
      "stdq": 0.0467,
      "stdqueue_k": 0.053,
      "stdqueue_q": 0.0,
      "step": 5300
    },
    {
      "accuracy": 41.2109,
      "active_queue_size": 16384.0,
      "cl_loss": 4.4912,
      "doc_norm": 1.8411,
      "encoder_q-embeddings": 1841.1473,
      "encoder_q-layer.0": 1369.3136,
      "encoder_q-layer.1": 1536.9348,
      "encoder_q-layer.10": 5052.1421,
      "encoder_q-layer.11": 5471.6035,
      "encoder_q-layer.2": 1813.4529,
      "encoder_q-layer.3": 1901.3992,
      "encoder_q-layer.4": 2086.4277,
      "encoder_q-layer.5": 2441.6709,
      "encoder_q-layer.6": 3108.3728,
      "encoder_q-layer.7": 4168.1455,
      "encoder_q-layer.8": 5556.9971,
      "encoder_q-layer.9": 5077.3506,
      "epoch": 0.05,
      "inbatch_neg_score": 0.5958,
      "inbatch_pos_score": 1.1475,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 4.4912,
      "norm_diff": 0.1529,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4885.8599,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5913,
      "query_norm": 1.6882,
      "queue_k_norm": 1.8463,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0059,
      "sent_len_1": 66.9315,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8725,
      "stdk": 0.0522,
      "stdq": 0.0471,
      "stdqueue_k": 0.0525,
      "stdqueue_q": 0.0,
      "step": 5400
    },
    {
      "accuracy": 41.3086,
      "active_queue_size": 16384.0,
      "cl_loss": 4.4481,
      "doc_norm": 1.8314,
      "encoder_q-embeddings": 1716.3674,
      "encoder_q-layer.0": 1342.3389,
      "encoder_q-layer.1": 1383.212,
      "encoder_q-layer.10": 2656.1667,
      "encoder_q-layer.11": 4333.1353,
      "encoder_q-layer.2": 1659.5106,
      "encoder_q-layer.3": 1717.2804,
      "encoder_q-layer.4": 1795.4296,
      "encoder_q-layer.5": 2052.7449,
      "encoder_q-layer.6": 2264.1504,
      "encoder_q-layer.7": 2304.5867,
      "encoder_q-layer.8": 2506.0586,
      "encoder_q-layer.9": 1829.9856,
      "epoch": 0.05,
      "inbatch_neg_score": 0.5446,
      "inbatch_pos_score": 1.0859,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 4.4481,
      "norm_diff": 0.1506,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3254.0831,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5415,
      "query_norm": 1.6809,
      "queue_k_norm": 1.825,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9985,
      "sent_len_1": 66.744,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8338,
      "stdk": 0.0523,
      "stdq": 0.0466,
      "stdqueue_k": 0.0522,
      "stdqueue_q": 0.0,
      "step": 5500
    },
    {
      "accuracy": 43.2617,
      "active_queue_size": 16384.0,
      "cl_loss": 4.4349,
      "doc_norm": 1.8098,
      "encoder_q-embeddings": 3650.1765,
      "encoder_q-layer.0": 3051.0227,
      "encoder_q-layer.1": 3041.3528,
      "encoder_q-layer.10": 3014.8486,
      "encoder_q-layer.11": 4155.3086,
      "encoder_q-layer.2": 3269.729,
      "encoder_q-layer.3": 3092.3315,
      "encoder_q-layer.4": 2612.7676,
      "encoder_q-layer.5": 2535.4465,
      "encoder_q-layer.6": 2623.2439,
      "encoder_q-layer.7": 2878.6738,
      "encoder_q-layer.8": 2724.3127,
      "encoder_q-layer.9": 2090.7109,
      "epoch": 0.05,
      "inbatch_neg_score": 0.6292,
      "inbatch_pos_score": 1.1855,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 4.4349,
      "norm_diff": 0.0833,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4482.265,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6255,
      "query_norm": 1.7264,
      "queue_k_norm": 1.8049,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0912,
      "sent_len_1": 66.9919,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4462,
      "stdk": 0.0519,
      "stdq": 0.0477,
      "stdqueue_k": 0.0518,
      "stdqueue_q": 0.0,
      "step": 5600
    },
    {
      "accuracy": 45.0195,
      "active_queue_size": 16384.0,
      "cl_loss": 4.3714,
      "doc_norm": 1.7925,
      "encoder_q-embeddings": 2831.5884,
      "encoder_q-layer.0": 2408.2375,
      "encoder_q-layer.1": 2394.2698,
      "encoder_q-layer.10": 1981.9097,
      "encoder_q-layer.11": 3269.958,
      "encoder_q-layer.2": 2907.8538,
      "encoder_q-layer.3": 2828.4126,
      "encoder_q-layer.4": 2506.3503,
      "encoder_q-layer.5": 2295.1133,
      "encoder_q-layer.6": 1729.2343,
      "encoder_q-layer.7": 1276.2903,
      "encoder_q-layer.8": 1460.2573,
      "encoder_q-layer.9": 1390.9647,
      "epoch": 0.06,
      "inbatch_neg_score": 0.6208,
      "inbatch_pos_score": 1.2041,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 4.3714,
      "norm_diff": 0.063,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3521.8281,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6157,
      "query_norm": 1.7295,
      "queue_k_norm": 1.7907,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9754,
      "sent_len_1": 66.8924,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.105,
      "stdk": 0.0516,
      "stdq": 0.0472,
      "stdqueue_k": 0.0515,
      "stdqueue_q": 0.0,
      "step": 5700
    },
    {
      "accuracy": 41.6992,
      "active_queue_size": 16384.0,
      "cl_loss": 4.3784,
      "doc_norm": 1.7726,
      "encoder_q-embeddings": 1545.6986,
      "encoder_q-layer.0": 1209.6715,
      "encoder_q-layer.1": 1304.6588,
      "encoder_q-layer.10": 3422.6655,
      "encoder_q-layer.11": 5485.7993,
      "encoder_q-layer.2": 1476.8188,
      "encoder_q-layer.3": 1566.3169,
      "encoder_q-layer.4": 1607.6489,
      "encoder_q-layer.5": 1649.0963,
      "encoder_q-layer.6": 1789.5554,
      "encoder_q-layer.7": 2298.0103,
      "encoder_q-layer.8": 3141.1279,
      "encoder_q-layer.9": 2866.8477,
      "epoch": 0.06,
      "inbatch_neg_score": 0.6422,
      "inbatch_pos_score": 1.1885,
      "learning_rate": 2.9e-05,
      "loss": 4.3784,
      "norm_diff": 0.0454,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3552.6526,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6382,
      "query_norm": 1.7271,
      "queue_k_norm": 1.7789,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8259,
      "sent_len_1": 66.7822,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7038,
      "stdk": 0.051,
      "stdq": 0.0468,
      "stdqueue_k": 0.0513,
      "stdqueue_q": 0.0,
      "step": 5800
    },
    {
      "accuracy": 43.0664,
      "active_queue_size": 16384.0,
      "cl_loss": 4.3466,
      "doc_norm": 1.7672,
      "encoder_q-embeddings": 1548.1401,
      "encoder_q-layer.0": 1223.3447,
      "encoder_q-layer.1": 1257.8894,
      "encoder_q-layer.10": 2138.834,
      "encoder_q-layer.11": 3755.585,
      "encoder_q-layer.2": 1496.9287,
      "encoder_q-layer.3": 1468.109,
      "encoder_q-layer.4": 1476.1311,
      "encoder_q-layer.5": 1548.4819,
      "encoder_q-layer.6": 1606.9347,
      "encoder_q-layer.7": 1576.9185,
      "encoder_q-layer.8": 2067.189,
      "encoder_q-layer.9": 1795.5177,
      "epoch": 0.06,
      "inbatch_neg_score": 0.5984,
      "inbatch_pos_score": 1.1543,
      "learning_rate": 2.95e-05,
      "loss": 4.3466,
      "norm_diff": 0.0217,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2727.1422,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5942,
      "query_norm": 1.7455,
      "queue_k_norm": 1.7663,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2334,
      "sent_len_1": 67.0884,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5075,
      "stdk": 0.0509,
      "stdq": 0.0479,
      "stdqueue_k": 0.051,
      "stdqueue_q": 0.0,
      "step": 5900
    },
    {
      "accuracy": 38.8672,
      "active_queue_size": 16384.0,
      "cl_loss": 4.3359,
      "doc_norm": 1.7571,
      "encoder_q-embeddings": 3082.6055,
      "encoder_q-layer.0": 2822.1558,
      "encoder_q-layer.1": 2980.2932,
      "encoder_q-layer.10": 3043.7395,
      "encoder_q-layer.11": 4254.3994,
      "encoder_q-layer.2": 3343.2732,
      "encoder_q-layer.3": 3362.7451,
      "encoder_q-layer.4": 3065.9329,
      "encoder_q-layer.5": 3111.5674,
      "encoder_q-layer.6": 3554.8521,
      "encoder_q-layer.7": 3700.5881,
      "encoder_q-layer.8": 3630.3933,
      "encoder_q-layer.9": 3149.9485,
      "epoch": 0.06,
      "inbatch_neg_score": 0.6268,
      "inbatch_pos_score": 1.1533,
      "learning_rate": 3e-05,
      "loss": 4.3359,
      "norm_diff": 0.0124,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4786.7029,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.623,
      "query_norm": 1.763,
      "queue_k_norm": 1.7564,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.086,
      "sent_len_1": 66.826,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1925,
      "stdk": 0.0507,
      "stdq": 0.0481,
      "stdqueue_k": 0.0509,
      "stdqueue_q": 0.0,
      "step": 6000
    },
    {
      "accuracy": 46.582,
      "active_queue_size": 16384.0,
      "cl_loss": 4.2948,
      "doc_norm": 1.7415,
      "encoder_q-embeddings": 3046.4209,
      "encoder_q-layer.0": 2611.6562,
      "encoder_q-layer.1": 2521.1807,
      "encoder_q-layer.10": 2185.2114,
      "encoder_q-layer.11": 3660.9963,
      "encoder_q-layer.2": 2754.1946,
      "encoder_q-layer.3": 2720.585,
      "encoder_q-layer.4": 2971.613,
      "encoder_q-layer.5": 2835.0684,
      "encoder_q-layer.6": 2358.311,
      "encoder_q-layer.7": 2241.8113,
      "encoder_q-layer.8": 2920.1975,
      "encoder_q-layer.9": 2208.7305,
      "epoch": 0.06,
      "inbatch_neg_score": 0.6275,
      "inbatch_pos_score": 1.2305,
      "learning_rate": 3.05e-05,
      "loss": 4.2948,
      "norm_diff": 0.0129,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3994.0291,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.623,
      "query_norm": 1.7286,
      "queue_k_norm": 1.7454,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6802,
      "sent_len_1": 66.6305,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9263,
      "stdk": 0.0505,
      "stdq": 0.0471,
      "stdqueue_k": 0.0506,
      "stdqueue_q": 0.0,
      "step": 6100
    },
    {
      "accuracy": 45.9961,
      "active_queue_size": 16384.0,
      "cl_loss": 4.2646,
      "doc_norm": 1.7398,
      "encoder_q-embeddings": 6157.5576,
      "encoder_q-layer.0": 5343.7695,
      "encoder_q-layer.1": 4875.6182,
      "encoder_q-layer.10": 4356.9248,
      "encoder_q-layer.11": 7092.4297,
      "encoder_q-layer.2": 4654.2915,
      "encoder_q-layer.3": 4447.3369,
      "encoder_q-layer.4": 4377.6328,
      "encoder_q-layer.5": 4027.7229,
      "encoder_q-layer.6": 4059.5732,
      "encoder_q-layer.7": 3215.4512,
      "encoder_q-layer.8": 3570.8674,
      "encoder_q-layer.9": 3956.7844,
      "epoch": 0.06,
      "inbatch_neg_score": 0.5919,
      "inbatch_pos_score": 1.1689,
      "learning_rate": 3.1e-05,
      "loss": 4.2646,
      "norm_diff": 0.0358,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7103.1361,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5894,
      "query_norm": 1.704,
      "queue_k_norm": 1.7342,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9869,
      "sent_len_1": 66.8279,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2637,
      "stdk": 0.0507,
      "stdq": 0.0453,
      "stdqueue_k": 0.0505,
      "stdqueue_q": 0.0,
      "step": 6200
    },
    {
      "accuracy": 43.1641,
      "active_queue_size": 16384.0,
      "cl_loss": 4.2653,
      "doc_norm": 1.7205,
      "encoder_q-embeddings": 12826.2129,
      "encoder_q-layer.0": 10555.5566,
      "encoder_q-layer.1": 11039.3877,
      "encoder_q-layer.10": 4343.8638,
      "encoder_q-layer.11": 6690.5459,
      "encoder_q-layer.2": 12639.8672,
      "encoder_q-layer.3": 13406.8174,
      "encoder_q-layer.4": 12872.6543,
      "encoder_q-layer.5": 11037.1172,
      "encoder_q-layer.6": 9463.3965,
      "encoder_q-layer.7": 6736.2422,
      "encoder_q-layer.8": 4578.3101,
      "encoder_q-layer.9": 3726.1072,
      "epoch": 0.06,
      "inbatch_neg_score": 0.5246,
      "inbatch_pos_score": 1.0645,
      "learning_rate": 3.15e-05,
      "loss": 4.2653,
      "norm_diff": 0.0195,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 15263.6055,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5176,
      "query_norm": 1.74,
      "queue_k_norm": 1.7215,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1236,
      "sent_len_1": 67.0016,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.3837,
      "stdk": 0.05,
      "stdq": 0.0456,
      "stdqueue_k": 0.0502,
      "stdqueue_q": 0.0,
      "step": 6300
    },
    {
      "accuracy": 40.8203,
      "active_queue_size": 16384.0,
      "cl_loss": 4.2605,
      "doc_norm": 1.7059,
      "encoder_q-embeddings": 4624.5234,
      "encoder_q-layer.0": 3635.5151,
      "encoder_q-layer.1": 3700.6951,
      "encoder_q-layer.10": 5153.3867,
      "encoder_q-layer.11": 8364.6055,
      "encoder_q-layer.2": 4237.5078,
      "encoder_q-layer.3": 3946.3887,
      "encoder_q-layer.4": 3934.7419,
      "encoder_q-layer.5": 3859.0947,
      "encoder_q-layer.6": 3625.865,
      "encoder_q-layer.7": 2988.3765,
      "encoder_q-layer.8": 3384.5146,
      "encoder_q-layer.9": 3572.4299,
      "epoch": 0.06,
      "inbatch_neg_score": 0.5231,
      "inbatch_pos_score": 1.0635,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 4.2605,
      "norm_diff": 0.0683,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6540.4522,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5205,
      "query_norm": 1.7742,
      "queue_k_norm": 1.71,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9553,
      "sent_len_1": 66.7705,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.275,
      "stdk": 0.0498,
      "stdq": 0.046,
      "stdqueue_k": 0.0501,
      "stdqueue_q": 0.0,
      "step": 6400
    },
    {
      "accuracy": 46.1914,
      "active_queue_size": 16384.0,
      "cl_loss": 4.2312,
      "doc_norm": 1.7001,
      "encoder_q-embeddings": 5830.876,
      "encoder_q-layer.0": 5105.1055,
      "encoder_q-layer.1": 5418.9136,
      "encoder_q-layer.10": 3831.0039,
      "encoder_q-layer.11": 6853.3076,
      "encoder_q-layer.2": 6586.0732,
      "encoder_q-layer.3": 6353.438,
      "encoder_q-layer.4": 5763.8188,
      "encoder_q-layer.5": 5450.5249,
      "encoder_q-layer.6": 4432.5786,
      "encoder_q-layer.7": 3076.1313,
      "encoder_q-layer.8": 3026.9827,
      "encoder_q-layer.9": 2791.1824,
      "epoch": 0.06,
      "inbatch_neg_score": 0.5256,
      "inbatch_pos_score": 1.1133,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 4.2312,
      "norm_diff": 0.1314,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7816.5019,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5186,
      "query_norm": 1.8314,
      "queue_k_norm": 1.7004,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0084,
      "sent_len_1": 66.6807,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6012,
      "stdk": 0.0499,
      "stdq": 0.0464,
      "stdqueue_k": 0.0499,
      "stdqueue_q": 0.0,
      "step": 6500
    },
    {
      "accuracy": 45.2148,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1838,
      "doc_norm": 1.6938,
      "encoder_q-embeddings": 10079.9873,
      "encoder_q-layer.0": 8442.5107,
      "encoder_q-layer.1": 9077.9365,
      "encoder_q-layer.10": 5757.5322,
      "encoder_q-layer.11": 9796.1016,
      "encoder_q-layer.2": 11250.4102,
      "encoder_q-layer.3": 11547.7246,
      "encoder_q-layer.4": 10636.3555,
      "encoder_q-layer.5": 10227.4463,
      "encoder_q-layer.6": 11099.1182,
      "encoder_q-layer.7": 10200.291,
      "encoder_q-layer.8": 7597.9126,
      "encoder_q-layer.9": 4664.3647,
      "epoch": 0.06,
      "inbatch_neg_score": 0.5374,
      "inbatch_pos_score": 1.1152,
      "learning_rate": 3.3e-05,
      "loss": 4.1838,
      "norm_diff": 0.1393,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 14138.5556,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5322,
      "query_norm": 1.8331,
      "queue_k_norm": 1.6924,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7995,
      "sent_len_1": 66.8571,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.4075,
      "stdk": 0.0499,
      "stdq": 0.0457,
      "stdqueue_k": 0.0499,
      "stdqueue_q": 0.0,
      "step": 6600
    },
    {
      "accuracy": 45.1172,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1983,
      "doc_norm": 1.6842,
      "encoder_q-embeddings": 9139.3242,
      "encoder_q-layer.0": 7671.4126,
      "encoder_q-layer.1": 7504.9771,
      "encoder_q-layer.10": 5081.4395,
      "encoder_q-layer.11": 7049.6768,
      "encoder_q-layer.2": 8521.3105,
      "encoder_q-layer.3": 9108.4268,
      "encoder_q-layer.4": 8384.2627,
      "encoder_q-layer.5": 8248.627,
      "encoder_q-layer.6": 6519.0693,
      "encoder_q-layer.7": 4759.1479,
      "encoder_q-layer.8": 4488.7783,
      "encoder_q-layer.9": 3878.7708,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5764,
      "inbatch_pos_score": 1.1436,
      "learning_rate": 3.35e-05,
      "loss": 4.1983,
      "norm_diff": 0.2071,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 10786.8664,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5684,
      "query_norm": 1.8913,
      "queue_k_norm": 1.6831,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8284,
      "sent_len_1": 66.7389,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.235,
      "stdk": 0.0497,
      "stdq": 0.0464,
      "stdqueue_k": 0.0497,
      "stdqueue_q": 0.0,
      "step": 6700
    },
    {
      "accuracy": 42.7734,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1806,
      "doc_norm": 1.6735,
      "encoder_q-embeddings": 4335.5176,
      "encoder_q-layer.0": 3438.823,
      "encoder_q-layer.1": 3650.8928,
      "encoder_q-layer.10": 3347.7729,
      "encoder_q-layer.11": 6336.6694,
      "encoder_q-layer.2": 4077.6787,
      "encoder_q-layer.3": 4217.5503,
      "encoder_q-layer.4": 4024.168,
      "encoder_q-layer.5": 4269.8677,
      "encoder_q-layer.6": 4670.251,
      "encoder_q-layer.7": 4280.6562,
      "encoder_q-layer.8": 5005.7451,
      "encoder_q-layer.9": 3321.8132,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5793,
      "inbatch_pos_score": 1.1338,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 4.1806,
      "norm_diff": 0.2444,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6254.1959,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5718,
      "query_norm": 1.9179,
      "queue_k_norm": 1.6786,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8582,
      "sent_len_1": 66.6977,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.9338,
      "stdk": 0.0493,
      "stdq": 0.0464,
      "stdqueue_k": 0.0497,
      "stdqueue_q": 0.0,
      "step": 6800
    },
    {
      "accuracy": 43.9453,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1264,
      "doc_norm": 1.6718,
      "encoder_q-embeddings": 6578.7891,
      "encoder_q-layer.0": 5056.333,
      "encoder_q-layer.1": 5000.7324,
      "encoder_q-layer.10": 3529.1582,
      "encoder_q-layer.11": 6308.2515,
      "encoder_q-layer.2": 5560.5181,
      "encoder_q-layer.3": 5883.2344,
      "encoder_q-layer.4": 5330.3965,
      "encoder_q-layer.5": 5508.5107,
      "encoder_q-layer.6": 5454.5415,
      "encoder_q-layer.7": 3664.938,
      "encoder_q-layer.8": 3023.4946,
      "encoder_q-layer.9": 2596.5898,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5938,
      "inbatch_pos_score": 1.1553,
      "learning_rate": 3.45e-05,
      "loss": 4.1264,
      "norm_diff": 0.2423,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7563.4759,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5845,
      "query_norm": 1.9141,
      "queue_k_norm": 1.6703,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0216,
      "sent_len_1": 66.8705,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9812,
      "stdk": 0.0494,
      "stdq": 0.0461,
      "stdqueue_k": 0.0494,
      "stdqueue_q": 0.0,
      "step": 6900
    },
    {
      "accuracy": 45.4102,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1054,
      "doc_norm": 1.6618,
      "encoder_q-embeddings": 4363.8872,
      "encoder_q-layer.0": 3404.1799,
      "encoder_q-layer.1": 3937.2637,
      "encoder_q-layer.10": 3475.4028,
      "encoder_q-layer.11": 6841.4736,
      "encoder_q-layer.2": 4338.2095,
      "encoder_q-layer.3": 4292.6089,
      "encoder_q-layer.4": 4549.0522,
      "encoder_q-layer.5": 4796.8149,
      "encoder_q-layer.6": 4553.6426,
      "encoder_q-layer.7": 3834.6418,
      "encoder_q-layer.8": 3516.1377,
      "encoder_q-layer.9": 2525.4731,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5613,
      "inbatch_pos_score": 1.126,
      "learning_rate": 3.5e-05,
      "loss": 4.1054,
      "norm_diff": 0.257,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6392.8717,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5542,
      "query_norm": 1.9188,
      "queue_k_norm": 1.6669,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1198,
      "sent_len_1": 66.9368,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.2188,
      "stdk": 0.0491,
      "stdq": 0.0465,
      "stdqueue_k": 0.0495,
      "stdqueue_q": 0.0,
      "step": 7000
    },
    {
      "accuracy": 48.7305,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1002,
      "doc_norm": 1.6576,
      "encoder_q-embeddings": 14933.0967,
      "encoder_q-layer.0": 11326.6113,
      "encoder_q-layer.1": 12658.6562,
      "encoder_q-layer.10": 4904.3467,
      "encoder_q-layer.11": 8805.7061,
      "encoder_q-layer.2": 15708.8232,
      "encoder_q-layer.3": 14747.7754,
      "encoder_q-layer.4": 13139.3584,
      "encoder_q-layer.5": 13481.3076,
      "encoder_q-layer.6": 12920.2861,
      "encoder_q-layer.7": 10571.1514,
      "encoder_q-layer.8": 5541.7769,
      "encoder_q-layer.9": 3708.8401,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5824,
      "inbatch_pos_score": 1.1895,
      "learning_rate": 3.55e-05,
      "loss": 4.1002,
      "norm_diff": 0.2674,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 17945.3169,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.5742,
      "query_norm": 1.925,
      "queue_k_norm": 1.6608,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7945,
      "sent_len_1": 66.6045,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5563,
      "stdk": 0.0491,
      "stdq": 0.0464,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 7100
    },
    {
      "accuracy": 45.7031,
      "active_queue_size": 16384.0,
      "cl_loss": 4.0972,
      "doc_norm": 1.6479,
      "encoder_q-embeddings": 5179.75,
      "encoder_q-layer.0": 4301.9121,
      "encoder_q-layer.1": 4461.6079,
      "encoder_q-layer.10": 2882.1272,
      "encoder_q-layer.11": 5831.6318,
      "encoder_q-layer.2": 5106.7314,
      "encoder_q-layer.3": 4771.0249,
      "encoder_q-layer.4": 4626.7495,
      "encoder_q-layer.5": 4646.9004,
      "encoder_q-layer.6": 3827.5779,
      "encoder_q-layer.7": 3061.1714,
      "encoder_q-layer.8": 2854.2827,
      "encoder_q-layer.9": 2403.7659,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5384,
      "inbatch_pos_score": 1.1035,
      "learning_rate": 3.6e-05,
      "loss": 4.0972,
      "norm_diff": 0.2782,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6427.3504,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5293,
      "query_norm": 1.9261,
      "queue_k_norm": 1.6573,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9208,
      "sent_len_1": 66.6823,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4613,
      "stdk": 0.0489,
      "stdq": 0.0457,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 7200
    },
    {
      "accuracy": 43.75,
      "active_queue_size": 16384.0,
      "cl_loss": 4.1085,
      "doc_norm": 1.6481,
      "encoder_q-embeddings": 4465.02,
      "encoder_q-layer.0": 3330.5779,
      "encoder_q-layer.1": 3203.8201,
      "encoder_q-layer.10": 3303.8386,
      "encoder_q-layer.11": 6765.4092,
      "encoder_q-layer.2": 3515.5962,
      "encoder_q-layer.3": 3619.8489,
      "encoder_q-layer.4": 3686.459,
      "encoder_q-layer.5": 3730.6997,
      "encoder_q-layer.6": 3719.6553,
      "encoder_q-layer.7": 3849.9509,
      "encoder_q-layer.8": 3681.0601,
      "encoder_q-layer.9": 2807.0762,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5504,
      "inbatch_pos_score": 1.1143,
      "learning_rate": 3.65e-05,
      "loss": 4.1085,
      "norm_diff": 0.2983,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5850.5879,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.542,
      "query_norm": 1.9464,
      "queue_k_norm": 1.6488,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9557,
      "sent_len_1": 66.7591,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.73,
      "stdk": 0.049,
      "stdq": 0.0467,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 7300
    },
    {
      "accuracy": 41.6016,
      "active_queue_size": 16384.0,
      "cl_loss": 4.039,
      "doc_norm": 1.6434,
      "encoder_q-embeddings": 3841.0474,
      "encoder_q-layer.0": 3134.6201,
      "encoder_q-layer.1": 3510.9939,
      "encoder_q-layer.10": 3997.1157,
      "encoder_q-layer.11": 6638.77,
      "encoder_q-layer.2": 4208.4033,
      "encoder_q-layer.3": 4737.3984,
      "encoder_q-layer.4": 4810.1948,
      "encoder_q-layer.5": 5033.7505,
      "encoder_q-layer.6": 4704.6885,
      "encoder_q-layer.7": 4102.0098,
      "encoder_q-layer.8": 3994.8623,
      "encoder_q-layer.9": 3879.7871,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5464,
      "inbatch_pos_score": 1.0801,
      "learning_rate": 3.7e-05,
      "loss": 4.039,
      "norm_diff": 0.2831,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6382.1788,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5376,
      "query_norm": 1.9265,
      "queue_k_norm": 1.6468,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9347,
      "sent_len_1": 66.7611,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6025,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 7400
    },
    {
      "accuracy": 45.1172,
      "active_queue_size": 16384.0,
      "cl_loss": 4.0199,
      "doc_norm": 1.6427,
      "encoder_q-embeddings": 3333.1528,
      "encoder_q-layer.0": 2427.4639,
      "encoder_q-layer.1": 2536.9778,
      "encoder_q-layer.10": 3342.0698,
      "encoder_q-layer.11": 6263.3555,
      "encoder_q-layer.2": 2927.8064,
      "encoder_q-layer.3": 2980.1267,
      "encoder_q-layer.4": 3164.6946,
      "encoder_q-layer.5": 3346.6018,
      "encoder_q-layer.6": 3398.051,
      "encoder_q-layer.7": 3368.5378,
      "encoder_q-layer.8": 3954.7695,
      "encoder_q-layer.9": 3110.9331,
      "epoch": 0.07,
      "inbatch_neg_score": 0.5246,
      "inbatch_pos_score": 1.0938,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 4.0199,
      "norm_diff": 0.2707,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5140.2205,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5156,
      "query_norm": 1.9134,
      "queue_k_norm": 1.6398,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0926,
      "sent_len_1": 66.9453,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4125,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 7500
    },
    {
      "accuracy": 45.1172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9981,
      "doc_norm": 1.6286,
      "encoder_q-embeddings": 3143.6313,
      "encoder_q-layer.0": 2298.6775,
      "encoder_q-layer.1": 2447.5061,
      "encoder_q-layer.10": 2765.3386,
      "encoder_q-layer.11": 5407.6099,
      "encoder_q-layer.2": 2796.2537,
      "encoder_q-layer.3": 2688.8979,
      "encoder_q-layer.4": 2810.3853,
      "encoder_q-layer.5": 2828.9753,
      "encoder_q-layer.6": 2626.9978,
      "encoder_q-layer.7": 2299.4333,
      "encoder_q-layer.8": 2409.6406,
      "encoder_q-layer.9": 2195.2214,
      "epoch": 0.07,
      "inbatch_neg_score": 0.4893,
      "inbatch_pos_score": 1.0586,
      "learning_rate": 3.8e-05,
      "loss": 3.9981,
      "norm_diff": 0.3189,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4365.0486,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4805,
      "query_norm": 1.9475,
      "queue_k_norm": 1.636,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9815,
      "sent_len_1": 66.8387,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7025,
      "stdk": 0.0485,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 7600
    },
    {
      "accuracy": 44.6289,
      "active_queue_size": 16384.0,
      "cl_loss": 4.0429,
      "doc_norm": 1.6322,
      "encoder_q-embeddings": 2745.519,
      "encoder_q-layer.0": 2144.1682,
      "encoder_q-layer.1": 2261.6624,
      "encoder_q-layer.10": 3021.8599,
      "encoder_q-layer.11": 6000.1904,
      "encoder_q-layer.2": 2489.7769,
      "encoder_q-layer.3": 2548.1135,
      "encoder_q-layer.4": 2670.9585,
      "encoder_q-layer.5": 2780.7493,
      "encoder_q-layer.6": 2696.4009,
      "encoder_q-layer.7": 2450.5454,
      "encoder_q-layer.8": 2494.5493,
      "encoder_q-layer.9": 2164.4553,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4894,
      "inbatch_pos_score": 1.0605,
      "learning_rate": 3.85e-05,
      "loss": 4.0429,
      "norm_diff": 0.264,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4371.821,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4832,
      "query_norm": 1.8962,
      "queue_k_norm": 1.6309,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8225,
      "sent_len_1": 66.5445,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0137,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 7700
    },
    {
      "accuracy": 49.4141,
      "active_queue_size": 16384.0,
      "cl_loss": 4.0201,
      "doc_norm": 1.6233,
      "encoder_q-embeddings": 2307.7273,
      "encoder_q-layer.0": 1677.8064,
      "encoder_q-layer.1": 1733.4159,
      "encoder_q-layer.10": 2546.0913,
      "encoder_q-layer.11": 5302.2407,
      "encoder_q-layer.2": 1924.5396,
      "encoder_q-layer.3": 1992.9025,
      "encoder_q-layer.4": 2115.551,
      "encoder_q-layer.5": 2027.1743,
      "encoder_q-layer.6": 2018.0228,
      "encoder_q-layer.7": 1999.8083,
      "encoder_q-layer.8": 2326.7268,
      "encoder_q-layer.9": 2027.1915,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4866,
      "inbatch_pos_score": 1.0938,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 4.0201,
      "norm_diff": 0.2667,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3653.6909,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4788,
      "query_norm": 1.8899,
      "queue_k_norm": 1.6241,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1475,
      "sent_len_1": 66.7962,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4062,
      "stdk": 0.0487,
      "stdq": 0.046,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 7800
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 4.0355,
      "doc_norm": 1.6185,
      "encoder_q-embeddings": 3930.7703,
      "encoder_q-layer.0": 3070.0327,
      "encoder_q-layer.1": 3082.45,
      "encoder_q-layer.10": 2552.324,
      "encoder_q-layer.11": 5134.9961,
      "encoder_q-layer.2": 3621.0925,
      "encoder_q-layer.3": 3685.8403,
      "encoder_q-layer.4": 3536.2622,
      "encoder_q-layer.5": 3571.3672,
      "encoder_q-layer.6": 3062.0022,
      "encoder_q-layer.7": 2531.6292,
      "encoder_q-layer.8": 2752.7292,
      "encoder_q-layer.9": 2239.9033,
      "epoch": 0.08,
      "inbatch_neg_score": 0.486,
      "inbatch_pos_score": 1.1016,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 4.0355,
      "norm_diff": 0.2838,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4995.2658,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4758,
      "query_norm": 1.9023,
      "queue_k_norm": 1.62,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7465,
      "sent_len_1": 66.6071,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.25,
      "stdk": 0.0487,
      "stdq": 0.0465,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 7900
    },
    {
      "accuracy": 45.0195,
      "active_queue_size": 16384.0,
      "cl_loss": 4.0528,
      "doc_norm": 1.6121,
      "encoder_q-embeddings": 3166.0195,
      "encoder_q-layer.0": 2272.5168,
      "encoder_q-layer.1": 2443.8884,
      "encoder_q-layer.10": 3152.6899,
      "encoder_q-layer.11": 6116.4502,
      "encoder_q-layer.2": 2812.2642,
      "encoder_q-layer.3": 2811.3955,
      "encoder_q-layer.4": 2997.0549,
      "encoder_q-layer.5": 3205.772,
      "encoder_q-layer.6": 3237.2151,
      "encoder_q-layer.7": 4032.5964,
      "encoder_q-layer.8": 3576.127,
      "encoder_q-layer.9": 2727.1023,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4831,
      "inbatch_pos_score": 1.0342,
      "learning_rate": 4e-05,
      "loss": 4.0528,
      "norm_diff": 0.2822,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4974.0772,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4751,
      "query_norm": 1.8944,
      "queue_k_norm": 1.6185,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.919,
      "sent_len_1": 66.4893,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9863,
      "stdk": 0.0485,
      "stdq": 0.0457,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 8000
    },
    {
      "accuracy": 45.8984,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9835,
      "doc_norm": 1.6101,
      "encoder_q-embeddings": 5378.2241,
      "encoder_q-layer.0": 4232.8379,
      "encoder_q-layer.1": 4464.2007,
      "encoder_q-layer.10": 2565.0317,
      "encoder_q-layer.11": 5325.9043,
      "encoder_q-layer.2": 4730.8809,
      "encoder_q-layer.3": 4513.9092,
      "encoder_q-layer.4": 4461.3076,
      "encoder_q-layer.5": 4624.9517,
      "encoder_q-layer.6": 4136.0415,
      "encoder_q-layer.7": 3244.355,
      "encoder_q-layer.8": 2798.5903,
      "encoder_q-layer.9": 2190.6545,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4856,
      "inbatch_pos_score": 1.0664,
      "learning_rate": 4.05e-05,
      "loss": 3.9835,
      "norm_diff": 0.3064,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6280.5552,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4763,
      "query_norm": 1.9166,
      "queue_k_norm": 1.612,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0903,
      "sent_len_1": 66.7248,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7438,
      "stdk": 0.0487,
      "stdq": 0.0464,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 8100
    },
    {
      "accuracy": 47.4609,
      "active_queue_size": 16384.0,
      "cl_loss": 3.964,
      "doc_norm": 1.6004,
      "encoder_q-embeddings": 5836.0229,
      "encoder_q-layer.0": 4536.7139,
      "encoder_q-layer.1": 4646.1846,
      "encoder_q-layer.10": 5046.1309,
      "encoder_q-layer.11": 11108.626,
      "encoder_q-layer.2": 5165.8174,
      "encoder_q-layer.3": 5071.2529,
      "encoder_q-layer.4": 5279.4668,
      "encoder_q-layer.5": 5407.5322,
      "encoder_q-layer.6": 4977.0146,
      "encoder_q-layer.7": 4587.9019,
      "encoder_q-layer.8": 5112.7275,
      "encoder_q-layer.9": 4052.0347,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4852,
      "inbatch_pos_score": 1.082,
      "learning_rate": 4.1e-05,
      "loss": 3.964,
      "norm_diff": 0.2734,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8536.7546,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4778,
      "query_norm": 1.8739,
      "queue_k_norm": 1.6076,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9377,
      "sent_len_1": 66.7896,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2675,
      "stdk": 0.0485,
      "stdq": 0.0458,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 8200
    },
    {
      "accuracy": 44.8242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9701,
      "doc_norm": 1.604,
      "encoder_q-embeddings": 11814.498,
      "encoder_q-layer.0": 9397.1992,
      "encoder_q-layer.1": 10738.8662,
      "encoder_q-layer.10": 5488.5713,
      "encoder_q-layer.11": 12476.7881,
      "encoder_q-layer.2": 12527.3916,
      "encoder_q-layer.3": 13987.6816,
      "encoder_q-layer.4": 13304.8887,
      "encoder_q-layer.5": 11308.9326,
      "encoder_q-layer.6": 10630.2988,
      "encoder_q-layer.7": 8877.582,
      "encoder_q-layer.8": 7255.1377,
      "encoder_q-layer.9": 4883.2471,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4968,
      "inbatch_pos_score": 1.0635,
      "learning_rate": 4.15e-05,
      "loss": 3.9701,
      "norm_diff": 0.2241,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 16218.581,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4922,
      "query_norm": 1.8281,
      "queue_k_norm": 1.6045,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0322,
      "sent_len_1": 66.6502,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.32,
      "stdk": 0.0486,
      "stdq": 0.0453,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 8300
    },
    {
      "accuracy": 46.6797,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9767,
      "doc_norm": 1.5939,
      "encoder_q-embeddings": 8513.5,
      "encoder_q-layer.0": 6485.2949,
      "encoder_q-layer.1": 6957.0928,
      "encoder_q-layer.10": 5007.5591,
      "encoder_q-layer.11": 11138.7158,
      "encoder_q-layer.2": 7758.8765,
      "encoder_q-layer.3": 7822.5059,
      "encoder_q-layer.4": 8511.5332,
      "encoder_q-layer.5": 8750.8359,
      "encoder_q-layer.6": 8830.6621,
      "encoder_q-layer.7": 7251.2158,
      "encoder_q-layer.8": 5624.1079,
      "encoder_q-layer.9": 4044.615,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4703,
      "inbatch_pos_score": 1.0439,
      "learning_rate": 4.2e-05,
      "loss": 3.9767,
      "norm_diff": 0.1648,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 11416.8741,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4653,
      "query_norm": 1.7587,
      "queue_k_norm": 1.6011,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.709,
      "sent_len_1": 66.8283,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.94,
      "stdk": 0.0484,
      "stdq": 0.0451,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 8400
    },
    {
      "accuracy": 42.9688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9471,
      "doc_norm": 1.5908,
      "encoder_q-embeddings": 14496.5781,
      "encoder_q-layer.0": 10713.9824,
      "encoder_q-layer.1": 11070.0381,
      "encoder_q-layer.10": 5808.6099,
      "encoder_q-layer.11": 11595.4756,
      "encoder_q-layer.2": 12815.8408,
      "encoder_q-layer.3": 13781.3408,
      "encoder_q-layer.4": 14669.5352,
      "encoder_q-layer.5": 14948.6846,
      "encoder_q-layer.6": 15221.8594,
      "encoder_q-layer.7": 13152.2998,
      "encoder_q-layer.8": 10046.7725,
      "encoder_q-layer.9": 6164.0352,
      "epoch": 0.08,
      "inbatch_neg_score": 0.451,
      "inbatch_pos_score": 1.0137,
      "learning_rate": 4.25e-05,
      "loss": 3.9471,
      "norm_diff": 0.1809,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 18192.4707,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.4473,
      "query_norm": 1.7717,
      "queue_k_norm": 1.5956,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0091,
      "sent_len_1": 66.7102,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9175,
      "stdk": 0.0484,
      "stdq": 0.047,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 8500
    },
    {
      "accuracy": 47.7539,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9375,
      "doc_norm": 1.5895,
      "encoder_q-embeddings": 6345.3765,
      "encoder_q-layer.0": 5102.5527,
      "encoder_q-layer.1": 5084.8115,
      "encoder_q-layer.10": 2255.1584,
      "encoder_q-layer.11": 4605.4111,
      "encoder_q-layer.2": 4386.0376,
      "encoder_q-layer.3": 4025.8145,
      "encoder_q-layer.4": 3812.7598,
      "encoder_q-layer.5": 4288.7979,
      "encoder_q-layer.6": 3868.6743,
      "encoder_q-layer.7": 2960.4236,
      "encoder_q-layer.8": 2346.3757,
      "encoder_q-layer.9": 1885.9329,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4395,
      "inbatch_pos_score": 1.0469,
      "learning_rate": 4.3e-05,
      "loss": 3.9375,
      "norm_diff": 0.1742,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6344.1302,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4355,
      "query_norm": 1.7637,
      "queue_k_norm": 1.5898,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8119,
      "sent_len_1": 66.7382,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0662,
      "stdk": 0.0486,
      "stdq": 0.0458,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 8600
    },
    {
      "accuracy": 47.7539,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9281,
      "doc_norm": 1.5915,
      "encoder_q-embeddings": 2426.4014,
      "encoder_q-layer.0": 1754.2914,
      "encoder_q-layer.1": 1874.1854,
      "encoder_q-layer.10": 2505.6128,
      "encoder_q-layer.11": 5364.6553,
      "encoder_q-layer.2": 2002.113,
      "encoder_q-layer.3": 2076.9189,
      "encoder_q-layer.4": 2284.7288,
      "encoder_q-layer.5": 2254.9426,
      "encoder_q-layer.6": 2245.4102,
      "encoder_q-layer.7": 2052.8594,
      "encoder_q-layer.8": 2324.8662,
      "encoder_q-layer.9": 2129.377,
      "epoch": 0.08,
      "inbatch_neg_score": 0.4545,
      "inbatch_pos_score": 1.0273,
      "learning_rate": 4.35e-05,
      "loss": 3.9281,
      "norm_diff": 0.1795,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3821.4707,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4492,
      "query_norm": 1.771,
      "queue_k_norm": 1.585,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.923,
      "sent_len_1": 66.8144,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7388,
      "stdk": 0.0488,
      "stdq": 0.0449,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 8700
    },
    {
      "accuracy": 46.7773,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8882,
      "doc_norm": 1.5922,
      "encoder_q-embeddings": 2237.5703,
      "encoder_q-layer.0": 1691.1095,
      "encoder_q-layer.1": 1758.0088,
      "encoder_q-layer.10": 2526.5715,
      "encoder_q-layer.11": 5309.3408,
      "encoder_q-layer.2": 1917.2933,
      "encoder_q-layer.3": 2107.8115,
      "encoder_q-layer.4": 2272.5217,
      "encoder_q-layer.5": 2338.7725,
      "encoder_q-layer.6": 2417.4688,
      "encoder_q-layer.7": 2855.7729,
      "encoder_q-layer.8": 2767.7241,
      "encoder_q-layer.9": 2192.061,
      "epoch": 0.09,
      "inbatch_neg_score": 0.4581,
      "inbatch_pos_score": 1.041,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 3.8882,
      "norm_diff": 0.2385,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3983.0947,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4524,
      "query_norm": 1.8307,
      "queue_k_norm": 1.5813,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9688,
      "sent_len_1": 66.841,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4737,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 8800
    },
    {
      "accuracy": 49.6094,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9092,
      "doc_norm": 1.5752,
      "encoder_q-embeddings": 4108.9956,
      "encoder_q-layer.0": 3037.6924,
      "encoder_q-layer.1": 3400.7451,
      "encoder_q-layer.10": 2490.0752,
      "encoder_q-layer.11": 5089.5552,
      "encoder_q-layer.2": 3858.3384,
      "encoder_q-layer.3": 4104.0342,
      "encoder_q-layer.4": 4065.3992,
      "encoder_q-layer.5": 3964.0813,
      "encoder_q-layer.6": 4026.21,
      "encoder_q-layer.7": 3029.4194,
      "encoder_q-layer.8": 2937.0071,
      "encoder_q-layer.9": 2062.4868,
      "epoch": 0.09,
      "inbatch_neg_score": 0.472,
      "inbatch_pos_score": 1.0684,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 3.9092,
      "norm_diff": 0.2397,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5520.3714,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4661,
      "query_norm": 1.815,
      "queue_k_norm": 1.5775,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9027,
      "sent_len_1": 66.7621,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2038,
      "stdk": 0.0483,
      "stdq": 0.0459,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 8900
    },
    {
      "accuracy": 45.8984,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9331,
      "doc_norm": 1.5733,
      "encoder_q-embeddings": 28428.25,
      "encoder_q-layer.0": 21571.3809,
      "encoder_q-layer.1": 26044.1523,
      "encoder_q-layer.10": 2285.6973,
      "encoder_q-layer.11": 5008.3965,
      "encoder_q-layer.2": 30118.0449,
      "encoder_q-layer.3": 33871.4141,
      "encoder_q-layer.4": 36140.0742,
      "encoder_q-layer.5": 32741.4531,
      "encoder_q-layer.6": 34471.875,
      "encoder_q-layer.7": 21305.8809,
      "encoder_q-layer.8": 10587.3799,
      "encoder_q-layer.9": 2321.0054,
      "epoch": 0.09,
      "inbatch_neg_score": 0.474,
      "inbatch_pos_score": 1.0508,
      "learning_rate": 4.5e-05,
      "loss": 3.9331,
      "norm_diff": 0.1872,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 37530.8416,
      "preclip_grad_norm_avg": 0.0003,
      "q@queue_neg_score": 0.4668,
      "query_norm": 1.7605,
      "queue_k_norm": 1.5707,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1418,
      "sent_len_1": 66.7443,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.975,
      "stdk": 0.0483,
      "stdq": 0.046,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 9000
    },
    {
      "accuracy": 46.582,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9052,
      "doc_norm": 1.5634,
      "encoder_q-embeddings": 2329.1304,
      "encoder_q-layer.0": 1826.3383,
      "encoder_q-layer.1": 1970.1744,
      "encoder_q-layer.10": 2476.2373,
      "encoder_q-layer.11": 5106.1118,
      "encoder_q-layer.2": 2055.3235,
      "encoder_q-layer.3": 2157.665,
      "encoder_q-layer.4": 2132.2021,
      "encoder_q-layer.5": 2112.8799,
      "encoder_q-layer.6": 2342.783,
      "encoder_q-layer.7": 2179.1875,
      "encoder_q-layer.8": 2350.6882,
      "encoder_q-layer.9": 1995.0621,
      "epoch": 0.09,
      "inbatch_neg_score": 0.4699,
      "inbatch_pos_score": 1.0566,
      "learning_rate": 4.55e-05,
      "loss": 3.9052,
      "norm_diff": 0.1518,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3631.6899,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4653,
      "query_norm": 1.7152,
      "queue_k_norm": 1.5677,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0246,
      "sent_len_1": 66.6872,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7113,
      "stdk": 0.0482,
      "stdq": 0.0463,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 9100
    },
    {
      "accuracy": 48.7305,
      "active_queue_size": 16384.0,
      "cl_loss": 3.9374,
      "doc_norm": 1.5644,
      "encoder_q-embeddings": 2091.7727,
      "encoder_q-layer.0": 1546.6447,
      "encoder_q-layer.1": 1638.9701,
      "encoder_q-layer.10": 2110.6799,
      "encoder_q-layer.11": 4823.5708,
      "encoder_q-layer.2": 1812.7776,
      "encoder_q-layer.3": 1949.2878,
      "encoder_q-layer.4": 1951.0656,
      "encoder_q-layer.5": 1983.5402,
      "encoder_q-layer.6": 2177.2021,
      "encoder_q-layer.7": 2120.3562,
      "encoder_q-layer.8": 2118.207,
      "encoder_q-layer.9": 1829.3379,
      "epoch": 0.09,
      "inbatch_neg_score": 0.4193,
      "inbatch_pos_score": 0.9941,
      "learning_rate": 4.600000000000001e-05,
      "loss": 3.9374,
      "norm_diff": 0.0883,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3443.4014,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4138,
      "query_norm": 1.6527,
      "queue_k_norm": 1.5632,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8553,
      "sent_len_1": 66.7163,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4688,
      "stdk": 0.0483,
      "stdq": 0.0444,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 9200
    },
    {
      "accuracy": 44.2383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8898,
      "doc_norm": 1.5653,
      "encoder_q-embeddings": 2218.9312,
      "encoder_q-layer.0": 1766.0363,
      "encoder_q-layer.1": 1895.9043,
      "encoder_q-layer.10": 2422.158,
      "encoder_q-layer.11": 5220.5703,
      "encoder_q-layer.2": 2068.668,
      "encoder_q-layer.3": 2100.2437,
      "encoder_q-layer.4": 2077.6531,
      "encoder_q-layer.5": 2152.5139,
      "encoder_q-layer.6": 2244.936,
      "encoder_q-layer.7": 2220.1985,
      "encoder_q-layer.8": 2362.0061,
      "encoder_q-layer.9": 1879.469,
      "epoch": 0.09,
      "inbatch_neg_score": 0.3858,
      "inbatch_pos_score": 0.9375,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 3.8898,
      "norm_diff": 0.1514,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3710.735,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3818,
      "query_norm": 1.7167,
      "queue_k_norm": 1.5601,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9834,
      "sent_len_1": 66.9512,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0712,
      "stdk": 0.0484,
      "stdq": 0.0454,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 9300
    },
    {
      "accuracy": 47.6562,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8729,
      "doc_norm": 1.5559,
      "encoder_q-embeddings": 3369.2881,
      "encoder_q-layer.0": 2620.5684,
      "encoder_q-layer.1": 2815.5239,
      "encoder_q-layer.10": 1913.45,
      "encoder_q-layer.11": 4345.3291,
      "encoder_q-layer.2": 3147.3723,
      "encoder_q-layer.3": 3240.8855,
      "encoder_q-layer.4": 3305.894,
      "encoder_q-layer.5": 3248.479,
      "encoder_q-layer.6": 2992.4417,
      "encoder_q-layer.7": 2381.6123,
      "encoder_q-layer.8": 2380.469,
      "encoder_q-layer.9": 1801.8413,
      "epoch": 0.09,
      "inbatch_neg_score": 0.3835,
      "inbatch_pos_score": 0.9595,
      "learning_rate": 4.7e-05,
      "loss": 3.8729,
      "norm_diff": 0.2672,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4405.2575,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3755,
      "query_norm": 1.8232,
      "queue_k_norm": 1.555,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1445,
      "sent_len_1": 66.5738,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9425,
      "stdk": 0.0484,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 9400
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8584,
      "doc_norm": 1.5533,
      "encoder_q-embeddings": 3717.4285,
      "encoder_q-layer.0": 2822.4722,
      "encoder_q-layer.1": 3090.3821,
      "encoder_q-layer.10": 1998.957,
      "encoder_q-layer.11": 3932.3835,
      "encoder_q-layer.2": 3385.2422,
      "encoder_q-layer.3": 3316.064,
      "encoder_q-layer.4": 3315.4941,
      "encoder_q-layer.5": 3212.0479,
      "encoder_q-layer.6": 3813.3418,
      "encoder_q-layer.7": 3512.3132,
      "encoder_q-layer.8": 3182.5552,
      "encoder_q-layer.9": 1708.0924,
      "epoch": 0.09,
      "inbatch_neg_score": 0.4157,
      "inbatch_pos_score": 1.0381,
      "learning_rate": 4.75e-05,
      "loss": 3.8584,
      "norm_diff": 0.3134,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4863.1222,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4072,
      "query_norm": 1.8666,
      "queue_k_norm": 1.5512,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8107,
      "sent_len_1": 66.7812,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.985,
      "stdk": 0.0485,
      "stdq": 0.0461,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 9500
    },
    {
      "accuracy": 46.9727,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8739,
      "doc_norm": 1.5477,
      "encoder_q-embeddings": 1844.7899,
      "encoder_q-layer.0": 1298.0183,
      "encoder_q-layer.1": 1377.6073,
      "encoder_q-layer.10": 2034.6401,
      "encoder_q-layer.11": 4578.0435,
      "encoder_q-layer.2": 1566.8417,
      "encoder_q-layer.3": 1690.0135,
      "encoder_q-layer.4": 1760.3304,
      "encoder_q-layer.5": 1844.8112,
      "encoder_q-layer.6": 1889.5553,
      "encoder_q-layer.7": 1775.187,
      "encoder_q-layer.8": 2020.3032,
      "encoder_q-layer.9": 1794.3826,
      "epoch": 0.09,
      "inbatch_neg_score": 0.4416,
      "inbatch_pos_score": 1.0244,
      "learning_rate": 4.8e-05,
      "loss": 3.8739,
      "norm_diff": 0.2028,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3202.4884,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4368,
      "query_norm": 1.7504,
      "queue_k_norm": 1.549,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8528,
      "sent_len_1": 66.5521,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.3638,
      "stdk": 0.0484,
      "stdq": 0.0441,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 9600
    },
    {
      "accuracy": 48.8281,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8596,
      "doc_norm": 1.5431,
      "encoder_q-embeddings": 3558.7866,
      "encoder_q-layer.0": 2880.3789,
      "encoder_q-layer.1": 2854.7271,
      "encoder_q-layer.10": 1949.1479,
      "encoder_q-layer.11": 4393.7417,
      "encoder_q-layer.2": 3158.1055,
      "encoder_q-layer.3": 3273.113,
      "encoder_q-layer.4": 3378.5376,
      "encoder_q-layer.5": 3444.7158,
      "encoder_q-layer.6": 3171.4875,
      "encoder_q-layer.7": 2401.4519,
      "encoder_q-layer.8": 2005.3352,
      "encoder_q-layer.9": 1695.6982,
      "epoch": 0.09,
      "inbatch_neg_score": 0.4255,
      "inbatch_pos_score": 1.041,
      "learning_rate": 4.85e-05,
      "loss": 3.8596,
      "norm_diff": 0.1356,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4564.656,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4211,
      "query_norm": 1.6787,
      "queue_k_norm": 1.5457,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0084,
      "sent_len_1": 66.7332,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7713,
      "stdk": 0.0483,
      "stdq": 0.0454,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 9700
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.88,
      "doc_norm": 1.5472,
      "encoder_q-embeddings": 2744.8625,
      "encoder_q-layer.0": 1941.8469,
      "encoder_q-layer.1": 2199.5332,
      "encoder_q-layer.10": 2007.4723,
      "encoder_q-layer.11": 4725.0591,
      "encoder_q-layer.2": 2307.0527,
      "encoder_q-layer.3": 2333.4902,
      "encoder_q-layer.4": 2351.4927,
      "encoder_q-layer.5": 2357.8501,
      "encoder_q-layer.6": 2570.4758,
      "encoder_q-layer.7": 2241.8457,
      "encoder_q-layer.8": 2334.7244,
      "encoder_q-layer.9": 1862.7498,
      "epoch": 0.1,
      "inbatch_neg_score": 0.4155,
      "inbatch_pos_score": 1.0078,
      "learning_rate": 4.9e-05,
      "loss": 3.88,
      "norm_diff": 0.1088,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3734.3654,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4097,
      "query_norm": 1.6561,
      "queue_k_norm": 1.5418,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0695,
      "sent_len_1": 66.8636,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.8187,
      "stdk": 0.0485,
      "stdq": 0.0446,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 9800
    },
    {
      "accuracy": 49.707,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8568,
      "doc_norm": 1.5415,
      "encoder_q-embeddings": 3521.6423,
      "encoder_q-layer.0": 2799.7722,
      "encoder_q-layer.1": 2855.229,
      "encoder_q-layer.10": 2274.8401,
      "encoder_q-layer.11": 4555.9673,
      "encoder_q-layer.2": 3607.9863,
      "encoder_q-layer.3": 3846.7373,
      "encoder_q-layer.4": 3728.3733,
      "encoder_q-layer.5": 3387.5396,
      "encoder_q-layer.6": 3067.1875,
      "encoder_q-layer.7": 2344.4485,
      "encoder_q-layer.8": 2202.8059,
      "encoder_q-layer.9": 1899.2386,
      "epoch": 0.1,
      "inbatch_neg_score": 0.371,
      "inbatch_pos_score": 1.002,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 3.8568,
      "norm_diff": 0.1678,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4858.9896,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3665,
      "query_norm": 1.7093,
      "queue_k_norm": 1.5392,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1624,
      "sent_len_1": 66.8869,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7337,
      "stdk": 0.0484,
      "stdq": 0.0464,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 9900
    },
    {
      "accuracy": 48.0469,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8404,
      "doc_norm": 1.531,
      "encoder_q-embeddings": 3037.1802,
      "encoder_q-layer.0": 2177.2124,
      "encoder_q-layer.1": 2348.6587,
      "encoder_q-layer.10": 2309.438,
      "encoder_q-layer.11": 4882.7256,
      "encoder_q-layer.2": 2721.4749,
      "encoder_q-layer.3": 2714.4055,
      "encoder_q-layer.4": 2788.7021,
      "encoder_q-layer.5": 2721.9194,
      "encoder_q-layer.6": 2684.6882,
      "encoder_q-layer.7": 2293.5112,
      "encoder_q-layer.8": 2341.2776,
      "encoder_q-layer.9": 2012.359,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3737,
      "inbatch_pos_score": 0.9526,
      "learning_rate": 5e-05,
      "loss": 3.8404,
      "norm_diff": 0.2067,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4082.5754,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3672,
      "query_norm": 1.7376,
      "queue_k_norm": 1.5345,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8207,
      "sent_len_1": 66.6558,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0488,
      "stdk": 0.0482,
      "stdq": 0.0451,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10000
    },
    {
      "dev_runtime": 37.3558,
      "dev_samples_per_second": 1.713,
      "dev_steps_per_second": 0.027,
      "epoch": 0.1,
      "step": 10000,
      "test_accuracy": 92.6513671875,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.42117059230804443,
      "test_doc_norm": 1.5123836994171143,
      "test_inbatch_neg_score": 0.7163257598876953,
      "test_inbatch_pos_score": 1.519754409790039,
      "test_loss": 0.42117059230804443,
      "test_loss_align": 2.13350772857666,
      "test_loss_unif": 3.730835437774658,
      "test_loss_unif_q@queue": 3.7308356761932373,
      "test_norm_diff": 0.18010081350803375,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.35869747400283813,
      "test_query_norm": 1.6924846172332764,
      "test_queue_k_norm": 1.534543752670288,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.041490890085697174,
      "test_stdq": 0.03906825929880142,
      "test_stdqueue_k": 0.04841500148177147,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 37.3558,
      "dev_samples_per_second": 1.713,
      "dev_steps_per_second": 0.027,
      "epoch": 0.1,
      "eval_beir-arguana_ndcg@10": 0.26148,
      "eval_beir-arguana_recall@10": 0.4495,
      "eval_beir-arguana_recall@100": 0.76885,
      "eval_beir-arguana_recall@20": 0.57824,
      "eval_beir-avg_ndcg@10": 0.1817200833333333,
      "eval_beir-avg_recall@10": 0.23396208333333335,
      "eval_beir-avg_recall@100": 0.40604483333333335,
      "eval_beir-avg_recall@20": 0.28441916666666667,
      "eval_beir-cqadupstack_ndcg@10": 0.13934083333333333,
      "eval_beir-cqadupstack_recall@10": 0.19650083333333335,
      "eval_beir-cqadupstack_recall@100": 0.37933833333333333,
      "eval_beir-cqadupstack_recall@20": 0.24443166666666669,
      "eval_beir-fiqa_ndcg@10": 0.09338,
      "eval_beir-fiqa_recall@10": 0.12618,
      "eval_beir-fiqa_recall@100": 0.30558,
      "eval_beir-fiqa_recall@20": 0.16784,
      "eval_beir-nfcorpus_ndcg@10": 0.16898,
      "eval_beir-nfcorpus_recall@10": 0.0785,
      "eval_beir-nfcorpus_recall@100": 0.19207,
      "eval_beir-nfcorpus_recall@20": 0.10706,
      "eval_beir-nq_ndcg@10": 0.10015,
      "eval_beir-nq_recall@10": 0.18342,
      "eval_beir-nq_recall@100": 0.47806,
      "eval_beir-nq_recall@20": 0.26511,
      "eval_beir-quora_ndcg@10": 0.17089,
      "eval_beir-quora_recall@10": 0.24223,
      "eval_beir-quora_recall@100": 0.40254,
      "eval_beir-quora_recall@20": 0.28077,
      "eval_beir-scidocs_ndcg@10": 0.08637,
      "eval_beir-scidocs_recall@10": 0.09362,
      "eval_beir-scidocs_recall@100": 0.2435,
      "eval_beir-scidocs_recall@20": 0.12987,
      "eval_beir-scifact_ndcg@10": 0.48329,
      "eval_beir-scifact_recall@10": 0.62172,
      "eval_beir-scifact_recall@100": 0.82844,
      "eval_beir-scifact_recall@20": 0.70294,
      "eval_beir-trec-covid_ndcg@10": 0.25987,
      "eval_beir-trec-covid_recall@10": 0.298,
      "eval_beir-trec-covid_recall@100": 0.2212,
      "eval_beir-trec-covid_recall@20": 0.287,
      "eval_beir-webis-touche2020_ndcg@10": 0.05345,
      "eval_beir-webis-touche2020_recall@10": 0.04995,
      "eval_beir-webis-touche2020_recall@100": 0.24087,
      "eval_beir-webis-touche2020_recall@20": 0.08093,
      "eval_senteval-avg_sts": 0.6678796689318702,
      "eval_senteval-sickr_spearman": 0.6164997442695914,
      "eval_senteval-stsb_spearman": 0.7192595935941491,
      "step": 10000,
      "test_accuracy": 92.6513671875,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.42117059230804443,
      "test_doc_norm": 1.5123836994171143,
      "test_inbatch_neg_score": 0.7163257598876953,
      "test_inbatch_pos_score": 1.519754409790039,
      "test_loss": 0.42117059230804443,
      "test_loss_align": 2.13350772857666,
      "test_loss_unif": 3.730835437774658,
      "test_loss_unif_q@queue": 3.7308356761932373,
      "test_norm_diff": 0.18010081350803375,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.35869747400283813,
      "test_query_norm": 1.6924846172332764,
      "test_queue_k_norm": 1.534543752670288,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.041490890085697174,
      "test_stdq": 0.03906825929880142,
      "test_stdqueue_k": 0.04841500148177147,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8527,
      "doc_norm": 1.5329,
      "encoder_q-embeddings": 9126.6426,
      "encoder_q-layer.0": 7649.1934,
      "encoder_q-layer.1": 7255.1392,
      "encoder_q-layer.10": 2286.7781,
      "encoder_q-layer.11": 4760.5239,
      "encoder_q-layer.2": 9067.2764,
      "encoder_q-layer.3": 10007.8398,
      "encoder_q-layer.4": 9854.9131,
      "encoder_q-layer.5": 9124.4883,
      "encoder_q-layer.6": 6630.6279,
      "encoder_q-layer.7": 4149.6177,
      "encoder_q-layer.8": 2981.0652,
      "encoder_q-layer.9": 2114.3831,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3936,
      "inbatch_pos_score": 1.002,
      "learning_rate": 4.994444444444445e-05,
      "loss": 3.8527,
      "norm_diff": 0.2115,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 10727.1283,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3875,
      "query_norm": 1.7445,
      "queue_k_norm": 1.5314,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8094,
      "sent_len_1": 66.6801,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4663,
      "stdk": 0.0483,
      "stdq": 0.0459,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10100
    },
    {
      "accuracy": 43.75,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8198,
      "doc_norm": 1.5327,
      "encoder_q-embeddings": 3807.8469,
      "encoder_q-layer.0": 2810.8528,
      "encoder_q-layer.1": 3383.6948,
      "encoder_q-layer.10": 2192.8823,
      "encoder_q-layer.11": 4729.8022,
      "encoder_q-layer.2": 4051.3137,
      "encoder_q-layer.3": 4327.147,
      "encoder_q-layer.4": 4400.4375,
      "encoder_q-layer.5": 4281.4028,
      "encoder_q-layer.6": 3957.0659,
      "encoder_q-layer.7": 2910.5562,
      "encoder_q-layer.8": 2732.2275,
      "encoder_q-layer.9": 2152.5439,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3915,
      "inbatch_pos_score": 0.9648,
      "learning_rate": 4.9888888888888894e-05,
      "loss": 3.8198,
      "norm_diff": 0.1514,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5377.5491,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3884,
      "query_norm": 1.6841,
      "queue_k_norm": 1.5307,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8705,
      "sent_len_1": 66.8151,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0325,
      "stdk": 0.0484,
      "stdq": 0.0455,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10200
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.826,
      "doc_norm": 1.53,
      "encoder_q-embeddings": 3528.8335,
      "encoder_q-layer.0": 2714.446,
      "encoder_q-layer.1": 2971.1436,
      "encoder_q-layer.10": 1990.5042,
      "encoder_q-layer.11": 4643.021,
      "encoder_q-layer.2": 3421.1052,
      "encoder_q-layer.3": 3610.9348,
      "encoder_q-layer.4": 3816.533,
      "encoder_q-layer.5": 3823.1963,
      "encoder_q-layer.6": 3806.8533,
      "encoder_q-layer.7": 3070.812,
      "encoder_q-layer.8": 2433.1572,
      "encoder_q-layer.9": 1802.348,
      "epoch": 0.1,
      "inbatch_neg_score": 0.4029,
      "inbatch_pos_score": 0.9976,
      "learning_rate": 4.9833333333333336e-05,
      "loss": 3.826,
      "norm_diff": 0.0889,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4903.2727,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3994,
      "query_norm": 1.6189,
      "queue_k_norm": 1.5261,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8732,
      "sent_len_1": 66.6755,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4837,
      "stdk": 0.0483,
      "stdq": 0.0454,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10300
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8361,
      "doc_norm": 1.5305,
      "encoder_q-embeddings": 5092.9858,
      "encoder_q-layer.0": 4160.6533,
      "encoder_q-layer.1": 4443.5386,
      "encoder_q-layer.10": 1863.7574,
      "encoder_q-layer.11": 4635.3784,
      "encoder_q-layer.2": 5174.2197,
      "encoder_q-layer.3": 5051.2676,
      "encoder_q-layer.4": 5035.7852,
      "encoder_q-layer.5": 5053.9819,
      "encoder_q-layer.6": 4500.6914,
      "encoder_q-layer.7": 3439.0391,
      "encoder_q-layer.8": 2555.9219,
      "encoder_q-layer.9": 1660.7036,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3886,
      "inbatch_pos_score": 0.9976,
      "learning_rate": 4.977777777777778e-05,
      "loss": 3.8361,
      "norm_diff": 0.0251,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6541.2822,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3867,
      "query_norm": 1.5556,
      "queue_k_norm": 1.5259,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0462,
      "sent_len_1": 66.7898,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2488,
      "stdk": 0.0485,
      "stdq": 0.0442,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10400
    },
    {
      "accuracy": 49.707,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8066,
      "doc_norm": 1.5202,
      "encoder_q-embeddings": 3183.2275,
      "encoder_q-layer.0": 2436.6934,
      "encoder_q-layer.1": 2649.2988,
      "encoder_q-layer.10": 1984.7385,
      "encoder_q-layer.11": 4489.6431,
      "encoder_q-layer.2": 3022.0579,
      "encoder_q-layer.3": 3347.6313,
      "encoder_q-layer.4": 3189.8015,
      "encoder_q-layer.5": 3283.4211,
      "encoder_q-layer.6": 3129.0393,
      "encoder_q-layer.7": 2726.1777,
      "encoder_q-layer.8": 2639.2686,
      "encoder_q-layer.9": 1914.3292,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3808,
      "inbatch_pos_score": 0.9624,
      "learning_rate": 4.972222222222223e-05,
      "loss": 3.8066,
      "norm_diff": 0.0622,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4417.4018,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3767,
      "query_norm": 1.5824,
      "queue_k_norm": 1.5241,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7236,
      "sent_len_1": 66.5268,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.0225,
      "stdk": 0.0482,
      "stdq": 0.0444,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10500
    },
    {
      "accuracy": 47.2656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8153,
      "doc_norm": 1.5204,
      "encoder_q-embeddings": 6467.0107,
      "encoder_q-layer.0": 5363.4141,
      "encoder_q-layer.1": 5938.9492,
      "encoder_q-layer.10": 1841.9348,
      "encoder_q-layer.11": 4281.978,
      "encoder_q-layer.2": 7152.3696,
      "encoder_q-layer.3": 7440.4116,
      "encoder_q-layer.4": 7049.3364,
      "encoder_q-layer.5": 7199.3096,
      "encoder_q-layer.6": 7204.8369,
      "encoder_q-layer.7": 7142.2749,
      "encoder_q-layer.8": 3946.3896,
      "encoder_q-layer.9": 1930.1853,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3854,
      "inbatch_pos_score": 0.9453,
      "learning_rate": 4.966666666666667e-05,
      "loss": 3.8153,
      "norm_diff": 0.0713,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8793.4157,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3813,
      "query_norm": 1.5918,
      "queue_k_norm": 1.5206,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8097,
      "sent_len_1": 66.6343,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8688,
      "stdk": 0.0482,
      "stdq": 0.0439,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 10600
    },
    {
      "accuracy": 47.8516,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8165,
      "doc_norm": 1.5187,
      "encoder_q-embeddings": 2093.6509,
      "encoder_q-layer.0": 1595.834,
      "encoder_q-layer.1": 1692.9539,
      "encoder_q-layer.10": 1823.4008,
      "encoder_q-layer.11": 4308.9785,
      "encoder_q-layer.2": 1775.3636,
      "encoder_q-layer.3": 1826.3744,
      "encoder_q-layer.4": 1931.1608,
      "encoder_q-layer.5": 1906.7373,
      "encoder_q-layer.6": 1905.8811,
      "encoder_q-layer.7": 2017.8956,
      "encoder_q-layer.8": 2036.4572,
      "encoder_q-layer.9": 1652.1947,
      "epoch": 0.1,
      "inbatch_neg_score": 0.3979,
      "inbatch_pos_score": 0.9692,
      "learning_rate": 4.961111111111111e-05,
      "loss": 3.8165,
      "norm_diff": 0.0697,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3247.6487,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3943,
      "query_norm": 1.5884,
      "queue_k_norm": 1.5201,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8869,
      "sent_len_1": 66.6298,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1925,
      "stdk": 0.0482,
      "stdq": 0.0435,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10700
    },
    {
      "accuracy": 48.9258,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8092,
      "doc_norm": 1.5166,
      "encoder_q-embeddings": 5026.9312,
      "encoder_q-layer.0": 3968.1172,
      "encoder_q-layer.1": 4203.2939,
      "encoder_q-layer.10": 1832.7747,
      "encoder_q-layer.11": 4365.3252,
      "encoder_q-layer.2": 4266.0493,
      "encoder_q-layer.3": 4388.186,
      "encoder_q-layer.4": 3903.1602,
      "encoder_q-layer.5": 3627.5889,
      "encoder_q-layer.6": 3520.8335,
      "encoder_q-layer.7": 3197.1594,
      "encoder_q-layer.8": 2447.0422,
      "encoder_q-layer.9": 1753.3865,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3925,
      "inbatch_pos_score": 1.0127,
      "learning_rate": 4.955555555555556e-05,
      "loss": 3.8092,
      "norm_diff": 0.1426,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5569.3297,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3894,
      "query_norm": 1.6592,
      "queue_k_norm": 1.5182,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0488,
      "sent_len_1": 66.7179,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9762,
      "stdk": 0.0481,
      "stdq": 0.0474,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 10800
    },
    {
      "accuracy": 45.8008,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7832,
      "doc_norm": 1.5186,
      "encoder_q-embeddings": 9006.2217,
      "encoder_q-layer.0": 6940.6367,
      "encoder_q-layer.1": 7265.8545,
      "encoder_q-layer.10": 1915.7791,
      "encoder_q-layer.11": 4700.2114,
      "encoder_q-layer.2": 9047.6377,
      "encoder_q-layer.3": 9197.207,
      "encoder_q-layer.4": 9395.9961,
      "encoder_q-layer.5": 10576.4092,
      "encoder_q-layer.6": 10681.1758,
      "encoder_q-layer.7": 10260.6982,
      "encoder_q-layer.8": 7573.458,
      "encoder_q-layer.9": 2245.3857,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3853,
      "inbatch_pos_score": 0.9492,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 3.7832,
      "norm_diff": 0.0536,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12171.1804,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3823,
      "query_norm": 1.5722,
      "queue_k_norm": 1.5181,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9138,
      "sent_len_1": 66.8879,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.1163,
      "stdk": 0.0483,
      "stdq": 0.0449,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 10900
    },
    {
      "accuracy": 50.7812,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8115,
      "doc_norm": 1.519,
      "encoder_q-embeddings": 4906.2681,
      "encoder_q-layer.0": 5366.8262,
      "encoder_q-layer.1": 5699.4824,
      "encoder_q-layer.10": 1641.1504,
      "encoder_q-layer.11": 3598.5437,
      "encoder_q-layer.2": 7036.3223,
      "encoder_q-layer.3": 5715.5039,
      "encoder_q-layer.4": 4678.5635,
      "encoder_q-layer.5": 4143.9629,
      "encoder_q-layer.6": 3898.6199,
      "encoder_q-layer.7": 3288.509,
      "encoder_q-layer.8": 2578.9526,
      "encoder_q-layer.9": 1565.1617,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3909,
      "inbatch_pos_score": 1.0029,
      "learning_rate": 4.9444444444444446e-05,
      "loss": 3.8115,
      "norm_diff": 0.0253,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7004.1396,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3877,
      "query_norm": 1.5443,
      "queue_k_norm": 1.5147,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6331,
      "sent_len_1": 66.6173,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8175,
      "stdk": 0.0483,
      "stdq": 0.0445,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 11000
    },
    {
      "accuracy": 48.3398,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8238,
      "doc_norm": 1.5179,
      "encoder_q-embeddings": 11654.9873,
      "encoder_q-layer.0": 10654.6865,
      "encoder_q-layer.1": 10746.0928,
      "encoder_q-layer.10": 2038.4668,
      "encoder_q-layer.11": 4614.9595,
      "encoder_q-layer.2": 13012.2959,
      "encoder_q-layer.3": 12863.9688,
      "encoder_q-layer.4": 13796.1162,
      "encoder_q-layer.5": 14591.7295,
      "encoder_q-layer.6": 11452.4561,
      "encoder_q-layer.7": 10041.4893,
      "encoder_q-layer.8": 6701.5894,
      "encoder_q-layer.9": 2291.3992,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3949,
      "inbatch_pos_score": 0.9849,
      "learning_rate": 4.938888888888889e-05,
      "loss": 3.8238,
      "norm_diff": 0.0198,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 15979.7269,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3916,
      "query_norm": 1.5378,
      "queue_k_norm": 1.5137,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0409,
      "sent_len_1": 66.8451,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6475,
      "stdk": 0.0483,
      "stdq": 0.0444,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 11100
    },
    {
      "accuracy": 47.9492,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7834,
      "doc_norm": 1.5192,
      "encoder_q-embeddings": 1962.4546,
      "encoder_q-layer.0": 1458.5718,
      "encoder_q-layer.1": 1617.8356,
      "encoder_q-layer.10": 848.2279,
      "encoder_q-layer.11": 1883.2726,
      "encoder_q-layer.2": 1879.4332,
      "encoder_q-layer.3": 1823.5004,
      "encoder_q-layer.4": 1765.0256,
      "encoder_q-layer.5": 1799.6224,
      "encoder_q-layer.6": 1716.7664,
      "encoder_q-layer.7": 1599.3937,
      "encoder_q-layer.8": 1682.4143,
      "encoder_q-layer.9": 1159.5663,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3977,
      "inbatch_pos_score": 0.9805,
      "learning_rate": 4.933333333333334e-05,
      "loss": 3.7834,
      "norm_diff": 0.0064,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2491.6134,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3948,
      "query_norm": 1.5173,
      "queue_k_norm": 1.515,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9445,
      "sent_len_1": 66.9824,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3325,
      "stdk": 0.0483,
      "stdq": 0.0432,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 11200
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8175,
      "doc_norm": 1.5084,
      "encoder_q-embeddings": 598.2411,
      "encoder_q-layer.0": 452.1927,
      "encoder_q-layer.1": 482.4727,
      "encoder_q-layer.10": 419.9093,
      "encoder_q-layer.11": 1103.7334,
      "encoder_q-layer.2": 529.6849,
      "encoder_q-layer.3": 534.6472,
      "encoder_q-layer.4": 539.7852,
      "encoder_q-layer.5": 534.9767,
      "encoder_q-layer.6": 526.1143,
      "encoder_q-layer.7": 542.257,
      "encoder_q-layer.8": 555.155,
      "encoder_q-layer.9": 409.3682,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3696,
      "inbatch_pos_score": 0.9795,
      "learning_rate": 4.927777777777778e-05,
      "loss": 3.8175,
      "norm_diff": 0.0191,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 852.8091,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3689,
      "query_norm": 1.521,
      "queue_k_norm": 1.5113,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9675,
      "sent_len_1": 66.6418,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8162,
      "stdk": 0.048,
      "stdq": 0.0441,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 11300
    },
    {
      "accuracy": 51.1719,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7944,
      "doc_norm": 1.5071,
      "encoder_q-embeddings": 1256.6489,
      "encoder_q-layer.0": 973.5815,
      "encoder_q-layer.1": 1042.5499,
      "encoder_q-layer.10": 436.2685,
      "encoder_q-layer.11": 1126.8124,
      "encoder_q-layer.2": 1162.8326,
      "encoder_q-layer.3": 1221.2096,
      "encoder_q-layer.4": 1269.3165,
      "encoder_q-layer.5": 1302.2251,
      "encoder_q-layer.6": 1189.7611,
      "encoder_q-layer.7": 1063.6256,
      "encoder_q-layer.8": 847.2612,
      "encoder_q-layer.9": 476.1317,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3657,
      "inbatch_pos_score": 0.957,
      "learning_rate": 4.922222222222222e-05,
      "loss": 3.7944,
      "norm_diff": 0.0099,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1590.7854,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3633,
      "query_norm": 1.4989,
      "queue_k_norm": 1.5106,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0055,
      "sent_len_1": 66.8457,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.17,
      "stdk": 0.048,
      "stdq": 0.0434,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 11400
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7666,
      "doc_norm": 1.5168,
      "encoder_q-embeddings": 781.6425,
      "encoder_q-layer.0": 599.9102,
      "encoder_q-layer.1": 623.575,
      "encoder_q-layer.10": 458.3023,
      "encoder_q-layer.11": 1046.6364,
      "encoder_q-layer.2": 732.7254,
      "encoder_q-layer.3": 788.2367,
      "encoder_q-layer.4": 823.9196,
      "encoder_q-layer.5": 839.5878,
      "encoder_q-layer.6": 888.6884,
      "encoder_q-layer.7": 873.8758,
      "encoder_q-layer.8": 916.8347,
      "encoder_q-layer.9": 533.5176,
      "epoch": 0.11,
      "inbatch_neg_score": 0.35,
      "inbatch_pos_score": 0.9912,
      "learning_rate": 4.9166666666666665e-05,
      "loss": 3.7666,
      "norm_diff": 0.0108,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1130.5347,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3491,
      "query_norm": 1.5172,
      "queue_k_norm": 1.5094,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7696,
      "sent_len_1": 66.6455,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8275,
      "stdk": 0.0484,
      "stdq": 0.0452,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 11500
    },
    {
      "accuracy": 47.9492,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7891,
      "doc_norm": 1.5101,
      "encoder_q-embeddings": 1715.0293,
      "encoder_q-layer.0": 1282.1475,
      "encoder_q-layer.1": 1421.3412,
      "encoder_q-layer.10": 452.36,
      "encoder_q-layer.11": 937.1433,
      "encoder_q-layer.2": 1458.2859,
      "encoder_q-layer.3": 1580.1786,
      "encoder_q-layer.4": 1601.8695,
      "encoder_q-layer.5": 1397.0217,
      "encoder_q-layer.6": 1232.0605,
      "encoder_q-layer.7": 1280.9523,
      "encoder_q-layer.8": 890.4059,
      "encoder_q-layer.9": 502.9158,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3522,
      "inbatch_pos_score": 0.936,
      "learning_rate": 4.9111111111111114e-05,
      "loss": 3.7891,
      "norm_diff": 0.0313,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1926.2904,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3501,
      "query_norm": 1.4787,
      "queue_k_norm": 1.507,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9652,
      "sent_len_1": 66.8126,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5263,
      "stdk": 0.0482,
      "stdq": 0.0443,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 11600
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7857,
      "doc_norm": 1.511,
      "encoder_q-embeddings": 1475.4409,
      "encoder_q-layer.0": 1081.061,
      "encoder_q-layer.1": 1234.4814,
      "encoder_q-layer.10": 428.0195,
      "encoder_q-layer.11": 1025.491,
      "encoder_q-layer.2": 1261.8141,
      "encoder_q-layer.3": 1326.8323,
      "encoder_q-layer.4": 1409.1532,
      "encoder_q-layer.5": 1509.5879,
      "encoder_q-layer.6": 1509.7963,
      "encoder_q-layer.7": 1629.637,
      "encoder_q-layer.8": 1627.3967,
      "encoder_q-layer.9": 758.9493,
      "epoch": 0.11,
      "inbatch_neg_score": 0.3525,
      "inbatch_pos_score": 0.9795,
      "learning_rate": 4.905555555555556e-05,
      "loss": 3.7857,
      "norm_diff": 0.0248,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1897.035,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3518,
      "query_norm": 1.4862,
      "queue_k_norm": 1.5069,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0807,
      "sent_len_1": 66.9321,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.745,
      "stdk": 0.0482,
      "stdq": 0.0445,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 11700
    },
    {
      "accuracy": 46.9727,
      "active_queue_size": 16384.0,
      "cl_loss": 3.755,
      "doc_norm": 1.4982,
      "encoder_q-embeddings": 1723.7433,
      "encoder_q-layer.0": 1755.9176,
      "encoder_q-layer.1": 1924.1892,
      "encoder_q-layer.10": 425.4989,
      "encoder_q-layer.11": 920.5051,
      "encoder_q-layer.2": 2064.9136,
      "encoder_q-layer.3": 1225.0648,
      "encoder_q-layer.4": 1254.9219,
      "encoder_q-layer.5": 1227.0945,
      "encoder_q-layer.6": 1204.5237,
      "encoder_q-layer.7": 1053.7251,
      "encoder_q-layer.8": 835.418,
      "encoder_q-layer.9": 532.4515,
      "epoch": 0.12,
      "inbatch_neg_score": 0.3383,
      "inbatch_pos_score": 0.9404,
      "learning_rate": 4.9e-05,
      "loss": 3.755,
      "norm_diff": 0.0049,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2053.2782,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3354,
      "query_norm": 1.4977,
      "queue_k_norm": 1.504,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9645,
      "sent_len_1": 66.996,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.59,
      "stdk": 0.0478,
      "stdq": 0.0448,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 11800
    },
    {
      "accuracy": 48.0469,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7907,
      "doc_norm": 1.5023,
      "encoder_q-embeddings": 1300.2075,
      "encoder_q-layer.0": 989.0618,
      "encoder_q-layer.1": 1105.1931,
      "encoder_q-layer.10": 429.1985,
      "encoder_q-layer.11": 924.9601,
      "encoder_q-layer.2": 1313.5441,
      "encoder_q-layer.3": 1345.287,
      "encoder_q-layer.4": 1423.4465,
      "encoder_q-layer.5": 1429.0365,
      "encoder_q-layer.6": 1400.5262,
      "encoder_q-layer.7": 1089.5223,
      "encoder_q-layer.8": 667.5069,
      "encoder_q-layer.9": 420.3381,
      "epoch": 0.12,
      "inbatch_neg_score": 0.3202,
      "inbatch_pos_score": 0.9189,
      "learning_rate": 4.894444444444445e-05,
      "loss": 3.7907,
      "norm_diff": 0.0427,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1681.81,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3193,
      "query_norm": 1.4596,
      "queue_k_norm": 1.502,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0614,
      "sent_len_1": 67.0893,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9888,
      "stdk": 0.048,
      "stdq": 0.0437,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 11900
    },
    {
      "accuracy": 46.6797,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7701,
      "doc_norm": 1.5014,
      "encoder_q-embeddings": 2925.7332,
      "encoder_q-layer.0": 2205.3018,
      "encoder_q-layer.1": 2256.5393,
      "encoder_q-layer.10": 431.851,
      "encoder_q-layer.11": 913.819,
      "encoder_q-layer.2": 2395.676,
      "encoder_q-layer.3": 2430.8835,
      "encoder_q-layer.4": 2358.5859,
      "encoder_q-layer.5": 2470.3872,
      "encoder_q-layer.6": 2583.0725,
      "encoder_q-layer.7": 2256.8599,
      "encoder_q-layer.8": 1329.7,
      "encoder_q-layer.9": 519.6412,
      "epoch": 0.12,
      "inbatch_neg_score": 0.3325,
      "inbatch_pos_score": 0.9253,
      "learning_rate": 4.888888888888889e-05,
      "loss": 3.7701,
      "norm_diff": 0.034,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3255.1087,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3301,
      "query_norm": 1.4674,
      "queue_k_norm": 1.4987,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7794,
      "sent_len_1": 66.5976,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.1725,
      "stdk": 0.0481,
      "stdq": 0.044,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 12000
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7758,
      "doc_norm": 1.5002,
      "encoder_q-embeddings": 1659.0184,
      "encoder_q-layer.0": 1230.9836,
      "encoder_q-layer.1": 1366.6635,
      "encoder_q-layer.10": 416.1913,
      "encoder_q-layer.11": 1232.2761,
      "encoder_q-layer.2": 1731.7726,
      "encoder_q-layer.3": 1879.1377,
      "encoder_q-layer.4": 1819.7449,
      "encoder_q-layer.5": 2067.6311,
      "encoder_q-layer.6": 1737.8129,
      "encoder_q-layer.7": 1241.5262,
      "encoder_q-layer.8": 833.4688,
      "encoder_q-layer.9": 476.087,
      "epoch": 0.12,
      "inbatch_neg_score": 0.2987,
      "inbatch_pos_score": 0.8984,
      "learning_rate": 4.883333333333334e-05,
      "loss": 3.7758,
      "norm_diff": 0.0344,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2155.0325,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2986,
      "query_norm": 1.4658,
      "queue_k_norm": 1.4955,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.858,
      "sent_len_1": 66.7496,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2862,
      "stdk": 0.0481,
      "stdq": 0.0441,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 12100
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7278,
      "doc_norm": 1.4922,
      "encoder_q-embeddings": 1070.7878,
      "encoder_q-layer.0": 786.878,
      "encoder_q-layer.1": 941.4999,
      "encoder_q-layer.10": 383.8946,
      "encoder_q-layer.11": 973.4283,
      "encoder_q-layer.2": 1093.4214,
      "encoder_q-layer.3": 1164.6896,
      "encoder_q-layer.4": 1290.7715,
      "encoder_q-layer.5": 1141.9205,
      "encoder_q-layer.6": 1323.0315,
      "encoder_q-layer.7": 1079.6373,
      "encoder_q-layer.8": 757.5689,
      "encoder_q-layer.9": 509.3365,
      "epoch": 0.12,
      "inbatch_neg_score": 0.287,
      "inbatch_pos_score": 0.9092,
      "learning_rate": 4.8777777777777775e-05,
      "loss": 3.7278,
      "norm_diff": 0.027,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1488.3412,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2864,
      "query_norm": 1.4656,
      "queue_k_norm": 1.4925,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1276,
      "sent_len_1": 67.0253,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7862,
      "stdk": 0.0479,
      "stdq": 0.0445,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 12200
    },
    {
      "accuracy": 48.7305,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7959,
      "doc_norm": 1.4882,
      "encoder_q-embeddings": 3817.4216,
      "encoder_q-layer.0": 2915.0354,
      "encoder_q-layer.1": 3526.8811,
      "encoder_q-layer.10": 428.5457,
      "encoder_q-layer.11": 1024.1432,
      "encoder_q-layer.2": 3564.1799,
      "encoder_q-layer.3": 3328.9626,
      "encoder_q-layer.4": 3126.2471,
      "encoder_q-layer.5": 2898.2136,
      "encoder_q-layer.6": 2388.1985,
      "encoder_q-layer.7": 1448.3685,
      "encoder_q-layer.8": 865.1682,
      "encoder_q-layer.9": 534.4011,
      "epoch": 0.12,
      "inbatch_neg_score": 0.2897,
      "inbatch_pos_score": 0.8979,
      "learning_rate": 4.8722222222222224e-05,
      "loss": 3.7959,
      "norm_diff": 0.024,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4030.6605,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2876,
      "query_norm": 1.4642,
      "queue_k_norm": 1.4919,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9609,
      "sent_len_1": 66.7459,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3688,
      "stdk": 0.0478,
      "stdq": 0.0449,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 12300
    },
    {
      "accuracy": 49.6094,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7613,
      "doc_norm": 1.4923,
      "encoder_q-embeddings": 3849.2231,
      "encoder_q-layer.0": 2869.9299,
      "encoder_q-layer.1": 2915.3867,
      "encoder_q-layer.10": 403.0017,
      "encoder_q-layer.11": 927.0325,
      "encoder_q-layer.2": 3165.5605,
      "encoder_q-layer.3": 3598.7512,
      "encoder_q-layer.4": 3609.0828,
      "encoder_q-layer.5": 3408.1492,
      "encoder_q-layer.6": 3036.3408,
      "encoder_q-layer.7": 2408.4207,
      "encoder_q-layer.8": 1595.8793,
      "encoder_q-layer.9": 610.8804,
      "epoch": 0.12,
      "inbatch_neg_score": 0.2949,
      "inbatch_pos_score": 0.9106,
      "learning_rate": 4.866666666666667e-05,
      "loss": 3.7613,
      "norm_diff": 0.0214,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4171.6513,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2939,
      "query_norm": 1.4709,
      "queue_k_norm": 1.4872,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7347,
      "sent_len_1": 66.6425,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.085,
      "stdk": 0.048,
      "stdq": 0.0446,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 12400
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8,
      "doc_norm": 1.4868,
      "encoder_q-embeddings": 1724.8857,
      "encoder_q-layer.0": 1306.5692,
      "encoder_q-layer.1": 1340.756,
      "encoder_q-layer.10": 392.868,
      "encoder_q-layer.11": 869.6208,
      "encoder_q-layer.2": 1379.0487,
      "encoder_q-layer.3": 1369.6888,
      "encoder_q-layer.4": 1269.3848,
      "encoder_q-layer.5": 1093.7417,
      "encoder_q-layer.6": 972.4558,
      "encoder_q-layer.7": 707.7662,
      "encoder_q-layer.8": 505.7879,
      "encoder_q-layer.9": 381.732,
      "epoch": 0.12,
      "inbatch_neg_score": 0.2922,
      "inbatch_pos_score": 0.9326,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 3.8,
      "norm_diff": 0.0151,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1703.9897,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2913,
      "query_norm": 1.4717,
      "queue_k_norm": 1.4842,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9319,
      "sent_len_1": 66.9227,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6012,
      "stdk": 0.0479,
      "stdq": 0.0441,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 12500
    },
    {
      "accuracy": 51.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7904,
      "doc_norm": 1.4799,
      "encoder_q-embeddings": 2584.7935,
      "encoder_q-layer.0": 1870.4818,
      "encoder_q-layer.1": 2363.5754,
      "encoder_q-layer.10": 404.9762,
      "encoder_q-layer.11": 999.093,
      "encoder_q-layer.2": 2561.6116,
      "encoder_q-layer.3": 2555.0261,
      "encoder_q-layer.4": 2778.4517,
      "encoder_q-layer.5": 3003.4722,
      "encoder_q-layer.6": 3175.4131,
      "encoder_q-layer.7": 2936.5498,
      "encoder_q-layer.8": 2223.2451,
      "encoder_q-layer.9": 842.515,
      "epoch": 0.12,
      "inbatch_neg_score": 0.2993,
      "inbatch_pos_score": 0.9199,
      "learning_rate": 4.855555555555556e-05,
      "loss": 3.7904,
      "norm_diff": 0.042,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3415.5143,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2998,
      "query_norm": 1.5219,
      "queue_k_norm": 1.4815,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8902,
      "sent_len_1": 66.8596,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8913,
      "stdk": 0.0477,
      "stdq": 0.045,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 12600
    },
    {
      "accuracy": 49.5117,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8249,
      "doc_norm": 1.4749,
      "encoder_q-embeddings": 1280.451,
      "encoder_q-layer.0": 879.083,
      "encoder_q-layer.1": 1032.9998,
      "encoder_q-layer.10": 389.7971,
      "encoder_q-layer.11": 1020.317,
      "encoder_q-layer.2": 1019.4603,
      "encoder_q-layer.3": 1159.4561,
      "encoder_q-layer.4": 1437.1007,
      "encoder_q-layer.5": 1294.5591,
      "encoder_q-layer.6": 1056.8418,
      "encoder_q-layer.7": 818.9913,
      "encoder_q-layer.8": 618.0167,
      "encoder_q-layer.9": 447.2755,
      "epoch": 0.12,
      "inbatch_neg_score": 0.3011,
      "inbatch_pos_score": 0.9014,
      "learning_rate": 4.85e-05,
      "loss": 3.8249,
      "norm_diff": 0.0466,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1498.7808,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3003,
      "query_norm": 1.5215,
      "queue_k_norm": 1.4765,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.952,
      "sent_len_1": 66.7794,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2937,
      "stdk": 0.0477,
      "stdq": 0.0446,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 12700
    },
    {
      "accuracy": 48.2422,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7746,
      "doc_norm": 1.467,
      "encoder_q-embeddings": 971.1915,
      "encoder_q-layer.0": 716.5073,
      "encoder_q-layer.1": 793.5526,
      "encoder_q-layer.10": 426.6133,
      "encoder_q-layer.11": 965.2045,
      "encoder_q-layer.2": 822.8741,
      "encoder_q-layer.3": 874.7257,
      "encoder_q-layer.4": 858.3135,
      "encoder_q-layer.5": 913.206,
      "encoder_q-layer.6": 849.5543,
      "encoder_q-layer.7": 694.137,
      "encoder_q-layer.8": 624.3719,
      "encoder_q-layer.9": 434.1303,
      "epoch": 0.12,
      "inbatch_neg_score": 0.3102,
      "inbatch_pos_score": 0.8989,
      "learning_rate": 4.844444444444445e-05,
      "loss": 3.7746,
      "norm_diff": 0.0391,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1171.5425,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3081,
      "query_norm": 1.5061,
      "queue_k_norm": 1.4713,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9294,
      "sent_len_1": 66.7471,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3088,
      "stdk": 0.0475,
      "stdq": 0.0438,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 12800
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8324,
      "doc_norm": 1.4648,
      "encoder_q-embeddings": 8559.5244,
      "encoder_q-layer.0": 6427.5249,
      "encoder_q-layer.1": 6577.3276,
      "encoder_q-layer.10": 423.8161,
      "encoder_q-layer.11": 1010.9144,
      "encoder_q-layer.2": 7428.4478,
      "encoder_q-layer.3": 7648.1519,
      "encoder_q-layer.4": 7765.6934,
      "encoder_q-layer.5": 6665.1035,
      "encoder_q-layer.6": 6173.5576,
      "encoder_q-layer.7": 5069.1138,
      "encoder_q-layer.8": 2539.0054,
      "encoder_q-layer.9": 694.0531,
      "epoch": 0.13,
      "inbatch_neg_score": 0.3133,
      "inbatch_pos_score": 0.937,
      "learning_rate": 4.838888888888889e-05,
      "loss": 3.8324,
      "norm_diff": 0.0501,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9114.6638,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3125,
      "query_norm": 1.5148,
      "queue_k_norm": 1.4663,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9291,
      "sent_len_1": 66.7826,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0025,
      "stdk": 0.0475,
      "stdq": 0.045,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 12900
    },
    {
      "accuracy": 48.3398,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8114,
      "doc_norm": 1.4696,
      "encoder_q-embeddings": 1792.8325,
      "encoder_q-layer.0": 1210.3773,
      "encoder_q-layer.1": 1469.3761,
      "encoder_q-layer.10": 428.5696,
      "encoder_q-layer.11": 954.15,
      "encoder_q-layer.2": 1693.5155,
      "encoder_q-layer.3": 1796.7406,
      "encoder_q-layer.4": 2054.5073,
      "encoder_q-layer.5": 1962.6506,
      "encoder_q-layer.6": 1618.5724,
      "encoder_q-layer.7": 1054.6565,
      "encoder_q-layer.8": 634.4353,
      "encoder_q-layer.9": 430.726,
      "epoch": 0.13,
      "inbatch_neg_score": 0.3195,
      "inbatch_pos_score": 0.9043,
      "learning_rate": 4.8333333333333334e-05,
      "loss": 3.8114,
      "norm_diff": 0.014,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2128.4061,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3188,
      "query_norm": 1.4813,
      "queue_k_norm": 1.4637,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.965,
      "sent_len_1": 66.8035,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4925,
      "stdk": 0.0478,
      "stdq": 0.0435,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 13000
    },
    {
      "accuracy": 47.6562,
      "active_queue_size": 16384.0,
      "cl_loss": 3.748,
      "doc_norm": 1.4594,
      "encoder_q-embeddings": 3085.5171,
      "encoder_q-layer.0": 2323.0063,
      "encoder_q-layer.1": 2660.0283,
      "encoder_q-layer.10": 407.7632,
      "encoder_q-layer.11": 1070.3639,
      "encoder_q-layer.2": 2930.7327,
      "encoder_q-layer.3": 3196.678,
      "encoder_q-layer.4": 3611.0891,
      "encoder_q-layer.5": 3837.4463,
      "encoder_q-layer.6": 3979.0002,
      "encoder_q-layer.7": 3357.2705,
      "encoder_q-layer.8": 1326.553,
      "encoder_q-layer.9": 606.3932,
      "epoch": 0.13,
      "inbatch_neg_score": 0.2741,
      "inbatch_pos_score": 0.8574,
      "learning_rate": 4.8277777777777776e-05,
      "loss": 3.748,
      "norm_diff": 0.0114,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4116.9472,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2715,
      "query_norm": 1.4615,
      "queue_k_norm": 1.4602,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6497,
      "sent_len_1": 66.7572,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.505,
      "stdk": 0.0474,
      "stdq": 0.0442,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 13100
    },
    {
      "accuracy": 48.5352,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7714,
      "doc_norm": 1.4572,
      "encoder_q-embeddings": 2867.1382,
      "encoder_q-layer.0": 2068.6909,
      "encoder_q-layer.1": 2501.4746,
      "encoder_q-layer.10": 232.044,
      "encoder_q-layer.11": 483.0477,
      "encoder_q-layer.2": 3178.6438,
      "encoder_q-layer.3": 3242.0679,
      "encoder_q-layer.4": 3549.3455,
      "encoder_q-layer.5": 3032.9878,
      "encoder_q-layer.6": 3557.541,
      "encoder_q-layer.7": 3228.9175,
      "encoder_q-layer.8": 1314.3352,
      "encoder_q-layer.9": 309.3714,
      "epoch": 0.13,
      "inbatch_neg_score": 0.273,
      "inbatch_pos_score": 0.8623,
      "learning_rate": 4.8222222222222225e-05,
      "loss": 3.7714,
      "norm_diff": 0.0131,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3941.7886,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2734,
      "query_norm": 1.4491,
      "queue_k_norm": 1.4578,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8796,
      "sent_len_1": 66.7485,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7063,
      "stdk": 0.0474,
      "stdq": 0.0435,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 13200
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7491,
      "doc_norm": 1.4534,
      "encoder_q-embeddings": 1620.7775,
      "encoder_q-layer.0": 1115.4755,
      "encoder_q-layer.1": 1232.1075,
      "encoder_q-layer.10": 200.7894,
      "encoder_q-layer.11": 523.566,
      "encoder_q-layer.2": 1512.2428,
      "encoder_q-layer.3": 1536.4409,
      "encoder_q-layer.4": 1456.811,
      "encoder_q-layer.5": 1227.5809,
      "encoder_q-layer.6": 954.4739,
      "encoder_q-layer.7": 685.7849,
      "encoder_q-layer.8": 420.0695,
      "encoder_q-layer.9": 211.886,
      "epoch": 0.13,
      "inbatch_neg_score": 0.2858,
      "inbatch_pos_score": 0.896,
      "learning_rate": 4.8166666666666674e-05,
      "loss": 3.7491,
      "norm_diff": 0.0236,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1677.6978,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2834,
      "query_norm": 1.4767,
      "queue_k_norm": 1.4568,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1258,
      "sent_len_1": 66.8954,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1262,
      "stdk": 0.0473,
      "stdq": 0.044,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 13300
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7654,
      "doc_norm": 1.4548,
      "encoder_q-embeddings": 1104.2333,
      "encoder_q-layer.0": 788.7063,
      "encoder_q-layer.1": 906.4,
      "encoder_q-layer.10": 196.6455,
      "encoder_q-layer.11": 461.0955,
      "encoder_q-layer.2": 1114.1243,
      "encoder_q-layer.3": 1090.0879,
      "encoder_q-layer.4": 1069.2125,
      "encoder_q-layer.5": 1000.2854,
      "encoder_q-layer.6": 967.7065,
      "encoder_q-layer.7": 811.5062,
      "encoder_q-layer.8": 439.9344,
      "encoder_q-layer.9": 203.2442,
      "epoch": 0.13,
      "inbatch_neg_score": 0.2884,
      "inbatch_pos_score": 0.9023,
      "learning_rate": 4.811111111111111e-05,
      "loss": 3.7654,
      "norm_diff": 0.0586,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1276.8469,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2866,
      "query_norm": 1.5134,
      "queue_k_norm": 1.4531,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9198,
      "sent_len_1": 66.6096,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7912,
      "stdk": 0.0475,
      "stdq": 0.0446,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 13400
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7721,
      "doc_norm": 1.4565,
      "encoder_q-embeddings": 798.809,
      "encoder_q-layer.0": 572.8371,
      "encoder_q-layer.1": 686.9156,
      "encoder_q-layer.10": 179.0298,
      "encoder_q-layer.11": 525.9814,
      "encoder_q-layer.2": 792.1835,
      "encoder_q-layer.3": 789.966,
      "encoder_q-layer.4": 694.6115,
      "encoder_q-layer.5": 715.4411,
      "encoder_q-layer.6": 715.8323,
      "encoder_q-layer.7": 539.1501,
      "encoder_q-layer.8": 336.9373,
      "encoder_q-layer.9": 185.6522,
      "epoch": 0.13,
      "inbatch_neg_score": 0.3018,
      "inbatch_pos_score": 0.9111,
      "learning_rate": 4.805555555555556e-05,
      "loss": 3.7721,
      "norm_diff": 0.0391,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 922.3438,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3013,
      "query_norm": 1.4956,
      "queue_k_norm": 1.4506,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8971,
      "sent_len_1": 66.6873,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0513,
      "stdk": 0.0476,
      "stdq": 0.044,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 13500
    },
    {
      "accuracy": 47.2656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8003,
      "doc_norm": 1.4398,
      "encoder_q-embeddings": 435.5981,
      "encoder_q-layer.0": 310.6275,
      "encoder_q-layer.1": 356.2498,
      "encoder_q-layer.10": 193.2817,
      "encoder_q-layer.11": 506.2979,
      "encoder_q-layer.2": 396.5273,
      "encoder_q-layer.3": 359.9361,
      "encoder_q-layer.4": 344.3542,
      "encoder_q-layer.5": 324.8602,
      "encoder_q-layer.6": 380.1005,
      "encoder_q-layer.7": 346.1803,
      "encoder_q-layer.8": 269.0069,
      "encoder_q-layer.9": 182.7588,
      "epoch": 0.13,
      "inbatch_neg_score": 0.317,
      "inbatch_pos_score": 0.9028,
      "learning_rate": 4.8e-05,
      "loss": 3.8003,
      "norm_diff": 0.1011,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 532.6078,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3145,
      "query_norm": 1.541,
      "queue_k_norm": 1.4474,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8766,
      "sent_len_1": 66.718,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6037,
      "stdk": 0.0471,
      "stdq": 0.044,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 13600
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8109,
      "doc_norm": 1.4457,
      "encoder_q-embeddings": 4997.5889,
      "encoder_q-layer.0": 3809.2131,
      "encoder_q-layer.1": 3694.7424,
      "encoder_q-layer.10": 231.4,
      "encoder_q-layer.11": 570.2198,
      "encoder_q-layer.2": 3817.0847,
      "encoder_q-layer.3": 3713.9468,
      "encoder_q-layer.4": 3590.8774,
      "encoder_q-layer.5": 3446.105,
      "encoder_q-layer.6": 3089.9011,
      "encoder_q-layer.7": 1863.7864,
      "encoder_q-layer.8": 900.6384,
      "encoder_q-layer.9": 282.9417,
      "epoch": 0.13,
      "inbatch_neg_score": 0.2943,
      "inbatch_pos_score": 0.8892,
      "learning_rate": 4.794444444444445e-05,
      "loss": 3.8109,
      "norm_diff": 0.0409,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4744.878,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2913,
      "query_norm": 1.4865,
      "queue_k_norm": 1.4435,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1424,
      "sent_len_1": 66.7509,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3837,
      "stdk": 0.0474,
      "stdq": 0.0433,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 13700
    },
    {
      "accuracy": 48.3398,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8204,
      "doc_norm": 1.4422,
      "encoder_q-embeddings": 1263.9434,
      "encoder_q-layer.0": 889.8225,
      "encoder_q-layer.1": 952.1967,
      "encoder_q-layer.10": 197.67,
      "encoder_q-layer.11": 552.4765,
      "encoder_q-layer.2": 999.1175,
      "encoder_q-layer.3": 975.7026,
      "encoder_q-layer.4": 959.0341,
      "encoder_q-layer.5": 939.6488,
      "encoder_q-layer.6": 912.5176,
      "encoder_q-layer.7": 714.6954,
      "encoder_q-layer.8": 470.8221,
      "encoder_q-layer.9": 205.3938,
      "epoch": 0.13,
      "inbatch_neg_score": 0.294,
      "inbatch_pos_score": 0.8989,
      "learning_rate": 4.7888888888888886e-05,
      "loss": 3.8204,
      "norm_diff": 0.1043,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1287.1149,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2917,
      "query_norm": 1.5465,
      "queue_k_norm": 1.4396,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0157,
      "sent_len_1": 66.9074,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0863,
      "stdk": 0.0473,
      "stdq": 0.0454,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 13800
    },
    {
      "accuracy": 50.7812,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8002,
      "doc_norm": 1.4369,
      "encoder_q-embeddings": 2190.3188,
      "encoder_q-layer.0": 1641.8998,
      "encoder_q-layer.1": 1760.7063,
      "encoder_q-layer.10": 193.0977,
      "encoder_q-layer.11": 498.0404,
      "encoder_q-layer.2": 2018.7343,
      "encoder_q-layer.3": 2104.27,
      "encoder_q-layer.4": 2344.3547,
      "encoder_q-layer.5": 2310.738,
      "encoder_q-layer.6": 2475.2764,
      "encoder_q-layer.7": 2305.7166,
      "encoder_q-layer.8": 1235.2456,
      "encoder_q-layer.9": 285.6683,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2862,
      "inbatch_pos_score": 0.8965,
      "learning_rate": 4.7833333333333335e-05,
      "loss": 3.8002,
      "norm_diff": 0.0424,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2726.532,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2847,
      "query_norm": 1.4792,
      "queue_k_norm": 1.4375,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0018,
      "sent_len_1": 66.8779,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.9875,
      "stdk": 0.0472,
      "stdq": 0.0442,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 13900
    },
    {
      "accuracy": 45.9961,
      "active_queue_size": 16384.0,
      "cl_loss": 3.8111,
      "doc_norm": 1.4357,
      "encoder_q-embeddings": 2751.6443,
      "encoder_q-layer.0": 1939.4822,
      "encoder_q-layer.1": 2168.0828,
      "encoder_q-layer.10": 212.5211,
      "encoder_q-layer.11": 434.0795,
      "encoder_q-layer.2": 2391.2361,
      "encoder_q-layer.3": 2669.8669,
      "encoder_q-layer.4": 2925.283,
      "encoder_q-layer.5": 2680.0696,
      "encoder_q-layer.6": 2443.1284,
      "encoder_q-layer.7": 2063.3191,
      "encoder_q-layer.8": 1215.4247,
      "encoder_q-layer.9": 333.4556,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2619,
      "inbatch_pos_score": 0.8413,
      "learning_rate": 4.7777777777777784e-05,
      "loss": 3.8111,
      "norm_diff": 0.0258,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3191.7227,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.26,
      "query_norm": 1.4615,
      "queue_k_norm": 1.434,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9663,
      "sent_len_1": 66.6392,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.68,
      "stdk": 0.0473,
      "stdq": 0.0445,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 14000
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7773,
      "doc_norm": 1.4322,
      "encoder_q-embeddings": 795.2767,
      "encoder_q-layer.0": 591.1854,
      "encoder_q-layer.1": 604.8469,
      "encoder_q-layer.10": 215.9927,
      "encoder_q-layer.11": 547.3765,
      "encoder_q-layer.2": 659.8074,
      "encoder_q-layer.3": 714.7322,
      "encoder_q-layer.4": 803.6387,
      "encoder_q-layer.5": 732.9355,
      "encoder_q-layer.6": 712.5629,
      "encoder_q-layer.7": 465.6397,
      "encoder_q-layer.8": 304.2286,
      "encoder_q-layer.9": 197.7587,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2428,
      "inbatch_pos_score": 0.8315,
      "learning_rate": 4.7722222222222226e-05,
      "loss": 3.7773,
      "norm_diff": 0.0584,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 896.1205,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2416,
      "query_norm": 1.4906,
      "queue_k_norm": 1.4317,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8507,
      "sent_len_1": 66.7548,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1637,
      "stdk": 0.0472,
      "stdq": 0.0442,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 14100
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7557,
      "doc_norm": 1.4255,
      "encoder_q-embeddings": 422.2756,
      "encoder_q-layer.0": 314.3701,
      "encoder_q-layer.1": 328.9502,
      "encoder_q-layer.10": 179.3673,
      "encoder_q-layer.11": 400.5014,
      "encoder_q-layer.2": 365.5537,
      "encoder_q-layer.3": 368.4449,
      "encoder_q-layer.4": 349.9974,
      "encoder_q-layer.5": 335.298,
      "encoder_q-layer.6": 350.9282,
      "encoder_q-layer.7": 359.1332,
      "encoder_q-layer.8": 317.0159,
      "encoder_q-layer.9": 209.487,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2332,
      "inbatch_pos_score": 0.856,
      "learning_rate": 4.766666666666667e-05,
      "loss": 3.7557,
      "norm_diff": 0.09,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 502.082,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2324,
      "query_norm": 1.5155,
      "queue_k_norm": 1.4282,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9606,
      "sent_len_1": 66.7424,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6012,
      "stdk": 0.0471,
      "stdq": 0.0451,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 14200
    },
    {
      "accuracy": 50.293,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7827,
      "doc_norm": 1.4333,
      "encoder_q-embeddings": 1015.936,
      "encoder_q-layer.0": 819.5609,
      "encoder_q-layer.1": 875.043,
      "encoder_q-layer.10": 192.5464,
      "encoder_q-layer.11": 451.9021,
      "encoder_q-layer.2": 917.2497,
      "encoder_q-layer.3": 931.4559,
      "encoder_q-layer.4": 936.2823,
      "encoder_q-layer.5": 955.656,
      "encoder_q-layer.6": 835.4227,
      "encoder_q-layer.7": 768.1489,
      "encoder_q-layer.8": 466.4324,
      "encoder_q-layer.9": 219.0986,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2526,
      "inbatch_pos_score": 0.8779,
      "learning_rate": 4.761111111111111e-05,
      "loss": 3.7827,
      "norm_diff": 0.0999,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1163.9857,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2524,
      "query_norm": 1.5333,
      "queue_k_norm": 1.4245,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8515,
      "sent_len_1": 66.6147,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5462,
      "stdk": 0.0475,
      "stdq": 0.0453,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 14300
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7477,
      "doc_norm": 1.4198,
      "encoder_q-embeddings": 1108.3652,
      "encoder_q-layer.0": 817.4711,
      "encoder_q-layer.1": 815.2624,
      "encoder_q-layer.10": 176.7507,
      "encoder_q-layer.11": 458.7482,
      "encoder_q-layer.2": 925.5131,
      "encoder_q-layer.3": 934.9206,
      "encoder_q-layer.4": 880.9213,
      "encoder_q-layer.5": 829.9568,
      "encoder_q-layer.6": 808.3671,
      "encoder_q-layer.7": 690.8621,
      "encoder_q-layer.8": 722.7192,
      "encoder_q-layer.9": 392.4698,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2635,
      "inbatch_pos_score": 0.8721,
      "learning_rate": 4.755555555555556e-05,
      "loss": 3.7477,
      "norm_diff": 0.0947,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1179.7032,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.262,
      "query_norm": 1.5145,
      "queue_k_norm": 1.4226,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8742,
      "sent_len_1": 66.7042,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2475,
      "stdk": 0.047,
      "stdq": 0.0443,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 14400
    },
    {
      "accuracy": 47.6562,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7719,
      "doc_norm": 1.415,
      "encoder_q-embeddings": 1131.4501,
      "encoder_q-layer.0": 827.1096,
      "encoder_q-layer.1": 883.7488,
      "encoder_q-layer.10": 236.7186,
      "encoder_q-layer.11": 555.2093,
      "encoder_q-layer.2": 1012.5291,
      "encoder_q-layer.3": 1152.9346,
      "encoder_q-layer.4": 1318.772,
      "encoder_q-layer.5": 1390.8135,
      "encoder_q-layer.6": 1136.7885,
      "encoder_q-layer.7": 818.8971,
      "encoder_q-layer.8": 475.4359,
      "encoder_q-layer.9": 254.5848,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2687,
      "inbatch_pos_score": 0.8296,
      "learning_rate": 4.75e-05,
      "loss": 3.7719,
      "norm_diff": 0.1031,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1386.6726,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2656,
      "query_norm": 1.5181,
      "queue_k_norm": 1.4232,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9217,
      "sent_len_1": 66.7756,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4875,
      "stdk": 0.0468,
      "stdq": 0.044,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 14500
    },
    {
      "accuracy": 48.8281,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7644,
      "doc_norm": 1.4217,
      "encoder_q-embeddings": 1059.2546,
      "encoder_q-layer.0": 769.2678,
      "encoder_q-layer.1": 867.0495,
      "encoder_q-layer.10": 205.9626,
      "encoder_q-layer.11": 473.9135,
      "encoder_q-layer.2": 971.1918,
      "encoder_q-layer.3": 991.6965,
      "encoder_q-layer.4": 901.6828,
      "encoder_q-layer.5": 933.2266,
      "encoder_q-layer.6": 803.4598,
      "encoder_q-layer.7": 564.1382,
      "encoder_q-layer.8": 310.6254,
      "encoder_q-layer.9": 188.0536,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2676,
      "inbatch_pos_score": 0.8599,
      "learning_rate": 4.7444444444444445e-05,
      "loss": 3.7644,
      "norm_diff": 0.1157,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1142.2735,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2642,
      "query_norm": 1.5374,
      "queue_k_norm": 1.4186,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9403,
      "sent_len_1": 66.7049,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2175,
      "stdk": 0.0471,
      "stdq": 0.0454,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 14600
    },
    {
      "accuracy": 45.8984,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7581,
      "doc_norm": 1.4146,
      "encoder_q-embeddings": 1587.9374,
      "encoder_q-layer.0": 1263.6409,
      "encoder_q-layer.1": 1207.2095,
      "encoder_q-layer.10": 250.047,
      "encoder_q-layer.11": 474.3844,
      "encoder_q-layer.2": 1465.3113,
      "encoder_q-layer.3": 1567.7484,
      "encoder_q-layer.4": 1477.5537,
      "encoder_q-layer.5": 1462.697,
      "encoder_q-layer.6": 1572.3326,
      "encoder_q-layer.7": 1710.2887,
      "encoder_q-layer.8": 1715.2892,
      "encoder_q-layer.9": 1002.7847,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2614,
      "inbatch_pos_score": 0.8506,
      "learning_rate": 4.7388888888888894e-05,
      "loss": 3.7581,
      "norm_diff": 0.1213,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2058.9713,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2598,
      "query_norm": 1.5358,
      "queue_k_norm": 1.4181,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0879,
      "sent_len_1": 66.7892,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.97,
      "stdk": 0.0469,
      "stdq": 0.0458,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 14700
    },
    {
      "accuracy": 50.9766,
      "active_queue_size": 16384.0,
      "cl_loss": 3.784,
      "doc_norm": 1.4183,
      "encoder_q-embeddings": 212.5462,
      "encoder_q-layer.0": 139.8297,
      "encoder_q-layer.1": 153.1564,
      "encoder_q-layer.10": 183.8877,
      "encoder_q-layer.11": 407.072,
      "encoder_q-layer.2": 173.479,
      "encoder_q-layer.3": 171.0556,
      "encoder_q-layer.4": 180.8121,
      "encoder_q-layer.5": 182.1551,
      "encoder_q-layer.6": 187.7803,
      "encoder_q-layer.7": 200.1177,
      "encoder_q-layer.8": 222.0048,
      "encoder_q-layer.9": 178.9089,
      "epoch": 0.14,
      "inbatch_neg_score": 0.2548,
      "inbatch_pos_score": 0.8696,
      "learning_rate": 4.7333333333333336e-05,
      "loss": 3.784,
      "norm_diff": 0.0925,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 312.1071,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2532,
      "query_norm": 1.5108,
      "queue_k_norm": 1.418,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.5792,
      "sent_len_1": 66.7561,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0575,
      "stdk": 0.0471,
      "stdq": 0.0448,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 14800
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7415,
      "doc_norm": 1.4188,
      "encoder_q-embeddings": 358.8188,
      "encoder_q-layer.0": 247.8521,
      "encoder_q-layer.1": 302.4949,
      "encoder_q-layer.10": 193.947,
      "encoder_q-layer.11": 453.4926,
      "encoder_q-layer.2": 342.6641,
      "encoder_q-layer.3": 378.2597,
      "encoder_q-layer.4": 444.5037,
      "encoder_q-layer.5": 458.0376,
      "encoder_q-layer.6": 562.1864,
      "encoder_q-layer.7": 511.6816,
      "encoder_q-layer.8": 283.0871,
      "encoder_q-layer.9": 190.5224,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2645,
      "inbatch_pos_score": 0.8608,
      "learning_rate": 4.727777777777778e-05,
      "loss": 3.7415,
      "norm_diff": 0.106,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 555.2857,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2644,
      "query_norm": 1.5249,
      "queue_k_norm": 1.4184,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0314,
      "sent_len_1": 67.0511,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5838,
      "stdk": 0.0472,
      "stdq": 0.0445,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 14900
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7283,
      "doc_norm": 1.4155,
      "encoder_q-embeddings": 1741.8225,
      "encoder_q-layer.0": 1236.5345,
      "encoder_q-layer.1": 1311.1008,
      "encoder_q-layer.10": 212.9071,
      "encoder_q-layer.11": 490.235,
      "encoder_q-layer.2": 1496.5485,
      "encoder_q-layer.3": 1504.3628,
      "encoder_q-layer.4": 1467.3281,
      "encoder_q-layer.5": 1293.5248,
      "encoder_q-layer.6": 1198.1888,
      "encoder_q-layer.7": 1080.5836,
      "encoder_q-layer.8": 976.5828,
      "encoder_q-layer.9": 636.894,
      "epoch": 0.15,
      "inbatch_neg_score": 0.266,
      "inbatch_pos_score": 0.8477,
      "learning_rate": 4.722222222222222e-05,
      "loss": 3.7283,
      "norm_diff": 0.0973,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1824.6911,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2654,
      "query_norm": 1.5127,
      "queue_k_norm": 1.4151,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9089,
      "sent_len_1": 66.7752,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4263,
      "stdk": 0.0471,
      "stdq": 0.0439,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15000
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7332,
      "doc_norm": 1.4147,
      "encoder_q-embeddings": 533.086,
      "encoder_q-layer.0": 416.5028,
      "encoder_q-layer.1": 458.2967,
      "encoder_q-layer.10": 180.9997,
      "encoder_q-layer.11": 390.2127,
      "encoder_q-layer.2": 495.113,
      "encoder_q-layer.3": 470.0929,
      "encoder_q-layer.4": 486.375,
      "encoder_q-layer.5": 519.9919,
      "encoder_q-layer.6": 523.8776,
      "encoder_q-layer.7": 489.1165,
      "encoder_q-layer.8": 297.7428,
      "encoder_q-layer.9": 181.5132,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2623,
      "inbatch_pos_score": 0.8945,
      "learning_rate": 4.716666666666667e-05,
      "loss": 3.7332,
      "norm_diff": 0.1143,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 651.2777,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2607,
      "query_norm": 1.529,
      "queue_k_norm": 1.4136,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0549,
      "sent_len_1": 66.9405,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.255,
      "stdk": 0.0471,
      "stdq": 0.0457,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15100
    },
    {
      "accuracy": 49.707,
      "active_queue_size": 16384.0,
      "cl_loss": 3.735,
      "doc_norm": 1.4135,
      "encoder_q-embeddings": 3169.7559,
      "encoder_q-layer.0": 2342.1697,
      "encoder_q-layer.1": 2507.304,
      "encoder_q-layer.10": 411.5363,
      "encoder_q-layer.11": 932.9662,
      "encoder_q-layer.2": 2915.4971,
      "encoder_q-layer.3": 3091.9917,
      "encoder_q-layer.4": 3362.2898,
      "encoder_q-layer.5": 3245.4194,
      "encoder_q-layer.6": 2635.144,
      "encoder_q-layer.7": 2008.2775,
      "encoder_q-layer.8": 1164.9126,
      "encoder_q-layer.9": 527.6869,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2383,
      "inbatch_pos_score": 0.8418,
      "learning_rate": 4.711111111111111e-05,
      "loss": 3.735,
      "norm_diff": 0.1085,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3639.1394,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2388,
      "query_norm": 1.522,
      "queue_k_norm": 1.4145,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.834,
      "sent_len_1": 66.7538,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0825,
      "stdk": 0.0471,
      "stdq": 0.0455,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 15200
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7476,
      "doc_norm": 1.4103,
      "encoder_q-embeddings": 10891.6543,
      "encoder_q-layer.0": 7714.4053,
      "encoder_q-layer.1": 7944.7563,
      "encoder_q-layer.10": 355.1263,
      "encoder_q-layer.11": 913.6475,
      "encoder_q-layer.2": 9287.6875,
      "encoder_q-layer.3": 10692.2109,
      "encoder_q-layer.4": 10407.0566,
      "encoder_q-layer.5": 9735.8125,
      "encoder_q-layer.6": 7649.7695,
      "encoder_q-layer.7": 4857.9585,
      "encoder_q-layer.8": 1943.9766,
      "encoder_q-layer.9": 499.6726,
      "epoch": 0.15,
      "inbatch_neg_score": 0.27,
      "inbatch_pos_score": 0.8667,
      "learning_rate": 4.7055555555555555e-05,
      "loss": 3.7476,
      "norm_diff": 0.099,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 11471.9891,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.269,
      "query_norm": 1.5093,
      "queue_k_norm": 1.4114,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.167,
      "sent_len_1": 66.7633,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6025,
      "stdk": 0.047,
      "stdq": 0.0447,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15300
    },
    {
      "accuracy": 48.7305,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7253,
      "doc_norm": 1.4123,
      "encoder_q-embeddings": 2200.8816,
      "encoder_q-layer.0": 1562.4109,
      "encoder_q-layer.1": 1727.1785,
      "encoder_q-layer.10": 388.364,
      "encoder_q-layer.11": 861.1769,
      "encoder_q-layer.2": 2250.5054,
      "encoder_q-layer.3": 2424.1499,
      "encoder_q-layer.4": 2412.5911,
      "encoder_q-layer.5": 2101.5083,
      "encoder_q-layer.6": 2077.0281,
      "encoder_q-layer.7": 1956.6439,
      "encoder_q-layer.8": 1129.1847,
      "encoder_q-layer.9": 459.9446,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2279,
      "inbatch_pos_score": 0.8188,
      "learning_rate": 4.7e-05,
      "loss": 3.7253,
      "norm_diff": 0.0759,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2669.7441,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2271,
      "query_norm": 1.4883,
      "queue_k_norm": 1.4125,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0852,
      "sent_len_1": 66.7339,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0462,
      "stdk": 0.0471,
      "stdq": 0.0451,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 15400
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7093,
      "doc_norm": 1.4176,
      "encoder_q-embeddings": 2254.3416,
      "encoder_q-layer.0": 1579.5554,
      "encoder_q-layer.1": 2081.0908,
      "encoder_q-layer.10": 416.1562,
      "encoder_q-layer.11": 848.3127,
      "encoder_q-layer.2": 2480.3091,
      "encoder_q-layer.3": 2741.7891,
      "encoder_q-layer.4": 2943.6089,
      "encoder_q-layer.5": 2622.7346,
      "encoder_q-layer.6": 2055.252,
      "encoder_q-layer.7": 1248.4773,
      "encoder_q-layer.8": 512.4978,
      "encoder_q-layer.9": 341.7864,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2304,
      "inbatch_pos_score": 0.8481,
      "learning_rate": 4.6944444444444446e-05,
      "loss": 3.7093,
      "norm_diff": 0.0643,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2832.4658,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2286,
      "query_norm": 1.4819,
      "queue_k_norm": 1.4101,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8778,
      "sent_len_1": 66.7929,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5625,
      "stdk": 0.0473,
      "stdq": 0.0446,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15500
    },
    {
      "accuracy": 47.2656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7382,
      "doc_norm": 1.4055,
      "encoder_q-embeddings": 912.9368,
      "encoder_q-layer.0": 678.1561,
      "encoder_q-layer.1": 745.2568,
      "encoder_q-layer.10": 376.242,
      "encoder_q-layer.11": 928.5197,
      "encoder_q-layer.2": 847.4238,
      "encoder_q-layer.3": 875.5267,
      "encoder_q-layer.4": 819.2725,
      "encoder_q-layer.5": 846.9387,
      "encoder_q-layer.6": 775.3784,
      "encoder_q-layer.7": 669.6934,
      "encoder_q-layer.8": 451.0255,
      "encoder_q-layer.9": 356.7258,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2417,
      "inbatch_pos_score": 0.8091,
      "learning_rate": 4.6888888888888895e-05,
      "loss": 3.7382,
      "norm_diff": 0.041,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1104.0046,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2394,
      "query_norm": 1.4465,
      "queue_k_norm": 1.4089,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2431,
      "sent_len_1": 66.8668,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0137,
      "stdk": 0.0469,
      "stdq": 0.0437,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15600
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7558,
      "doc_norm": 1.4024,
      "encoder_q-embeddings": 1048.4872,
      "encoder_q-layer.0": 714.452,
      "encoder_q-layer.1": 709.6117,
      "encoder_q-layer.10": 457.4622,
      "encoder_q-layer.11": 970.8304,
      "encoder_q-layer.2": 738.3916,
      "encoder_q-layer.3": 731.2146,
      "encoder_q-layer.4": 739.3929,
      "encoder_q-layer.5": 792.5068,
      "encoder_q-layer.6": 760.2808,
      "encoder_q-layer.7": 748.2123,
      "encoder_q-layer.8": 673.7017,
      "encoder_q-layer.9": 455.4038,
      "epoch": 0.15,
      "inbatch_neg_score": 0.2412,
      "inbatch_pos_score": 0.9175,
      "learning_rate": 4.683333333333334e-05,
      "loss": 3.7558,
      "norm_diff": 0.1402,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1126.2983,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2382,
      "query_norm": 1.5426,
      "queue_k_norm": 1.4078,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.082,
      "sent_len_1": 66.8991,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.13,
      "stdk": 0.0468,
      "stdq": 0.0478,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15700
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7408,
      "doc_norm": 1.4088,
      "encoder_q-embeddings": 2626.095,
      "encoder_q-layer.0": 1833.373,
      "encoder_q-layer.1": 1935.2697,
      "encoder_q-layer.10": 339.0569,
      "encoder_q-layer.11": 773.4079,
      "encoder_q-layer.2": 2663.7603,
      "encoder_q-layer.3": 2213.3801,
      "encoder_q-layer.4": 1721.7786,
      "encoder_q-layer.5": 1557.9772,
      "encoder_q-layer.6": 1431.899,
      "encoder_q-layer.7": 1333.0714,
      "encoder_q-layer.8": 911.9082,
      "encoder_q-layer.9": 380.6698,
      "epoch": 0.15,
      "inbatch_neg_score": 0.235,
      "inbatch_pos_score": 0.832,
      "learning_rate": 4.677777777777778e-05,
      "loss": 3.7408,
      "norm_diff": 0.0685,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2582.0204,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2343,
      "query_norm": 1.4773,
      "queue_k_norm": 1.4067,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.988,
      "sent_len_1": 66.5382,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7688,
      "stdk": 0.0471,
      "stdq": 0.0451,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 15800
    },
    {
      "accuracy": 50.4883,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7039,
      "doc_norm": 1.4063,
      "encoder_q-embeddings": 1252.1451,
      "encoder_q-layer.0": 939.847,
      "encoder_q-layer.1": 1008.0081,
      "encoder_q-layer.10": 371.2855,
      "encoder_q-layer.11": 863.2305,
      "encoder_q-layer.2": 1048.7152,
      "encoder_q-layer.3": 1110.6632,
      "encoder_q-layer.4": 1054.385,
      "encoder_q-layer.5": 1034.4124,
      "encoder_q-layer.6": 869.2734,
      "encoder_q-layer.7": 730.7847,
      "encoder_q-layer.8": 441.3317,
      "encoder_q-layer.9": 333.2159,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2404,
      "inbatch_pos_score": 0.8677,
      "learning_rate": 4.672222222222222e-05,
      "loss": 3.7039,
      "norm_diff": 0.0688,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1364.4632,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2402,
      "query_norm": 1.4751,
      "queue_k_norm": 1.4073,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0708,
      "sent_len_1": 66.9053,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.0488,
      "stdk": 0.0471,
      "stdq": 0.0458,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 15900
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7178,
      "doc_norm": 1.4135,
      "encoder_q-embeddings": 28072.6504,
      "encoder_q-layer.0": 20548.6719,
      "encoder_q-layer.1": 18052.1699,
      "encoder_q-layer.10": 241.4423,
      "encoder_q-layer.11": 453.7159,
      "encoder_q-layer.2": 22249.3223,
      "encoder_q-layer.3": 20311.4668,
      "encoder_q-layer.4": 17888.834,
      "encoder_q-layer.5": 16491.0742,
      "encoder_q-layer.6": 16818.793,
      "encoder_q-layer.7": 10798.9414,
      "encoder_q-layer.8": 5461.9761,
      "encoder_q-layer.9": 936.6177,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2608,
      "inbatch_pos_score": 0.8848,
      "learning_rate": 4.666666666666667e-05,
      "loss": 3.7178,
      "norm_diff": 0.0596,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 25686.5623,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.2607,
      "query_norm": 1.4731,
      "queue_k_norm": 1.4067,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9276,
      "sent_len_1": 66.9262,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6763,
      "stdk": 0.0474,
      "stdq": 0.0451,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 16000
    },
    {
      "accuracy": 48.1445,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7148,
      "doc_norm": 1.413,
      "encoder_q-embeddings": 1182.7577,
      "encoder_q-layer.0": 827.0007,
      "encoder_q-layer.1": 902.8774,
      "encoder_q-layer.10": 212.3741,
      "encoder_q-layer.11": 468.0152,
      "encoder_q-layer.2": 993.398,
      "encoder_q-layer.3": 1034.8196,
      "encoder_q-layer.4": 1120.3062,
      "encoder_q-layer.5": 1138.1552,
      "encoder_q-layer.6": 1368.4474,
      "encoder_q-layer.7": 1787.4846,
      "encoder_q-layer.8": 2005.4099,
      "encoder_q-layer.9": 950.6409,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2665,
      "inbatch_pos_score": 0.8657,
      "learning_rate": 4.6611111111111114e-05,
      "loss": 3.7148,
      "norm_diff": 0.0761,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1793.4853,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2651,
      "query_norm": 1.4891,
      "queue_k_norm": 1.4063,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0663,
      "sent_len_1": 66.8392,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5312,
      "stdk": 0.0473,
      "stdq": 0.0459,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 16100
    },
    {
      "accuracy": 48.5352,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7036,
      "doc_norm": 1.4067,
      "encoder_q-embeddings": 701.0217,
      "encoder_q-layer.0": 475.1675,
      "encoder_q-layer.1": 538.0632,
      "encoder_q-layer.10": 201.0547,
      "encoder_q-layer.11": 490.5031,
      "encoder_q-layer.2": 613.6672,
      "encoder_q-layer.3": 648.1406,
      "encoder_q-layer.4": 702.6463,
      "encoder_q-layer.5": 584.5582,
      "encoder_q-layer.6": 504.18,
      "encoder_q-layer.7": 305.4601,
      "encoder_q-layer.8": 231.5768,
      "encoder_q-layer.9": 186.2904,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2641,
      "inbatch_pos_score": 0.8535,
      "learning_rate": 4.6555555555555556e-05,
      "loss": 3.7036,
      "norm_diff": 0.0497,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 797.4695,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2637,
      "query_norm": 1.4564,
      "queue_k_norm": 1.4051,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0441,
      "sent_len_1": 66.8308,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1962,
      "stdk": 0.0471,
      "stdq": 0.044,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 16200
    },
    {
      "accuracy": 48.2422,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7043,
      "doc_norm": 1.4074,
      "encoder_q-embeddings": 841.4246,
      "encoder_q-layer.0": 624.8605,
      "encoder_q-layer.1": 643.5723,
      "encoder_q-layer.10": 191.2644,
      "encoder_q-layer.11": 438.1519,
      "encoder_q-layer.2": 758.8859,
      "encoder_q-layer.3": 785.2599,
      "encoder_q-layer.4": 899.8054,
      "encoder_q-layer.5": 806.9026,
      "encoder_q-layer.6": 715.1705,
      "encoder_q-layer.7": 586.3497,
      "encoder_q-layer.8": 306.8842,
      "encoder_q-layer.9": 191.9863,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2563,
      "inbatch_pos_score": 0.8491,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 3.7043,
      "norm_diff": 0.0599,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 960.9096,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2534,
      "query_norm": 1.4673,
      "queue_k_norm": 1.4032,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9168,
      "sent_len_1": 66.7472,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2612,
      "stdk": 0.0471,
      "stdq": 0.045,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 16300
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6831,
      "doc_norm": 1.402,
      "encoder_q-embeddings": 2075.948,
      "encoder_q-layer.0": 1503.925,
      "encoder_q-layer.1": 1534.533,
      "encoder_q-layer.10": 238.4414,
      "encoder_q-layer.11": 520.8948,
      "encoder_q-layer.2": 1842.5541,
      "encoder_q-layer.3": 1896.3843,
      "encoder_q-layer.4": 1931.1184,
      "encoder_q-layer.5": 2091.8738,
      "encoder_q-layer.6": 2217.9817,
      "encoder_q-layer.7": 1444.4355,
      "encoder_q-layer.8": 411.2575,
      "encoder_q-layer.9": 218.521,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2614,
      "inbatch_pos_score": 0.8589,
      "learning_rate": 4.644444444444445e-05,
      "loss": 3.6831,
      "norm_diff": 0.0495,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2387.6488,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2603,
      "query_norm": 1.4515,
      "queue_k_norm": 1.404,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2199,
      "sent_len_1": 66.9548,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4062,
      "stdk": 0.047,
      "stdq": 0.0437,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 16400
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.679,
      "doc_norm": 1.4014,
      "encoder_q-embeddings": 582.403,
      "encoder_q-layer.0": 422.6291,
      "encoder_q-layer.1": 488.9368,
      "encoder_q-layer.10": 208.8659,
      "encoder_q-layer.11": 439.2017,
      "encoder_q-layer.2": 603.7133,
      "encoder_q-layer.3": 657.0931,
      "encoder_q-layer.4": 560.5911,
      "encoder_q-layer.5": 489.5139,
      "encoder_q-layer.6": 493.6404,
      "encoder_q-layer.7": 410.53,
      "encoder_q-layer.8": 246.5434,
      "encoder_q-layer.9": 186.7983,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2554,
      "inbatch_pos_score": 0.8633,
      "learning_rate": 4.638888888888889e-05,
      "loss": 3.679,
      "norm_diff": 0.0873,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 704.0172,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2532,
      "query_norm": 1.4886,
      "queue_k_norm": 1.4034,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0985,
      "sent_len_1": 66.8092,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.1138,
      "stdk": 0.0469,
      "stdq": 0.0456,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 16500
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7149,
      "doc_norm": 1.4024,
      "encoder_q-embeddings": 320.9492,
      "encoder_q-layer.0": 237.4339,
      "encoder_q-layer.1": 265.9968,
      "encoder_q-layer.10": 188.2703,
      "encoder_q-layer.11": 422.2701,
      "encoder_q-layer.2": 276.2502,
      "encoder_q-layer.3": 280.3205,
      "encoder_q-layer.4": 279.9183,
      "encoder_q-layer.5": 274.705,
      "encoder_q-layer.6": 292.8484,
      "encoder_q-layer.7": 236.5598,
      "encoder_q-layer.8": 207.2865,
      "encoder_q-layer.9": 171.8089,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2435,
      "inbatch_pos_score": 0.8706,
      "learning_rate": 4.633333333333333e-05,
      "loss": 3.7149,
      "norm_diff": 0.0897,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 399.6498,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2433,
      "query_norm": 1.4921,
      "queue_k_norm": 1.4037,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7783,
      "sent_len_1": 66.9321,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8862,
      "stdk": 0.047,
      "stdq": 0.0461,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 16600
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6983,
      "doc_norm": 1.3997,
      "encoder_q-embeddings": 744.143,
      "encoder_q-layer.0": 504.9815,
      "encoder_q-layer.1": 594.9958,
      "encoder_q-layer.10": 215.2236,
      "encoder_q-layer.11": 452.3077,
      "encoder_q-layer.2": 700.6279,
      "encoder_q-layer.3": 791.2682,
      "encoder_q-layer.4": 919.1357,
      "encoder_q-layer.5": 870.3819,
      "encoder_q-layer.6": 956.6893,
      "encoder_q-layer.7": 544.0125,
      "encoder_q-layer.8": 246.8699,
      "encoder_q-layer.9": 183.5899,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2615,
      "inbatch_pos_score": 0.8892,
      "learning_rate": 4.627777777777778e-05,
      "loss": 3.6983,
      "norm_diff": 0.1141,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 959.787,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2593,
      "query_norm": 1.5138,
      "queue_k_norm": 1.4039,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9366,
      "sent_len_1": 66.9835,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2837,
      "stdk": 0.0469,
      "stdq": 0.0462,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 16700
    },
    {
      "accuracy": 50.7812,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6969,
      "doc_norm": 1.4047,
      "encoder_q-embeddings": 329.9292,
      "encoder_q-layer.0": 223.3442,
      "encoder_q-layer.1": 244.1288,
      "encoder_q-layer.10": 182.2275,
      "encoder_q-layer.11": 413.9463,
      "encoder_q-layer.2": 275.2245,
      "encoder_q-layer.3": 279.997,
      "encoder_q-layer.4": 280.0677,
      "encoder_q-layer.5": 282.5807,
      "encoder_q-layer.6": 317.522,
      "encoder_q-layer.7": 336.8553,
      "encoder_q-layer.8": 359.5425,
      "encoder_q-layer.9": 230.6185,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2553,
      "inbatch_pos_score": 0.8848,
      "learning_rate": 4.6222222222222224e-05,
      "loss": 3.6969,
      "norm_diff": 0.0875,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 440.8046,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2534,
      "query_norm": 1.4922,
      "queue_k_norm": 1.4049,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9766,
      "sent_len_1": 66.7595,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9762,
      "stdk": 0.0471,
      "stdq": 0.046,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 16800
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6832,
      "doc_norm": 1.4027,
      "encoder_q-embeddings": 1294.7759,
      "encoder_q-layer.0": 930.2153,
      "encoder_q-layer.1": 1025.7069,
      "encoder_q-layer.10": 170.0993,
      "encoder_q-layer.11": 408.0142,
      "encoder_q-layer.2": 1196.2966,
      "encoder_q-layer.3": 1222.6422,
      "encoder_q-layer.4": 1281.0061,
      "encoder_q-layer.5": 1227.757,
      "encoder_q-layer.6": 1029.2234,
      "encoder_q-layer.7": 731.5608,
      "encoder_q-layer.8": 289.2668,
      "encoder_q-layer.9": 172.4314,
      "epoch": 0.16,
      "inbatch_neg_score": 0.2556,
      "inbatch_pos_score": 0.8711,
      "learning_rate": 4.6166666666666666e-05,
      "loss": 3.6832,
      "norm_diff": 0.0725,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1411.4619,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2542,
      "query_norm": 1.4751,
      "queue_k_norm": 1.4039,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0245,
      "sent_len_1": 66.6111,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.3725,
      "stdk": 0.047,
      "stdq": 0.0448,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 16900
    },
    {
      "accuracy": 48.8281,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6509,
      "doc_norm": 1.3949,
      "encoder_q-embeddings": 5786.2402,
      "encoder_q-layer.0": 4663.4385,
      "encoder_q-layer.1": 4832.209,
      "encoder_q-layer.10": 176.3794,
      "encoder_q-layer.11": 403.2226,
      "encoder_q-layer.2": 5899.3018,
      "encoder_q-layer.3": 6673.5156,
      "encoder_q-layer.4": 6754.8599,
      "encoder_q-layer.5": 7430.8662,
      "encoder_q-layer.6": 6523.9233,
      "encoder_q-layer.7": 4032.9512,
      "encoder_q-layer.8": 1683.1449,
      "encoder_q-layer.9": 401.7845,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2635,
      "inbatch_pos_score": 0.8481,
      "learning_rate": 4.6111111111111115e-05,
      "loss": 3.6509,
      "norm_diff": 0.0681,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7397.5867,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2617,
      "query_norm": 1.463,
      "queue_k_norm": 1.4053,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9539,
      "sent_len_1": 66.8168,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2575,
      "stdk": 0.0468,
      "stdq": 0.0446,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 17000
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6853,
      "doc_norm": 1.4035,
      "encoder_q-embeddings": 515.8788,
      "encoder_q-layer.0": 372.5059,
      "encoder_q-layer.1": 440.0763,
      "encoder_q-layer.10": 210.877,
      "encoder_q-layer.11": 465.8947,
      "encoder_q-layer.2": 539.7372,
      "encoder_q-layer.3": 528.1481,
      "encoder_q-layer.4": 589.9855,
      "encoder_q-layer.5": 526.9517,
      "encoder_q-layer.6": 468.4024,
      "encoder_q-layer.7": 377.8751,
      "encoder_q-layer.8": 288.2725,
      "encoder_q-layer.9": 199.2092,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2627,
      "inbatch_pos_score": 0.9204,
      "learning_rate": 4.605555555555556e-05,
      "loss": 3.6853,
      "norm_diff": 0.1178,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 654.1217,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.261,
      "query_norm": 1.5213,
      "queue_k_norm": 1.4045,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0057,
      "sent_len_1": 66.8655,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9437,
      "stdk": 0.0471,
      "stdq": 0.0471,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 17100
    },
    {
      "accuracy": 48.7305,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7034,
      "doc_norm": 1.4094,
      "encoder_q-embeddings": 1657.6027,
      "encoder_q-layer.0": 1157.9812,
      "encoder_q-layer.1": 1236.9668,
      "encoder_q-layer.10": 186.4518,
      "encoder_q-layer.11": 484.6173,
      "encoder_q-layer.2": 1402.939,
      "encoder_q-layer.3": 1532.244,
      "encoder_q-layer.4": 1647.7589,
      "encoder_q-layer.5": 1693.3566,
      "encoder_q-layer.6": 2025.5414,
      "encoder_q-layer.7": 2038.4303,
      "encoder_q-layer.8": 1210.8477,
      "encoder_q-layer.9": 294.932,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2641,
      "inbatch_pos_score": 0.8682,
      "learning_rate": 4.600000000000001e-05,
      "loss": 3.7034,
      "norm_diff": 0.0782,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2165.1222,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2646,
      "query_norm": 1.4876,
      "queue_k_norm": 1.4022,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9778,
      "sent_len_1": 66.9516,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9512,
      "stdk": 0.0473,
      "stdq": 0.0462,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 17200
    },
    {
      "accuracy": 50.7812,
      "active_queue_size": 16384.0,
      "cl_loss": 3.752,
      "doc_norm": 1.405,
      "encoder_q-embeddings": 637.668,
      "encoder_q-layer.0": 453.5364,
      "encoder_q-layer.1": 501.9983,
      "encoder_q-layer.10": 197.6379,
      "encoder_q-layer.11": 495.2677,
      "encoder_q-layer.2": 567.5832,
      "encoder_q-layer.3": 584.4075,
      "encoder_q-layer.4": 661.741,
      "encoder_q-layer.5": 584.7655,
      "encoder_q-layer.6": 616.3918,
      "encoder_q-layer.7": 578.8821,
      "encoder_q-layer.8": 400.3969,
      "encoder_q-layer.9": 193.3036,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2869,
      "inbatch_pos_score": 0.8867,
      "learning_rate": 4.594444444444444e-05,
      "loss": 3.752,
      "norm_diff": 0.065,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 793.9456,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2832,
      "query_norm": 1.4699,
      "queue_k_norm": 1.4042,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8245,
      "sent_len_1": 66.7684,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.63,
      "stdk": 0.047,
      "stdq": 0.045,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 17300
    },
    {
      "accuracy": 50.0977,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7117,
      "doc_norm": 1.4052,
      "encoder_q-embeddings": 1126.4989,
      "encoder_q-layer.0": 753.838,
      "encoder_q-layer.1": 811.7047,
      "encoder_q-layer.10": 214.9194,
      "encoder_q-layer.11": 499.5439,
      "encoder_q-layer.2": 969.0172,
      "encoder_q-layer.3": 992.5029,
      "encoder_q-layer.4": 1000.4146,
      "encoder_q-layer.5": 985.7787,
      "encoder_q-layer.6": 1015.6859,
      "encoder_q-layer.7": 827.111,
      "encoder_q-layer.8": 469.3026,
      "encoder_q-layer.9": 239.1533,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2834,
      "inbatch_pos_score": 0.8818,
      "learning_rate": 4.588888888888889e-05,
      "loss": 3.7117,
      "norm_diff": 0.0649,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1235.039,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.283,
      "query_norm": 1.4702,
      "queue_k_norm": 1.4065,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8101,
      "sent_len_1": 66.6826,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5737,
      "stdk": 0.047,
      "stdq": 0.0446,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 17400
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7036,
      "doc_norm": 1.4022,
      "encoder_q-embeddings": 5209.1621,
      "encoder_q-layer.0": 3171.3943,
      "encoder_q-layer.1": 3032.1838,
      "encoder_q-layer.10": 164.1274,
      "encoder_q-layer.11": 390.6615,
      "encoder_q-layer.2": 2679.125,
      "encoder_q-layer.3": 2691.0713,
      "encoder_q-layer.4": 2803.6394,
      "encoder_q-layer.5": 2158.6279,
      "encoder_q-layer.6": 1745.0059,
      "encoder_q-layer.7": 1623.4635,
      "encoder_q-layer.8": 875.9292,
      "encoder_q-layer.9": 265.5596,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2937,
      "inbatch_pos_score": 0.916,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 3.7036,
      "norm_diff": 0.0916,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4086.8895,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.293,
      "query_norm": 1.4938,
      "queue_k_norm": 1.4087,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.109,
      "sent_len_1": 66.9925,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8288,
      "stdk": 0.0469,
      "stdq": 0.0451,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 17500
    },
    {
      "accuracy": 47.9492,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6947,
      "doc_norm": 1.4005,
      "encoder_q-embeddings": 3387.6172,
      "encoder_q-layer.0": 2316.6272,
      "encoder_q-layer.1": 2653.1011,
      "encoder_q-layer.10": 198.856,
      "encoder_q-layer.11": 465.8602,
      "encoder_q-layer.2": 2715.9873,
      "encoder_q-layer.3": 3163.7808,
      "encoder_q-layer.4": 3028.0928,
      "encoder_q-layer.5": 2317.7349,
      "encoder_q-layer.6": 2260.6221,
      "encoder_q-layer.7": 2149.5535,
      "encoder_q-layer.8": 1303.3611,
      "encoder_q-layer.9": 319.3143,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2955,
      "inbatch_pos_score": 0.8564,
      "learning_rate": 4.577777777777778e-05,
      "loss": 3.6947,
      "norm_diff": 0.0464,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3517.4546,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2937,
      "query_norm": 1.4469,
      "queue_k_norm": 1.4085,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1006,
      "sent_len_1": 66.6582,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2063,
      "stdk": 0.0468,
      "stdq": 0.0436,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 17600
    },
    {
      "accuracy": 50.4883,
      "active_queue_size": 16384.0,
      "cl_loss": 3.697,
      "doc_norm": 1.4085,
      "encoder_q-embeddings": 1357.1682,
      "encoder_q-layer.0": 969.774,
      "encoder_q-layer.1": 1099.7329,
      "encoder_q-layer.10": 198.0203,
      "encoder_q-layer.11": 438.6694,
      "encoder_q-layer.2": 1081.1906,
      "encoder_q-layer.3": 1139.3794,
      "encoder_q-layer.4": 1117.7994,
      "encoder_q-layer.5": 924.501,
      "encoder_q-layer.6": 900.6042,
      "encoder_q-layer.7": 780.9453,
      "encoder_q-layer.8": 433.614,
      "encoder_q-layer.9": 220.0501,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2955,
      "inbatch_pos_score": 0.8926,
      "learning_rate": 4.572222222222222e-05,
      "loss": 3.697,
      "norm_diff": 0.0599,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1374.4682,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2954,
      "query_norm": 1.4684,
      "queue_k_norm": 1.4098,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8575,
      "sent_len_1": 67.0099,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8775,
      "stdk": 0.0471,
      "stdq": 0.0441,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 17700
    },
    {
      "accuracy": 50.9766,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7108,
      "doc_norm": 1.4144,
      "encoder_q-embeddings": 1674.6956,
      "encoder_q-layer.0": 1015.9118,
      "encoder_q-layer.1": 1043.0096,
      "encoder_q-layer.10": 192.2271,
      "encoder_q-layer.11": 453.877,
      "encoder_q-layer.2": 1032.1792,
      "encoder_q-layer.3": 1141.4756,
      "encoder_q-layer.4": 1174.6505,
      "encoder_q-layer.5": 1158.7028,
      "encoder_q-layer.6": 1137.4181,
      "encoder_q-layer.7": 1058.8611,
      "encoder_q-layer.8": 630.6962,
      "encoder_q-layer.9": 216.6463,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2987,
      "inbatch_pos_score": 0.9248,
      "learning_rate": 4.566666666666667e-05,
      "loss": 3.7108,
      "norm_diff": 0.0922,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1581.0842,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2971,
      "query_norm": 1.5066,
      "queue_k_norm": 1.4085,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0613,
      "sent_len_1": 66.8495,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9238,
      "stdk": 0.0473,
      "stdq": 0.0463,
      "stdqueue_k": 0.0471,
      "stdqueue_q": 0.0,
      "step": 17800
    },
    {
      "accuracy": 46.9727,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7068,
      "doc_norm": 1.4138,
      "encoder_q-embeddings": 851.5773,
      "encoder_q-layer.0": 631.8976,
      "encoder_q-layer.1": 663.7853,
      "encoder_q-layer.10": 203.5052,
      "encoder_q-layer.11": 491.9795,
      "encoder_q-layer.2": 679.9135,
      "encoder_q-layer.3": 715.6171,
      "encoder_q-layer.4": 693.3563,
      "encoder_q-layer.5": 699.3101,
      "encoder_q-layer.6": 723.6272,
      "encoder_q-layer.7": 605.5711,
      "encoder_q-layer.8": 342.183,
      "encoder_q-layer.9": 191.9534,
      "epoch": 0.17,
      "inbatch_neg_score": 0.2918,
      "inbatch_pos_score": 0.8789,
      "learning_rate": 4.561111111111112e-05,
      "loss": 3.7068,
      "norm_diff": 0.0478,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 928.5365,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2908,
      "query_norm": 1.4616,
      "queue_k_norm": 1.4102,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9925,
      "sent_len_1": 66.7714,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7925,
      "stdk": 0.0472,
      "stdq": 0.0445,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 17900
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7132,
      "doc_norm": 1.4115,
      "encoder_q-embeddings": 9034.6309,
      "encoder_q-layer.0": 6595.9736,
      "encoder_q-layer.1": 6569.293,
      "encoder_q-layer.10": 473.9172,
      "encoder_q-layer.11": 1081.7544,
      "encoder_q-layer.2": 6733.7554,
      "encoder_q-layer.3": 6520.188,
      "encoder_q-layer.4": 6005.0889,
      "encoder_q-layer.5": 6219.5884,
      "encoder_q-layer.6": 5745.5698,
      "encoder_q-layer.7": 4334.1035,
      "encoder_q-layer.8": 1921.8358,
      "encoder_q-layer.9": 555.2422,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2812,
      "inbatch_pos_score": 0.8823,
      "learning_rate": 4.555555555555556e-05,
      "loss": 3.7132,
      "norm_diff": 0.0491,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8680.7333,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2817,
      "query_norm": 1.4606,
      "queue_k_norm": 1.4129,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7364,
      "sent_len_1": 66.6119,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.81,
      "stdk": 0.0471,
      "stdq": 0.0454,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 18000
    },
    {
      "accuracy": 48.0469,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7152,
      "doc_norm": 1.4127,
      "encoder_q-embeddings": 1066.0984,
      "encoder_q-layer.0": 783.1487,
      "encoder_q-layer.1": 882.4213,
      "encoder_q-layer.10": 393.5575,
      "encoder_q-layer.11": 1004.3089,
      "encoder_q-layer.2": 936.4866,
      "encoder_q-layer.3": 966.8971,
      "encoder_q-layer.4": 927.2573,
      "encoder_q-layer.5": 849.6568,
      "encoder_q-layer.6": 858.0427,
      "encoder_q-layer.7": 858.9401,
      "encoder_q-layer.8": 623.6135,
      "encoder_q-layer.9": 395.9897,
      "epoch": 0.18,
      "inbatch_neg_score": 0.298,
      "inbatch_pos_score": 0.8926,
      "learning_rate": 4.55e-05,
      "loss": 3.7152,
      "norm_diff": 0.0328,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1279.0283,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2981,
      "query_norm": 1.4454,
      "queue_k_norm": 1.4131,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0911,
      "sent_len_1": 66.5922,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4462,
      "stdk": 0.0471,
      "stdq": 0.0447,
      "stdqueue_k": 0.0472,
      "stdqueue_q": 0.0,
      "step": 18100
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7171,
      "doc_norm": 1.4175,
      "encoder_q-embeddings": 7370.5942,
      "encoder_q-layer.0": 5560.5107,
      "encoder_q-layer.1": 6291.6392,
      "encoder_q-layer.10": 411.3118,
      "encoder_q-layer.11": 951.916,
      "encoder_q-layer.2": 6831.312,
      "encoder_q-layer.3": 7607.2876,
      "encoder_q-layer.4": 8384.6963,
      "encoder_q-layer.5": 9515.5469,
      "encoder_q-layer.6": 7224.2617,
      "encoder_q-layer.7": 6729.625,
      "encoder_q-layer.8": 2577.2354,
      "encoder_q-layer.9": 617.1458,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2889,
      "inbatch_pos_score": 0.896,
      "learning_rate": 4.5444444444444444e-05,
      "loss": 3.7171,
      "norm_diff": 0.0562,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9314.2602,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2888,
      "query_norm": 1.4737,
      "queue_k_norm": 1.4169,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8967,
      "sent_len_1": 66.613,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0725,
      "stdk": 0.0473,
      "stdq": 0.0461,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 18200
    },
    {
      "accuracy": 48.6328,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6935,
      "doc_norm": 1.4166,
      "encoder_q-embeddings": 1582.0549,
      "encoder_q-layer.0": 1121.9978,
      "encoder_q-layer.1": 1213.204,
      "encoder_q-layer.10": 371.2173,
      "encoder_q-layer.11": 935.5358,
      "encoder_q-layer.2": 1388.0028,
      "encoder_q-layer.3": 1480.9048,
      "encoder_q-layer.4": 1492.691,
      "encoder_q-layer.5": 1296.7861,
      "encoder_q-layer.6": 1408.8463,
      "encoder_q-layer.7": 1195.0363,
      "encoder_q-layer.8": 800.1848,
      "encoder_q-layer.9": 383.4941,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2828,
      "inbatch_pos_score": 0.8662,
      "learning_rate": 4.538888888888889e-05,
      "loss": 3.6935,
      "norm_diff": 0.0108,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1828.6239,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2827,
      "query_norm": 1.4195,
      "queue_k_norm": 1.4173,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0201,
      "sent_len_1": 66.9928,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.4275,
      "stdk": 0.0472,
      "stdq": 0.044,
      "stdqueue_k": 0.0473,
      "stdqueue_q": 0.0,
      "step": 18300
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.698,
      "doc_norm": 1.4161,
      "encoder_q-embeddings": 1917.1768,
      "encoder_q-layer.0": 1400.553,
      "encoder_q-layer.1": 1473.6007,
      "encoder_q-layer.10": 342.4746,
      "encoder_q-layer.11": 876.2931,
      "encoder_q-layer.2": 1807.5647,
      "encoder_q-layer.3": 1839.1068,
      "encoder_q-layer.4": 1703.7872,
      "encoder_q-layer.5": 1806.1366,
      "encoder_q-layer.6": 1556.5087,
      "encoder_q-layer.7": 1121.4736,
      "encoder_q-layer.8": 537.0094,
      "encoder_q-layer.9": 337.6828,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2775,
      "inbatch_pos_score": 0.9263,
      "learning_rate": 4.5333333333333335e-05,
      "loss": 3.698,
      "norm_diff": 0.0287,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2120.0878,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2769,
      "query_norm": 1.4442,
      "queue_k_norm": 1.4181,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8453,
      "sent_len_1": 66.6626,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8475,
      "stdk": 0.0472,
      "stdq": 0.045,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 18400
    },
    {
      "accuracy": 49.5117,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6867,
      "doc_norm": 1.42,
      "encoder_q-embeddings": 850.2382,
      "encoder_q-layer.0": 606.9716,
      "encoder_q-layer.1": 657.2075,
      "encoder_q-layer.10": 351.2126,
      "encoder_q-layer.11": 923.2324,
      "encoder_q-layer.2": 726.4685,
      "encoder_q-layer.3": 855.2262,
      "encoder_q-layer.4": 773.5623,
      "encoder_q-layer.5": 742.8327,
      "encoder_q-layer.6": 608.9172,
      "encoder_q-layer.7": 540.7654,
      "encoder_q-layer.8": 474.1552,
      "encoder_q-layer.9": 346.6003,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2651,
      "inbatch_pos_score": 0.8491,
      "learning_rate": 4.527777777777778e-05,
      "loss": 3.6867,
      "norm_diff": 0.0099,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1024.4933,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2651,
      "query_norm": 1.4225,
      "queue_k_norm": 1.4169,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0281,
      "sent_len_1": 66.9267,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3487,
      "stdk": 0.0474,
      "stdq": 0.0444,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 18500
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7181,
      "doc_norm": 1.4191,
      "encoder_q-embeddings": 3529.3538,
      "encoder_q-layer.0": 2613.1453,
      "encoder_q-layer.1": 3096.8091,
      "encoder_q-layer.10": 344.2018,
      "encoder_q-layer.11": 831.6951,
      "encoder_q-layer.2": 3935.0105,
      "encoder_q-layer.3": 3979.6201,
      "encoder_q-layer.4": 3903.5308,
      "encoder_q-layer.5": 4278.9902,
      "encoder_q-layer.6": 3819.6604,
      "encoder_q-layer.7": 2883.311,
      "encoder_q-layer.8": 1395.9659,
      "encoder_q-layer.9": 422.3774,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2712,
      "inbatch_pos_score": 0.9214,
      "learning_rate": 4.522222222222223e-05,
      "loss": 3.7181,
      "norm_diff": 0.0598,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4474.0669,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2708,
      "query_norm": 1.4789,
      "queue_k_norm": 1.417,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7645,
      "sent_len_1": 66.5222,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.915,
      "stdk": 0.0474,
      "stdq": 0.0463,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 18600
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6604,
      "doc_norm": 1.4072,
      "encoder_q-embeddings": 1180.1023,
      "encoder_q-layer.0": 837.0622,
      "encoder_q-layer.1": 1045.6249,
      "encoder_q-layer.10": 367.8062,
      "encoder_q-layer.11": 868.9481,
      "encoder_q-layer.2": 1185.8783,
      "encoder_q-layer.3": 1153.547,
      "encoder_q-layer.4": 1017.0213,
      "encoder_q-layer.5": 1059.5062,
      "encoder_q-layer.6": 954.7859,
      "encoder_q-layer.7": 849.492,
      "encoder_q-layer.8": 494.3997,
      "encoder_q-layer.9": 349.8059,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2661,
      "inbatch_pos_score": 0.877,
      "learning_rate": 4.516666666666667e-05,
      "loss": 3.6604,
      "norm_diff": 0.0339,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1363.2353,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2649,
      "query_norm": 1.441,
      "queue_k_norm": 1.4181,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.002,
      "sent_len_1": 66.9397,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.87,
      "stdk": 0.0469,
      "stdq": 0.045,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 18700
    },
    {
      "accuracy": 48.7305,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7105,
      "doc_norm": 1.4191,
      "encoder_q-embeddings": 537.825,
      "encoder_q-layer.0": 432.1359,
      "encoder_q-layer.1": 422.2435,
      "encoder_q-layer.10": 368.3806,
      "encoder_q-layer.11": 841.0267,
      "encoder_q-layer.2": 447.5409,
      "encoder_q-layer.3": 456.2989,
      "encoder_q-layer.4": 495.9431,
      "encoder_q-layer.5": 451.4252,
      "encoder_q-layer.6": 539.4727,
      "encoder_q-layer.7": 644.567,
      "encoder_q-layer.8": 692.5669,
      "encoder_q-layer.9": 494.3228,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2647,
      "inbatch_pos_score": 0.874,
      "learning_rate": 4.511111111111112e-05,
      "loss": 3.7105,
      "norm_diff": 0.0216,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 820.0407,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2646,
      "query_norm": 1.4375,
      "queue_k_norm": 1.4157,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7787,
      "sent_len_1": 66.8917,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.435,
      "stdk": 0.0475,
      "stdq": 0.0453,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 18800
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7065,
      "doc_norm": 1.4165,
      "encoder_q-embeddings": 2604.0552,
      "encoder_q-layer.0": 1809.871,
      "encoder_q-layer.1": 2005.1831,
      "encoder_q-layer.10": 351.7495,
      "encoder_q-layer.11": 883.3115,
      "encoder_q-layer.2": 2369.7439,
      "encoder_q-layer.3": 2702.147,
      "encoder_q-layer.4": 2743.5701,
      "encoder_q-layer.5": 2926.5417,
      "encoder_q-layer.6": 3060.3096,
      "encoder_q-layer.7": 2069.6462,
      "encoder_q-layer.8": 1377.5507,
      "encoder_q-layer.9": 740.8237,
      "epoch": 0.18,
      "inbatch_neg_score": 0.2689,
      "inbatch_pos_score": 0.875,
      "learning_rate": 4.5055555555555554e-05,
      "loss": 3.7065,
      "norm_diff": 0.0092,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3181.6385,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2683,
      "query_norm": 1.4195,
      "queue_k_norm": 1.4169,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7128,
      "sent_len_1": 66.7373,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5425,
      "stdk": 0.0474,
      "stdq": 0.0443,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 18900
    },
    {
      "accuracy": 47.3633,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6766,
      "doc_norm": 1.4117,
      "encoder_q-embeddings": 841.6739,
      "encoder_q-layer.0": 578.217,
      "encoder_q-layer.1": 655.1666,
      "encoder_q-layer.10": 389.3698,
      "encoder_q-layer.11": 927.1432,
      "encoder_q-layer.2": 741.6503,
      "encoder_q-layer.3": 777.184,
      "encoder_q-layer.4": 744.1923,
      "encoder_q-layer.5": 723.9521,
      "encoder_q-layer.6": 626.8442,
      "encoder_q-layer.7": 586.6871,
      "encoder_q-layer.8": 520.6122,
      "encoder_q-layer.9": 389.704,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2665,
      "inbatch_pos_score": 0.8564,
      "learning_rate": 4.5e-05,
      "loss": 3.6766,
      "norm_diff": 0.0508,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1017.9742,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2651,
      "query_norm": 1.4625,
      "queue_k_norm": 1.4152,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0856,
      "sent_len_1": 66.6779,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7413,
      "stdk": 0.0472,
      "stdq": 0.0458,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 19000
    },
    {
      "accuracy": 47.9492,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6841,
      "doc_norm": 1.411,
      "encoder_q-embeddings": 1435.6006,
      "encoder_q-layer.0": 919.4254,
      "encoder_q-layer.1": 1148.8915,
      "encoder_q-layer.10": 390.1514,
      "encoder_q-layer.11": 999.1635,
      "encoder_q-layer.2": 1414.1001,
      "encoder_q-layer.3": 1389.9095,
      "encoder_q-layer.4": 1471.1411,
      "encoder_q-layer.5": 1565.064,
      "encoder_q-layer.6": 1256.8174,
      "encoder_q-layer.7": 1105.4943,
      "encoder_q-layer.8": 887.4532,
      "encoder_q-layer.9": 459.4852,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2558,
      "inbatch_pos_score": 0.853,
      "learning_rate": 4.4944444444444445e-05,
      "loss": 3.6841,
      "norm_diff": 0.051,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1746.1664,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2561,
      "query_norm": 1.462,
      "queue_k_norm": 1.414,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9936,
      "sent_len_1": 66.9517,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3913,
      "stdk": 0.0473,
      "stdq": 0.0455,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 19100
    },
    {
      "accuracy": 49.6094,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7105,
      "doc_norm": 1.4096,
      "encoder_q-embeddings": 906.2639,
      "encoder_q-layer.0": 638.5667,
      "encoder_q-layer.1": 683.2079,
      "encoder_q-layer.10": 394.6326,
      "encoder_q-layer.11": 1039.0537,
      "encoder_q-layer.2": 722.9813,
      "encoder_q-layer.3": 780.7156,
      "encoder_q-layer.4": 862.5294,
      "encoder_q-layer.5": 831.6731,
      "encoder_q-layer.6": 808.9592,
      "encoder_q-layer.7": 643.3441,
      "encoder_q-layer.8": 634.231,
      "encoder_q-layer.9": 435.3466,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2636,
      "inbatch_pos_score": 0.8525,
      "learning_rate": 4.4888888888888894e-05,
      "loss": 3.7105,
      "norm_diff": 0.0196,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1128.605,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2639,
      "query_norm": 1.4292,
      "queue_k_norm": 1.4124,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0158,
      "sent_len_1": 66.5321,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.98,
      "stdk": 0.0472,
      "stdq": 0.0442,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 19200
    },
    {
      "accuracy": 51.5625,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6779,
      "doc_norm": 1.4111,
      "encoder_q-embeddings": 4355.2681,
      "encoder_q-layer.0": 3277.9062,
      "encoder_q-layer.1": 3889.6436,
      "encoder_q-layer.10": 343.6989,
      "encoder_q-layer.11": 910.6744,
      "encoder_q-layer.2": 3658.7234,
      "encoder_q-layer.3": 4069.844,
      "encoder_q-layer.4": 4905.4619,
      "encoder_q-layer.5": 4545.8511,
      "encoder_q-layer.6": 4165.7329,
      "encoder_q-layer.7": 3857.793,
      "encoder_q-layer.8": 2449.2114,
      "encoder_q-layer.9": 556.5851,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2692,
      "inbatch_pos_score": 0.8984,
      "learning_rate": 4.483333333333333e-05,
      "loss": 3.6779,
      "norm_diff": 0.0415,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5447.5908,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2686,
      "query_norm": 1.4526,
      "queue_k_norm": 1.4109,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0261,
      "sent_len_1": 66.9331,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8562,
      "stdk": 0.0473,
      "stdq": 0.0452,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 19300
    },
    {
      "accuracy": 44.8242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6694,
      "doc_norm": 1.4096,
      "encoder_q-embeddings": 1714.5846,
      "encoder_q-layer.0": 1322.3607,
      "encoder_q-layer.1": 1218.5131,
      "encoder_q-layer.10": 389.3027,
      "encoder_q-layer.11": 932.1772,
      "encoder_q-layer.2": 1497.8259,
      "encoder_q-layer.3": 1390.214,
      "encoder_q-layer.4": 1577.8846,
      "encoder_q-layer.5": 1358.1139,
      "encoder_q-layer.6": 1219.8013,
      "encoder_q-layer.7": 1103.5815,
      "encoder_q-layer.8": 735.0831,
      "encoder_q-layer.9": 367.3231,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2455,
      "inbatch_pos_score": 0.8247,
      "learning_rate": 4.477777777777778e-05,
      "loss": 3.6694,
      "norm_diff": 0.0398,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1879.7838,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2454,
      "query_norm": 1.4494,
      "queue_k_norm": 1.4112,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9522,
      "sent_len_1": 66.6016,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4512,
      "stdk": 0.0473,
      "stdq": 0.0458,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 19400
    },
    {
      "accuracy": 50.9766,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6645,
      "doc_norm": 1.4099,
      "encoder_q-embeddings": 768.418,
      "encoder_q-layer.0": 514.7711,
      "encoder_q-layer.1": 544.7568,
      "encoder_q-layer.10": 394.1882,
      "encoder_q-layer.11": 1011.0898,
      "encoder_q-layer.2": 592.8517,
      "encoder_q-layer.3": 625.7061,
      "encoder_q-layer.4": 649.1873,
      "encoder_q-layer.5": 644.9188,
      "encoder_q-layer.6": 678.933,
      "encoder_q-layer.7": 695.6519,
      "encoder_q-layer.8": 631.4683,
      "encoder_q-layer.9": 403.2744,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2563,
      "inbatch_pos_score": 0.8901,
      "learning_rate": 4.472222222222223e-05,
      "loss": 3.6645,
      "norm_diff": 0.0653,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 992.9612,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2566,
      "query_norm": 1.4752,
      "queue_k_norm": 1.4123,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9196,
      "sent_len_1": 66.6443,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.475,
      "stdk": 0.0473,
      "stdq": 0.0462,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 19500
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.7062,
      "doc_norm": 1.4122,
      "encoder_q-embeddings": 405.7465,
      "encoder_q-layer.0": 257.8206,
      "encoder_q-layer.1": 266.2433,
      "encoder_q-layer.10": 354.1282,
      "encoder_q-layer.11": 879.9171,
      "encoder_q-layer.2": 313.5178,
      "encoder_q-layer.3": 328.8215,
      "encoder_q-layer.4": 347.3798,
      "encoder_q-layer.5": 357.4089,
      "encoder_q-layer.6": 391.5436,
      "encoder_q-layer.7": 418.9671,
      "encoder_q-layer.8": 413.9175,
      "encoder_q-layer.9": 351.6747,
      "epoch": 0.19,
      "inbatch_neg_score": 0.265,
      "inbatch_pos_score": 0.9229,
      "learning_rate": 4.466666666666667e-05,
      "loss": 3.7062,
      "norm_diff": 0.0528,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 640.892,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2656,
      "query_norm": 1.465,
      "queue_k_norm": 1.4107,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1508,
      "sent_len_1": 66.9686,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3925,
      "stdk": 0.0475,
      "stdq": 0.046,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 19600
    },
    {
      "accuracy": 52.9297,
      "active_queue_size": 16384.0,
      "cl_loss": 3.672,
      "doc_norm": 1.4143,
      "encoder_q-embeddings": 1640.1128,
      "encoder_q-layer.0": 1259.7136,
      "encoder_q-layer.1": 1392.3971,
      "encoder_q-layer.10": 367.5541,
      "encoder_q-layer.11": 909.3582,
      "encoder_q-layer.2": 1427.1775,
      "encoder_q-layer.3": 1351.9202,
      "encoder_q-layer.4": 1372.3774,
      "encoder_q-layer.5": 1088.1914,
      "encoder_q-layer.6": 1106.3669,
      "encoder_q-layer.7": 849.8629,
      "encoder_q-layer.8": 523.7574,
      "encoder_q-layer.9": 356.1806,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2545,
      "inbatch_pos_score": 0.8906,
      "learning_rate": 4.461111111111111e-05,
      "loss": 3.672,
      "norm_diff": 0.0388,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1747.0146,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2529,
      "query_norm": 1.4531,
      "queue_k_norm": 1.4107,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8396,
      "sent_len_1": 66.7518,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7075,
      "stdk": 0.0475,
      "stdq": 0.046,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 19700
    },
    {
      "accuracy": 44.3359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6734,
      "doc_norm": 1.4108,
      "encoder_q-embeddings": 2004.9823,
      "encoder_q-layer.0": 1528.8389,
      "encoder_q-layer.1": 1695.5734,
      "encoder_q-layer.10": 398.7404,
      "encoder_q-layer.11": 1009.4551,
      "encoder_q-layer.2": 2071.167,
      "encoder_q-layer.3": 2283.2966,
      "encoder_q-layer.4": 2285.8606,
      "encoder_q-layer.5": 2390.7617,
      "encoder_q-layer.6": 2379.7996,
      "encoder_q-layer.7": 1900.4156,
      "encoder_q-layer.8": 1438.2664,
      "encoder_q-layer.9": 730.7394,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2595,
      "inbatch_pos_score": 0.8311,
      "learning_rate": 4.4555555555555555e-05,
      "loss": 3.6734,
      "norm_diff": 0.0141,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2766.6239,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2598,
      "query_norm": 1.4242,
      "queue_k_norm": 1.4086,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.019,
      "sent_len_1": 66.9759,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7075,
      "stdk": 0.0474,
      "stdq": 0.0451,
      "stdqueue_k": 0.0474,
      "stdqueue_q": 0.0,
      "step": 19800
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6805,
      "doc_norm": 1.4064,
      "encoder_q-embeddings": 1884.6013,
      "encoder_q-layer.0": 1273.4453,
      "encoder_q-layer.1": 1402.7159,
      "encoder_q-layer.10": 358.041,
      "encoder_q-layer.11": 919.1593,
      "encoder_q-layer.2": 1530.1965,
      "encoder_q-layer.3": 1641.0219,
      "encoder_q-layer.4": 1648.6068,
      "encoder_q-layer.5": 1517.2518,
      "encoder_q-layer.6": 1763.938,
      "encoder_q-layer.7": 1259.98,
      "encoder_q-layer.8": 569.788,
      "encoder_q-layer.9": 336.459,
      "epoch": 0.19,
      "inbatch_neg_score": 0.2498,
      "inbatch_pos_score": 0.8667,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 3.6805,
      "norm_diff": 0.0206,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2016.006,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2507,
      "query_norm": 1.427,
      "queue_k_norm": 1.4095,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0469,
      "sent_len_1": 66.6552,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.68,
      "stdk": 0.0473,
      "stdq": 0.0454,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 19900
    },
    {
      "accuracy": 48.8281,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6897,
      "doc_norm": 1.4125,
      "encoder_q-embeddings": 1515.3854,
      "encoder_q-layer.0": 973.5408,
      "encoder_q-layer.1": 1053.8535,
      "encoder_q-layer.10": 712.5607,
      "encoder_q-layer.11": 1931.8512,
      "encoder_q-layer.2": 1214.9441,
      "encoder_q-layer.3": 1237.6625,
      "encoder_q-layer.4": 1373.9677,
      "encoder_q-layer.5": 1391.6843,
      "encoder_q-layer.6": 1477.0901,
      "encoder_q-layer.7": 1419.6503,
      "encoder_q-layer.8": 1195.9203,
      "encoder_q-layer.9": 710.5776,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2443,
      "inbatch_pos_score": 0.8462,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 3.6897,
      "norm_diff": 0.012,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1984.8954,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2441,
      "query_norm": 1.4102,
      "queue_k_norm": 1.4122,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9698,
      "sent_len_1": 66.7501,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9412,
      "stdk": 0.0475,
      "stdq": 0.045,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20000
    },
    {
      "dev_runtime": 37.8767,
      "dev_samples_per_second": 1.69,
      "dev_steps_per_second": 0.026,
      "epoch": 0.2,
      "step": 20000,
      "test_accuracy": 93.4814453125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.38453954458236694,
      "test_doc_norm": 1.4031305313110352,
      "test_inbatch_neg_score": 0.5904818773269653,
      "test_inbatch_pos_score": 1.4654431343078613,
      "test_loss": 0.38453954458236694,
      "test_loss_align": 1.1792842149734497,
      "test_loss_unif": 3.8744869232177734,
      "test_loss_unif_q@queue": 3.8744866847991943,
      "test_norm_diff": 0.051639288663864136,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.2303803563117981,
      "test_query_norm": 1.4547698497772217,
      "test_queue_k_norm": 1.4126696586608887,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04144138842821121,
      "test_stdq": 0.04164225608110428,
      "test_stdqueue_k": 0.04765129089355469,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 37.8767,
      "dev_samples_per_second": 1.69,
      "dev_steps_per_second": 0.026,
      "epoch": 0.2,
      "eval_beir-arguana_ndcg@10": 0.32258,
      "eval_beir-arguana_recall@10": 0.55263,
      "eval_beir-arguana_recall@100": 0.85846,
      "eval_beir-arguana_recall@20": 0.69132,
      "eval_beir-avg_ndcg@10": 0.31629966666666665,
      "eval_beir-avg_recall@10": 0.37913633333333335,
      "eval_beir-avg_recall@100": 0.5573134166666667,
      "eval_beir-avg_recall@20": 0.43658425,
      "eval_beir-cqadupstack_ndcg@10": 0.21258666666666662,
      "eval_beir-cqadupstack_recall@10": 0.29197333333333336,
      "eval_beir-cqadupstack_recall@100": 0.5054841666666667,
      "eval_beir-cqadupstack_recall@20": 0.3496225,
      "eval_beir-fiqa_ndcg@10": 0.15151,
      "eval_beir-fiqa_recall@10": 0.19531,
      "eval_beir-fiqa_recall@100": 0.42902,
      "eval_beir-fiqa_recall@20": 0.24829,
      "eval_beir-nfcorpus_ndcg@10": 0.23706,
      "eval_beir-nfcorpus_recall@10": 0.11762,
      "eval_beir-nfcorpus_recall@100": 0.23963,
      "eval_beir-nfcorpus_recall@20": 0.14578,
      "eval_beir-nq_ndcg@10": 0.20267,
      "eval_beir-nq_recall@10": 0.34299,
      "eval_beir-nq_recall@100": 0.67017,
      "eval_beir-nq_recall@20": 0.45184,
      "eval_beir-quora_ndcg@10": 0.76882,
      "eval_beir-quora_recall@10": 0.87736,
      "eval_beir-quora_recall@100": 0.97153,
      "eval_beir-quora_recall@20": 0.9176,
      "eval_beir-scidocs_ndcg@10": 0.11634,
      "eval_beir-scidocs_recall@10": 0.12388,
      "eval_beir-scidocs_recall@100": 0.29848,
      "eval_beir-scidocs_recall@20": 0.17203,
      "eval_beir-scifact_ndcg@10": 0.5606,
      "eval_beir-scifact_recall@10": 0.71783,
      "eval_beir-scifact_recall@100": 0.86589,
      "eval_beir-scifact_recall@20": 0.76867,
      "eval_beir-trec-covid_ndcg@10": 0.42215,
      "eval_beir-trec-covid_recall@10": 0.448,
      "eval_beir-trec-covid_recall@100": 0.3212,
      "eval_beir-trec-covid_recall@20": 0.431,
      "eval_beir-webis-touche2020_ndcg@10": 0.16868,
      "eval_beir-webis-touche2020_recall@10": 0.12377,
      "eval_beir-webis-touche2020_recall@100": 0.41327,
      "eval_beir-webis-touche2020_recall@20": 0.18969,
      "eval_senteval-avg_sts": 0.7037160105314686,
      "eval_senteval-sickr_spearman": 0.6874844287739638,
      "eval_senteval-stsb_spearman": 0.7199475922889734,
      "step": 20000,
      "test_accuracy": 93.4814453125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.38453954458236694,
      "test_doc_norm": 1.4031305313110352,
      "test_inbatch_neg_score": 0.5904818773269653,
      "test_inbatch_pos_score": 1.4654431343078613,
      "test_loss": 0.38453954458236694,
      "test_loss_align": 1.1792842149734497,
      "test_loss_unif": 3.8744869232177734,
      "test_loss_unif_q@queue": 3.8744866847991943,
      "test_norm_diff": 0.051639288663864136,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.2303803563117981,
      "test_query_norm": 1.4547698497772217,
      "test_queue_k_norm": 1.4126696586608887,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04144138842821121,
      "test_stdq": 0.04164225608110428,
      "test_stdqueue_k": 0.04765129089355469,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 51.0742,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6744,
      "doc_norm": 1.4068,
      "encoder_q-embeddings": 1238.1823,
      "encoder_q-layer.0": 925.5707,
      "encoder_q-layer.1": 949.5054,
      "encoder_q-layer.10": 732.6279,
      "encoder_q-layer.11": 1768.1814,
      "encoder_q-layer.2": 1155.6656,
      "encoder_q-layer.3": 1195.7097,
      "encoder_q-layer.4": 1233.547,
      "encoder_q-layer.5": 1157.1794,
      "encoder_q-layer.6": 903.8389,
      "encoder_q-layer.7": 916.5795,
      "encoder_q-layer.8": 787.5237,
      "encoder_q-layer.9": 641.6774,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2212,
      "inbatch_pos_score": 0.833,
      "learning_rate": 4.438888888888889e-05,
      "loss": 3.6744,
      "norm_diff": 0.0091,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1655.3701,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2202,
      "query_norm": 1.4073,
      "queue_k_norm": 1.4094,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7549,
      "sent_len_1": 66.6159,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1687,
      "stdk": 0.0474,
      "stdq": 0.0452,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20100
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6623,
      "doc_norm": 1.409,
      "encoder_q-embeddings": 1439.5933,
      "encoder_q-layer.0": 929.0284,
      "encoder_q-layer.1": 996.3851,
      "encoder_q-layer.10": 730.9801,
      "encoder_q-layer.11": 1850.4895,
      "encoder_q-layer.2": 1296.5708,
      "encoder_q-layer.3": 1459.0968,
      "encoder_q-layer.4": 1264.9927,
      "encoder_q-layer.5": 1038.6907,
      "encoder_q-layer.6": 935.4612,
      "encoder_q-layer.7": 875.5612,
      "encoder_q-layer.8": 812.1811,
      "encoder_q-layer.9": 663.8297,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2161,
      "inbatch_pos_score": 0.7925,
      "learning_rate": 4.433333333333334e-05,
      "loss": 3.6623,
      "norm_diff": 0.0302,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1762.4213,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2156,
      "query_norm": 1.3788,
      "queue_k_norm": 1.4098,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9586,
      "sent_len_1": 66.8587,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6725,
      "stdk": 0.0475,
      "stdq": 0.044,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20200
    },
    {
      "accuracy": 49.2188,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6457,
      "doc_norm": 1.4042,
      "encoder_q-embeddings": 3970.1138,
      "encoder_q-layer.0": 2643.5166,
      "encoder_q-layer.1": 3236.8169,
      "encoder_q-layer.10": 724.2651,
      "encoder_q-layer.11": 1924.8302,
      "encoder_q-layer.2": 3779.6345,
      "encoder_q-layer.3": 3338.1926,
      "encoder_q-layer.4": 2633.03,
      "encoder_q-layer.5": 2155.969,
      "encoder_q-layer.6": 2113.4915,
      "encoder_q-layer.7": 1804.7622,
      "encoder_q-layer.8": 1245.2444,
      "encoder_q-layer.9": 670.4775,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2329,
      "inbatch_pos_score": 0.8203,
      "learning_rate": 4.427777777777778e-05,
      "loss": 3.6457,
      "norm_diff": 0.0145,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3940.9865,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2329,
      "query_norm": 1.4032,
      "queue_k_norm": 1.4064,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0194,
      "sent_len_1": 66.7255,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.07,
      "stdk": 0.0474,
      "stdq": 0.0447,
      "stdqueue_k": 0.0475,
      "stdqueue_q": 0.0,
      "step": 20300
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6531,
      "doc_norm": 1.4052,
      "encoder_q-embeddings": 4181.7427,
      "encoder_q-layer.0": 3084.0969,
      "encoder_q-layer.1": 3130.3386,
      "encoder_q-layer.10": 753.1086,
      "encoder_q-layer.11": 1938.9761,
      "encoder_q-layer.2": 3454.6311,
      "encoder_q-layer.3": 3467.418,
      "encoder_q-layer.4": 3281.8911,
      "encoder_q-layer.5": 3250.7256,
      "encoder_q-layer.6": 3268.5459,
      "encoder_q-layer.7": 2430.9236,
      "encoder_q-layer.8": 1431.8525,
      "encoder_q-layer.9": 755.4531,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2249,
      "inbatch_pos_score": 0.8477,
      "learning_rate": 4.422222222222222e-05,
      "loss": 3.6531,
      "norm_diff": 0.02,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4415.1079,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2236,
      "query_norm": 1.3857,
      "queue_k_norm": 1.4066,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9556,
      "sent_len_1": 66.855,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.25,
      "stdk": 0.0474,
      "stdq": 0.0445,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20400
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6482,
      "doc_norm": 1.4114,
      "encoder_q-embeddings": 16276.501,
      "encoder_q-layer.0": 10585.6572,
      "encoder_q-layer.1": 11991.8564,
      "encoder_q-layer.10": 679.3613,
      "encoder_q-layer.11": 1942.0946,
      "encoder_q-layer.2": 14036.0742,
      "encoder_q-layer.3": 12844.5117,
      "encoder_q-layer.4": 11718.5664,
      "encoder_q-layer.5": 11564.8691,
      "encoder_q-layer.6": 9798.3516,
      "encoder_q-layer.7": 6338.7764,
      "encoder_q-layer.8": 2268.8647,
      "encoder_q-layer.9": 777.4878,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2219,
      "inbatch_pos_score": 0.8105,
      "learning_rate": 4.4166666666666665e-05,
      "loss": 3.6482,
      "norm_diff": 0.0207,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 15443.6663,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2211,
      "query_norm": 1.3907,
      "queue_k_norm": 1.4066,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9714,
      "sent_len_1": 66.7399,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0575,
      "stdk": 0.0477,
      "stdq": 0.0448,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20500
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6503,
      "doc_norm": 1.4086,
      "encoder_q-embeddings": 871.1415,
      "encoder_q-layer.0": 625.0912,
      "encoder_q-layer.1": 672.2936,
      "encoder_q-layer.10": 714.1212,
      "encoder_q-layer.11": 1742.2745,
      "encoder_q-layer.2": 704.2947,
      "encoder_q-layer.3": 736.4985,
      "encoder_q-layer.4": 760.8802,
      "encoder_q-layer.5": 757.4772,
      "encoder_q-layer.6": 741.2864,
      "encoder_q-layer.7": 717.8003,
      "encoder_q-layer.8": 764.9335,
      "encoder_q-layer.9": 633.3941,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2176,
      "inbatch_pos_score": 0.8765,
      "learning_rate": 4.4111111111111114e-05,
      "loss": 3.6503,
      "norm_diff": 0.0101,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1303.0323,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.218,
      "query_norm": 1.4108,
      "queue_k_norm": 1.4055,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9286,
      "sent_len_1": 66.7078,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.53,
      "stdk": 0.0476,
      "stdq": 0.0458,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20600
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6541,
      "doc_norm": 1.4052,
      "encoder_q-embeddings": 2042.7272,
      "encoder_q-layer.0": 1346.1987,
      "encoder_q-layer.1": 1533.4705,
      "encoder_q-layer.10": 824.86,
      "encoder_q-layer.11": 1810.4769,
      "encoder_q-layer.2": 1749.7375,
      "encoder_q-layer.3": 1957.0193,
      "encoder_q-layer.4": 2044.6406,
      "encoder_q-layer.5": 2293.2798,
      "encoder_q-layer.6": 2289.8018,
      "encoder_q-layer.7": 1950.1404,
      "encoder_q-layer.8": 1499.7225,
      "encoder_q-layer.9": 719.4613,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2127,
      "inbatch_pos_score": 0.8423,
      "learning_rate": 4.4055555555555557e-05,
      "loss": 3.6541,
      "norm_diff": 0.0311,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2612.9304,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2124,
      "query_norm": 1.3741,
      "queue_k_norm": 1.4066,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0558,
      "sent_len_1": 66.6157,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.2587,
      "stdk": 0.0475,
      "stdq": 0.0445,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 20700
    },
    {
      "accuracy": 51.7578,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6291,
      "doc_norm": 1.404,
      "encoder_q-embeddings": 1006.7381,
      "encoder_q-layer.0": 713.5001,
      "encoder_q-layer.1": 830.8566,
      "encoder_q-layer.10": 683.2622,
      "encoder_q-layer.11": 1700.9886,
      "encoder_q-layer.2": 1034.4989,
      "encoder_q-layer.3": 1172.0814,
      "encoder_q-layer.4": 1445.7532,
      "encoder_q-layer.5": 1414.3739,
      "encoder_q-layer.6": 1437.4263,
      "encoder_q-layer.7": 1283.6648,
      "encoder_q-layer.8": 899.0558,
      "encoder_q-layer.9": 658.41,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2134,
      "inbatch_pos_score": 0.8481,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 3.6291,
      "norm_diff": 0.0082,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1715.4243,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.213,
      "query_norm": 1.4087,
      "queue_k_norm": 1.4054,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0729,
      "sent_len_1": 67.0935,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1662,
      "stdk": 0.0475,
      "stdq": 0.0457,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20800
    },
    {
      "accuracy": 49.1211,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6512,
      "doc_norm": 1.3966,
      "encoder_q-embeddings": 1974.3002,
      "encoder_q-layer.0": 1432.636,
      "encoder_q-layer.1": 1569.4801,
      "encoder_q-layer.10": 692.991,
      "encoder_q-layer.11": 1889.0093,
      "encoder_q-layer.2": 1758.0071,
      "encoder_q-layer.3": 2012.0502,
      "encoder_q-layer.4": 2043.1879,
      "encoder_q-layer.5": 1716.4247,
      "encoder_q-layer.6": 1480.1908,
      "encoder_q-layer.7": 1156.9004,
      "encoder_q-layer.8": 876.6447,
      "encoder_q-layer.9": 665.3245,
      "epoch": 0.2,
      "inbatch_neg_score": 0.2342,
      "inbatch_pos_score": 0.8184,
      "learning_rate": 4.394444444444445e-05,
      "loss": 3.6512,
      "norm_diff": 0.0087,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2392.7868,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2323,
      "query_norm": 1.3986,
      "queue_k_norm": 1.4043,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9641,
      "sent_len_1": 66.7494,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8975,
      "stdk": 0.0472,
      "stdq": 0.0444,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 20900
    },
    {
      "accuracy": 51.5625,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6157,
      "doc_norm": 1.4101,
      "encoder_q-embeddings": 5810.2817,
      "encoder_q-layer.0": 4015.791,
      "encoder_q-layer.1": 4422.1763,
      "encoder_q-layer.10": 670.6324,
      "encoder_q-layer.11": 1809.5466,
      "encoder_q-layer.2": 4801.3589,
      "encoder_q-layer.3": 5022.9961,
      "encoder_q-layer.4": 5381.0029,
      "encoder_q-layer.5": 5293.3213,
      "encoder_q-layer.6": 4826.5776,
      "encoder_q-layer.7": 3938.0117,
      "encoder_q-layer.8": 2190.3606,
      "encoder_q-layer.9": 801.5458,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2404,
      "inbatch_pos_score": 0.8662,
      "learning_rate": 4.388888888888889e-05,
      "loss": 3.6157,
      "norm_diff": 0.0098,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6286.9261,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2394,
      "query_norm": 1.4196,
      "queue_k_norm": 1.4043,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0225,
      "sent_len_1": 66.9417,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.3125,
      "stdk": 0.0477,
      "stdq": 0.0453,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21000
    },
    {
      "accuracy": 51.0742,
      "active_queue_size": 16384.0,
      "cl_loss": 3.656,
      "doc_norm": 1.4026,
      "encoder_q-embeddings": 5646.7222,
      "encoder_q-layer.0": 4256.7993,
      "encoder_q-layer.1": 4694.647,
      "encoder_q-layer.10": 734.9106,
      "encoder_q-layer.11": 1840.3044,
      "encoder_q-layer.2": 5579.1372,
      "encoder_q-layer.3": 5157.7061,
      "encoder_q-layer.4": 4995.8496,
      "encoder_q-layer.5": 4540.4292,
      "encoder_q-layer.6": 4041.5437,
      "encoder_q-layer.7": 3336.3159,
      "encoder_q-layer.8": 1565.9414,
      "encoder_q-layer.9": 748.8441,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2333,
      "inbatch_pos_score": 0.8481,
      "learning_rate": 4.383333333333334e-05,
      "loss": 3.656,
      "norm_diff": 0.0206,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6086.5939,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2319,
      "query_norm": 1.4208,
      "queue_k_norm": 1.4045,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9441,
      "sent_len_1": 66.7035,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.95,
      "stdk": 0.0474,
      "stdq": 0.0455,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21100
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6412,
      "doc_norm": 1.4017,
      "encoder_q-embeddings": 3397.7815,
      "encoder_q-layer.0": 2265.3176,
      "encoder_q-layer.1": 2412.0835,
      "encoder_q-layer.10": 682.3354,
      "encoder_q-layer.11": 1834.3969,
      "encoder_q-layer.2": 2439.4292,
      "encoder_q-layer.3": 2590.4866,
      "encoder_q-layer.4": 2695.3855,
      "encoder_q-layer.5": 2467.2727,
      "encoder_q-layer.6": 2469.6465,
      "encoder_q-layer.7": 2196.1777,
      "encoder_q-layer.8": 1317.3826,
      "encoder_q-layer.9": 707.8461,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2357,
      "inbatch_pos_score": 0.8682,
      "learning_rate": 4.377777777777778e-05,
      "loss": 3.6412,
      "norm_diff": 0.0127,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3498.2947,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2351,
      "query_norm": 1.4102,
      "queue_k_norm": 1.4039,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8642,
      "sent_len_1": 66.8108,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2463,
      "stdk": 0.0474,
      "stdq": 0.0451,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21200
    },
    {
      "accuracy": 50.7812,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6352,
      "doc_norm": 1.4067,
      "encoder_q-embeddings": 1585.1516,
      "encoder_q-layer.0": 1246.2286,
      "encoder_q-layer.1": 1371.1304,
      "encoder_q-layer.10": 693.8217,
      "encoder_q-layer.11": 1801.6388,
      "encoder_q-layer.2": 1323.1305,
      "encoder_q-layer.3": 1247.5039,
      "encoder_q-layer.4": 1215.8278,
      "encoder_q-layer.5": 1159.7771,
      "encoder_q-layer.6": 1224.9232,
      "encoder_q-layer.7": 1083.1996,
      "encoder_q-layer.8": 871.6111,
      "encoder_q-layer.9": 681.1541,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2299,
      "inbatch_pos_score": 0.8818,
      "learning_rate": 4.3722222222222224e-05,
      "loss": 3.6352,
      "norm_diff": 0.0376,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1897.4008,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2316,
      "query_norm": 1.4443,
      "queue_k_norm": 1.4035,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0378,
      "sent_len_1": 66.6547,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4487,
      "stdk": 0.0476,
      "stdq": 0.0462,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21300
    },
    {
      "accuracy": 51.0742,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6103,
      "doc_norm": 1.4075,
      "encoder_q-embeddings": 2062.8748,
      "encoder_q-layer.0": 1451.9348,
      "encoder_q-layer.1": 1659.4902,
      "encoder_q-layer.10": 728.5291,
      "encoder_q-layer.11": 1870.8025,
      "encoder_q-layer.2": 1936.7646,
      "encoder_q-layer.3": 1959.7657,
      "encoder_q-layer.4": 2234.4285,
      "encoder_q-layer.5": 2147.4204,
      "encoder_q-layer.6": 1967.3094,
      "encoder_q-layer.7": 1845.4354,
      "encoder_q-layer.8": 1254.1143,
      "encoder_q-layer.9": 728.0189,
      "epoch": 0.21,
      "inbatch_neg_score": 0.246,
      "inbatch_pos_score": 0.8613,
      "learning_rate": 4.3666666666666666e-05,
      "loss": 3.6103,
      "norm_diff": 0.0335,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2639.4306,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2455,
      "query_norm": 1.441,
      "queue_k_norm": 1.4039,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9402,
      "sent_len_1": 66.9216,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6125,
      "stdk": 0.0476,
      "stdq": 0.0461,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21400
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6427,
      "doc_norm": 1.4075,
      "encoder_q-embeddings": 1065.8875,
      "encoder_q-layer.0": 688.2351,
      "encoder_q-layer.1": 746.5103,
      "encoder_q-layer.10": 653.9229,
      "encoder_q-layer.11": 1761.7258,
      "encoder_q-layer.2": 763.0791,
      "encoder_q-layer.3": 759.4487,
      "encoder_q-layer.4": 769.9341,
      "encoder_q-layer.5": 751.7296,
      "encoder_q-layer.6": 731.0483,
      "encoder_q-layer.7": 771.0158,
      "encoder_q-layer.8": 748.4575,
      "encoder_q-layer.9": 641.0685,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2496,
      "inbatch_pos_score": 0.8638,
      "learning_rate": 4.3611111111111116e-05,
      "loss": 3.6427,
      "norm_diff": 0.0159,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1400.7476,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.249,
      "query_norm": 1.4234,
      "queue_k_norm": 1.405,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9799,
      "sent_len_1": 66.6345,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.57,
      "stdk": 0.0476,
      "stdq": 0.0455,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21500
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6355,
      "doc_norm": 1.4073,
      "encoder_q-embeddings": 2269.9583,
      "encoder_q-layer.0": 1546.8578,
      "encoder_q-layer.1": 1677.6477,
      "encoder_q-layer.10": 725.0486,
      "encoder_q-layer.11": 2000.2567,
      "encoder_q-layer.2": 1898.5203,
      "encoder_q-layer.3": 2024.0823,
      "encoder_q-layer.4": 1989.0905,
      "encoder_q-layer.5": 1732.9651,
      "encoder_q-layer.6": 1795.0535,
      "encoder_q-layer.7": 1702.813,
      "encoder_q-layer.8": 1158.6821,
      "encoder_q-layer.9": 704.8473,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2494,
      "inbatch_pos_score": 0.8682,
      "learning_rate": 4.355555555555556e-05,
      "loss": 3.6355,
      "norm_diff": 0.0231,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2592.9102,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2498,
      "query_norm": 1.4304,
      "queue_k_norm": 1.405,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7478,
      "sent_len_1": 66.685,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2912,
      "stdk": 0.0476,
      "stdq": 0.0457,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21600
    },
    {
      "accuracy": 49.707,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6368,
      "doc_norm": 1.4099,
      "encoder_q-embeddings": 2372.5952,
      "encoder_q-layer.0": 1738.2679,
      "encoder_q-layer.1": 1783.7593,
      "encoder_q-layer.10": 721.6495,
      "encoder_q-layer.11": 1936.2815,
      "encoder_q-layer.2": 2052.4722,
      "encoder_q-layer.3": 2236.1294,
      "encoder_q-layer.4": 2327.7346,
      "encoder_q-layer.5": 2883.3545,
      "encoder_q-layer.6": 2531.7109,
      "encoder_q-layer.7": 2248.4209,
      "encoder_q-layer.8": 1980.8938,
      "encoder_q-layer.9": 1234.1089,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2501,
      "inbatch_pos_score": 0.8701,
      "learning_rate": 4.35e-05,
      "loss": 3.6368,
      "norm_diff": 0.0284,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3092.4001,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2499,
      "query_norm": 1.4383,
      "queue_k_norm": 1.4058,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9446,
      "sent_len_1": 66.8822,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1962,
      "stdk": 0.0477,
      "stdq": 0.0458,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21700
    },
    {
      "accuracy": 49.9023,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6365,
      "doc_norm": 1.4105,
      "encoder_q-embeddings": 1175.5316,
      "encoder_q-layer.0": 806.7217,
      "encoder_q-layer.1": 881.6093,
      "encoder_q-layer.10": 728.8169,
      "encoder_q-layer.11": 1954.3879,
      "encoder_q-layer.2": 983.6409,
      "encoder_q-layer.3": 1095.2211,
      "encoder_q-layer.4": 1220.3379,
      "encoder_q-layer.5": 1112.9543,
      "encoder_q-layer.6": 1138.5565,
      "encoder_q-layer.7": 1181.7294,
      "encoder_q-layer.8": 957.9765,
      "encoder_q-layer.9": 673.194,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2456,
      "inbatch_pos_score": 0.8623,
      "learning_rate": 4.344444444444445e-05,
      "loss": 3.6365,
      "norm_diff": 0.021,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1711.0514,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2471,
      "query_norm": 1.4315,
      "queue_k_norm": 1.406,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9044,
      "sent_len_1": 66.9669,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2212,
      "stdk": 0.0477,
      "stdq": 0.0456,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21800
    },
    {
      "accuracy": 48.6328,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6067,
      "doc_norm": 1.4021,
      "encoder_q-embeddings": 892.0833,
      "encoder_q-layer.0": 614.1625,
      "encoder_q-layer.1": 672.5717,
      "encoder_q-layer.10": 740.2905,
      "encoder_q-layer.11": 1902.2957,
      "encoder_q-layer.2": 692.403,
      "encoder_q-layer.3": 711.4027,
      "encoder_q-layer.4": 703.5872,
      "encoder_q-layer.5": 666.347,
      "encoder_q-layer.6": 728.6129,
      "encoder_q-layer.7": 790.4483,
      "encoder_q-layer.8": 836.1508,
      "encoder_q-layer.9": 675.3912,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2405,
      "inbatch_pos_score": 0.8467,
      "learning_rate": 4.338888888888889e-05,
      "loss": 3.6067,
      "norm_diff": 0.0373,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1377.3941,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2404,
      "query_norm": 1.4394,
      "queue_k_norm": 1.407,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8823,
      "sent_len_1": 66.8593,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4688,
      "stdk": 0.0474,
      "stdq": 0.0458,
      "stdqueue_k": 0.0476,
      "stdqueue_q": 0.0,
      "step": 21900
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6101,
      "doc_norm": 1.4152,
      "encoder_q-embeddings": 2503.8142,
      "encoder_q-layer.0": 1998.1382,
      "encoder_q-layer.1": 2137.5024,
      "encoder_q-layer.10": 1318.1263,
      "encoder_q-layer.11": 3457.5308,
      "encoder_q-layer.2": 2175.4136,
      "encoder_q-layer.3": 2040.3018,
      "encoder_q-layer.4": 2039.0209,
      "encoder_q-layer.5": 1890.4874,
      "encoder_q-layer.6": 1991.7964,
      "encoder_q-layer.7": 1963.7295,
      "encoder_q-layer.8": 1730.4539,
      "encoder_q-layer.9": 1318.6471,
      "epoch": 0.21,
      "inbatch_neg_score": 0.2459,
      "inbatch_pos_score": 0.8872,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 3.6101,
      "norm_diff": 0.0177,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3240.5889,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2458,
      "query_norm": 1.4268,
      "queue_k_norm": 1.4079,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1557,
      "sent_len_1": 67.2133,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3688,
      "stdk": 0.0479,
      "stdq": 0.0452,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 22000
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6051,
      "doc_norm": 1.4069,
      "encoder_q-embeddings": 3706.0234,
      "encoder_q-layer.0": 2532.8706,
      "encoder_q-layer.1": 2861.2012,
      "encoder_q-layer.10": 1579.2969,
      "encoder_q-layer.11": 4130.9136,
      "encoder_q-layer.2": 2835.8088,
      "encoder_q-layer.3": 3017.5691,
      "encoder_q-layer.4": 3304.2693,
      "encoder_q-layer.5": 3140.1267,
      "encoder_q-layer.6": 3269.9512,
      "encoder_q-layer.7": 3058.936,
      "encoder_q-layer.8": 2337.7449,
      "encoder_q-layer.9": 1430.2625,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2445,
      "inbatch_pos_score": 0.8354,
      "learning_rate": 4.3277777777777776e-05,
      "loss": 3.6051,
      "norm_diff": 0.0064,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4508.6883,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2434,
      "query_norm": 1.402,
      "queue_k_norm": 1.4082,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0459,
      "sent_len_1": 66.7114,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3787,
      "stdk": 0.0475,
      "stdq": 0.0444,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 22100
    },
    {
      "accuracy": 50.4883,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5998,
      "doc_norm": 1.4075,
      "encoder_q-embeddings": 3496.5762,
      "encoder_q-layer.0": 2316.728,
      "encoder_q-layer.1": 2661.2573,
      "encoder_q-layer.10": 1424.775,
      "encoder_q-layer.11": 3567.1377,
      "encoder_q-layer.2": 2636.9011,
      "encoder_q-layer.3": 2617.9202,
      "encoder_q-layer.4": 2501.2041,
      "encoder_q-layer.5": 2559.8503,
      "encoder_q-layer.6": 2073.1841,
      "encoder_q-layer.7": 2142.2771,
      "encoder_q-layer.8": 1627.7817,
      "encoder_q-layer.9": 1267.4159,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2448,
      "inbatch_pos_score": 0.8735,
      "learning_rate": 4.3222222222222226e-05,
      "loss": 3.5998,
      "norm_diff": 0.0246,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3840.7987,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2439,
      "query_norm": 1.4316,
      "queue_k_norm": 1.4081,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9578,
      "sent_len_1": 66.8299,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8613,
      "stdk": 0.0475,
      "stdq": 0.0457,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 22200
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6219,
      "doc_norm": 1.4092,
      "encoder_q-embeddings": 13431.8135,
      "encoder_q-layer.0": 9404.4385,
      "encoder_q-layer.1": 11312.5293,
      "encoder_q-layer.10": 1336.5458,
      "encoder_q-layer.11": 3306.5383,
      "encoder_q-layer.2": 11131.9424,
      "encoder_q-layer.3": 12300.3291,
      "encoder_q-layer.4": 10938.8848,
      "encoder_q-layer.5": 10187.208,
      "encoder_q-layer.6": 10392.1484,
      "encoder_q-layer.7": 8817.3623,
      "encoder_q-layer.8": 4345.5991,
      "encoder_q-layer.9": 1631.3113,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2318,
      "inbatch_pos_score": 0.8608,
      "learning_rate": 4.316666666666667e-05,
      "loss": 3.6219,
      "norm_diff": 0.0111,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 14531.1955,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2308,
      "query_norm": 1.4161,
      "queue_k_norm": 1.4076,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0311,
      "sent_len_1": 66.7417,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5525,
      "stdk": 0.0476,
      "stdq": 0.0458,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 22300
    },
    {
      "accuracy": 50.6836,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6391,
      "doc_norm": 1.4131,
      "encoder_q-embeddings": 4512.4009,
      "encoder_q-layer.0": 3167.5908,
      "encoder_q-layer.1": 3646.2915,
      "encoder_q-layer.10": 1375.1631,
      "encoder_q-layer.11": 3260.7505,
      "encoder_q-layer.2": 4019.3982,
      "encoder_q-layer.3": 4284.8374,
      "encoder_q-layer.4": 4318.9541,
      "encoder_q-layer.5": 4306.3828,
      "encoder_q-layer.6": 4780.7871,
      "encoder_q-layer.7": 4829.311,
      "encoder_q-layer.8": 2718.4395,
      "encoder_q-layer.9": 1455.2354,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2024,
      "inbatch_pos_score": 0.8315,
      "learning_rate": 4.311111111111111e-05,
      "loss": 3.6391,
      "norm_diff": 0.0269,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5658.6522,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2029,
      "query_norm": 1.3862,
      "queue_k_norm": 1.4084,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9493,
      "sent_len_1": 66.7765,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.735,
      "stdk": 0.0478,
      "stdq": 0.0456,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 22400
    },
    {
      "accuracy": 48.1445,
      "active_queue_size": 16384.0,
      "cl_loss": 3.624,
      "doc_norm": 1.4048,
      "encoder_q-embeddings": 3292.4585,
      "encoder_q-layer.0": 2327.605,
      "encoder_q-layer.1": 2542.7039,
      "encoder_q-layer.10": 1378.4287,
      "encoder_q-layer.11": 3434.1226,
      "encoder_q-layer.2": 2738.8411,
      "encoder_q-layer.3": 2811.7319,
      "encoder_q-layer.4": 2820.4282,
      "encoder_q-layer.5": 2735.2168,
      "encoder_q-layer.6": 2652.7061,
      "encoder_q-layer.7": 2550.0569,
      "encoder_q-layer.8": 2021.9401,
      "encoder_q-layer.9": 1355.5892,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2081,
      "inbatch_pos_score": 0.7983,
      "learning_rate": 4.305555555555556e-05,
      "loss": 3.624,
      "norm_diff": 0.0128,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3924.3975,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.207,
      "query_norm": 1.392,
      "queue_k_norm": 1.4078,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8468,
      "sent_len_1": 66.5989,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.725,
      "stdk": 0.0475,
      "stdq": 0.0452,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 22500
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6243,
      "doc_norm": 1.4056,
      "encoder_q-embeddings": 1801.0687,
      "encoder_q-layer.0": 1231.5109,
      "encoder_q-layer.1": 1377.1641,
      "encoder_q-layer.10": 1331.8094,
      "encoder_q-layer.11": 3080.2461,
      "encoder_q-layer.2": 1466.7032,
      "encoder_q-layer.3": 1597.0383,
      "encoder_q-layer.4": 1783.7549,
      "encoder_q-layer.5": 1912.4165,
      "encoder_q-layer.6": 2223.6235,
      "encoder_q-layer.7": 2199.4973,
      "encoder_q-layer.8": 1922.5112,
      "encoder_q-layer.9": 1390.3849,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2007,
      "inbatch_pos_score": 0.8916,
      "learning_rate": 4.3e-05,
      "loss": 3.6243,
      "norm_diff": 0.0262,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2770.7319,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1996,
      "query_norm": 1.4318,
      "queue_k_norm": 1.4075,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6868,
      "sent_len_1": 66.4877,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8113,
      "stdk": 0.0476,
      "stdq": 0.0467,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 22600
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5978,
      "doc_norm": 1.4101,
      "encoder_q-embeddings": 2018.0083,
      "encoder_q-layer.0": 1432.0581,
      "encoder_q-layer.1": 1557.7783,
      "encoder_q-layer.10": 1268.5103,
      "encoder_q-layer.11": 3108.3628,
      "encoder_q-layer.2": 1681.326,
      "encoder_q-layer.3": 1780.8315,
      "encoder_q-layer.4": 1977.7126,
      "encoder_q-layer.5": 2017.4597,
      "encoder_q-layer.6": 2057.0273,
      "encoder_q-layer.7": 1892.4855,
      "encoder_q-layer.8": 1681.3557,
      "encoder_q-layer.9": 1314.4568,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2175,
      "inbatch_pos_score": 0.8467,
      "learning_rate": 4.294444444444445e-05,
      "loss": 3.5978,
      "norm_diff": 0.015,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2841.8286,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2169,
      "query_norm": 1.403,
      "queue_k_norm": 1.408,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8527,
      "sent_len_1": 66.9319,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4375,
      "stdk": 0.0478,
      "stdq": 0.0453,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 22700
    },
    {
      "accuracy": 50.6836,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6238,
      "doc_norm": 1.4009,
      "encoder_q-embeddings": 1844.6071,
      "encoder_q-layer.0": 1198.5017,
      "encoder_q-layer.1": 1314.2782,
      "encoder_q-layer.10": 1481.4857,
      "encoder_q-layer.11": 3119.3772,
      "encoder_q-layer.2": 1571.9701,
      "encoder_q-layer.3": 1825.545,
      "encoder_q-layer.4": 1963.4264,
      "encoder_q-layer.5": 1970.0953,
      "encoder_q-layer.6": 2174.064,
      "encoder_q-layer.7": 2512.6648,
      "encoder_q-layer.8": 2711.3445,
      "encoder_q-layer.9": 1902.424,
      "epoch": 0.22,
      "inbatch_neg_score": 0.1926,
      "inbatch_pos_score": 0.8345,
      "learning_rate": 4.2888888888888886e-05,
      "loss": 3.6238,
      "norm_diff": 0.0145,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2960.4967,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1937,
      "query_norm": 1.4146,
      "queue_k_norm": 1.4077,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1964,
      "sent_len_1": 66.9495,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4437,
      "stdk": 0.0475,
      "stdq": 0.0459,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 22800
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5799,
      "doc_norm": 1.4042,
      "encoder_q-embeddings": 2163.4851,
      "encoder_q-layer.0": 1412.8202,
      "encoder_q-layer.1": 1488.662,
      "encoder_q-layer.10": 1298.1719,
      "encoder_q-layer.11": 3003.1184,
      "encoder_q-layer.2": 1647.1156,
      "encoder_q-layer.3": 1711.807,
      "encoder_q-layer.4": 1814.46,
      "encoder_q-layer.5": 1762.7004,
      "encoder_q-layer.6": 1758.7183,
      "encoder_q-layer.7": 1697.3467,
      "encoder_q-layer.8": 1655.4149,
      "encoder_q-layer.9": 1330.0714,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2253,
      "inbatch_pos_score": 0.8638,
      "learning_rate": 4.2833333333333335e-05,
      "loss": 3.5799,
      "norm_diff": 0.0118,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2746.1556,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2263,
      "query_norm": 1.4147,
      "queue_k_norm": 1.4079,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9606,
      "sent_len_1": 66.8906,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9613,
      "stdk": 0.0476,
      "stdq": 0.0453,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 22900
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5935,
      "doc_norm": 1.4049,
      "encoder_q-embeddings": 3588.0007,
      "encoder_q-layer.0": 2472.6519,
      "encoder_q-layer.1": 2917.5098,
      "encoder_q-layer.10": 1236.9369,
      "encoder_q-layer.11": 3083.9822,
      "encoder_q-layer.2": 3453.0034,
      "encoder_q-layer.3": 3422.2983,
      "encoder_q-layer.4": 3741.229,
      "encoder_q-layer.5": 3745.3049,
      "encoder_q-layer.6": 3604.8086,
      "encoder_q-layer.7": 3647.1389,
      "encoder_q-layer.8": 2187.8438,
      "encoder_q-layer.9": 1391.3909,
      "epoch": 0.22,
      "inbatch_neg_score": 0.2323,
      "inbatch_pos_score": 0.8306,
      "learning_rate": 4.277777777777778e-05,
      "loss": 3.5935,
      "norm_diff": 0.0113,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4695.7255,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2311,
      "query_norm": 1.3943,
      "queue_k_norm": 1.4053,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0192,
      "sent_len_1": 66.9587,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.3413,
      "stdk": 0.0476,
      "stdq": 0.0447,
      "stdqueue_k": 0.0477,
      "stdqueue_q": 0.0,
      "step": 23000
    },
    {
      "accuracy": 51.1719,
      "active_queue_size": 16384.0,
      "cl_loss": 3.617,
      "doc_norm": 1.4062,
      "encoder_q-embeddings": 2197.012,
      "encoder_q-layer.0": 1520.8915,
      "encoder_q-layer.1": 1608.5641,
      "encoder_q-layer.10": 1254.6937,
      "encoder_q-layer.11": 3132.0479,
      "encoder_q-layer.2": 1853.3009,
      "encoder_q-layer.3": 1947.9946,
      "encoder_q-layer.4": 2009.8712,
      "encoder_q-layer.5": 1903.3256,
      "encoder_q-layer.6": 1923.4484,
      "encoder_q-layer.7": 1844.0907,
      "encoder_q-layer.8": 1763.6982,
      "encoder_q-layer.9": 1238.5104,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2278,
      "inbatch_pos_score": 0.833,
      "learning_rate": 4.272222222222223e-05,
      "loss": 3.617,
      "norm_diff": 0.0385,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2919.8752,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2266,
      "query_norm": 1.3676,
      "queue_k_norm": 1.4078,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8986,
      "sent_len_1": 66.6515,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0087,
      "stdk": 0.0476,
      "stdq": 0.044,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 23100
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5956,
      "doc_norm": 1.4121,
      "encoder_q-embeddings": 4292.5698,
      "encoder_q-layer.0": 3177.3362,
      "encoder_q-layer.1": 3719.3362,
      "encoder_q-layer.10": 1426.2882,
      "encoder_q-layer.11": 3540.4192,
      "encoder_q-layer.2": 3746.6934,
      "encoder_q-layer.3": 4066.0945,
      "encoder_q-layer.4": 4192.5449,
      "encoder_q-layer.5": 3953.0811,
      "encoder_q-layer.6": 3746.8706,
      "encoder_q-layer.7": 3528.5908,
      "encoder_q-layer.8": 2575.1982,
      "encoder_q-layer.9": 1451.1039,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2337,
      "inbatch_pos_score": 0.8501,
      "learning_rate": 4.266666666666667e-05,
      "loss": 3.5956,
      "norm_diff": 0.0212,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5249.3057,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2338,
      "query_norm": 1.391,
      "queue_k_norm": 1.4093,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8447,
      "sent_len_1": 66.7732,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2025,
      "stdk": 0.0479,
      "stdq": 0.0448,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 23200
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.574,
      "doc_norm": 1.3986,
      "encoder_q-embeddings": 2588.0908,
      "encoder_q-layer.0": 1838.6442,
      "encoder_q-layer.1": 1906.0819,
      "encoder_q-layer.10": 1294.6283,
      "encoder_q-layer.11": 3004.9729,
      "encoder_q-layer.2": 2200.6113,
      "encoder_q-layer.3": 2349.104,
      "encoder_q-layer.4": 2683.8206,
      "encoder_q-layer.5": 2586.3337,
      "encoder_q-layer.6": 2391.667,
      "encoder_q-layer.7": 2396.2837,
      "encoder_q-layer.8": 1961.424,
      "encoder_q-layer.9": 1391.3147,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2287,
      "inbatch_pos_score": 0.8594,
      "learning_rate": 4.261111111111111e-05,
      "loss": 3.574,
      "norm_diff": 0.0243,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3357.456,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2284,
      "query_norm": 1.4223,
      "queue_k_norm": 1.4078,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9872,
      "sent_len_1": 66.7806,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.58,
      "stdk": 0.0474,
      "stdq": 0.046,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 23300
    },
    {
      "accuracy": 48.0469,
      "active_queue_size": 16384.0,
      "cl_loss": 3.6014,
      "doc_norm": 1.4033,
      "encoder_q-embeddings": 3225.189,
      "encoder_q-layer.0": 2353.8491,
      "encoder_q-layer.1": 2622.6938,
      "encoder_q-layer.10": 1323.7844,
      "encoder_q-layer.11": 3080.4714,
      "encoder_q-layer.2": 3041.4473,
      "encoder_q-layer.3": 3037.9509,
      "encoder_q-layer.4": 3332.1294,
      "encoder_q-layer.5": 3574.7417,
      "encoder_q-layer.6": 3202.762,
      "encoder_q-layer.7": 3208.0903,
      "encoder_q-layer.8": 2330.0215,
      "encoder_q-layer.9": 1405.6007,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2262,
      "inbatch_pos_score": 0.7998,
      "learning_rate": 4.255555555555556e-05,
      "loss": 3.6014,
      "norm_diff": 0.026,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4205.5623,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.225,
      "query_norm": 1.3773,
      "queue_k_norm": 1.4061,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1183,
      "sent_len_1": 66.8537,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3113,
      "stdk": 0.0475,
      "stdq": 0.0442,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 23400
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5789,
      "doc_norm": 1.4135,
      "encoder_q-embeddings": 2982.509,
      "encoder_q-layer.0": 1970.8741,
      "encoder_q-layer.1": 2270.9568,
      "encoder_q-layer.10": 1333.4247,
      "encoder_q-layer.11": 2947.635,
      "encoder_q-layer.2": 2582.9915,
      "encoder_q-layer.3": 2717.551,
      "encoder_q-layer.4": 2472.6238,
      "encoder_q-layer.5": 2529.8774,
      "encoder_q-layer.6": 2836.8904,
      "encoder_q-layer.7": 2287.7085,
      "encoder_q-layer.8": 1776.8845,
      "encoder_q-layer.9": 1352.8851,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2302,
      "inbatch_pos_score": 0.8584,
      "learning_rate": 4.25e-05,
      "loss": 3.5789,
      "norm_diff": 0.0187,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3564.6937,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2307,
      "query_norm": 1.3948,
      "queue_k_norm": 1.4092,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9249,
      "sent_len_1": 66.7481,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8237,
      "stdk": 0.0479,
      "stdq": 0.0447,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 23500
    },
    {
      "accuracy": 49.8047,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5691,
      "doc_norm": 1.406,
      "encoder_q-embeddings": 2322.4875,
      "encoder_q-layer.0": 1586.5276,
      "encoder_q-layer.1": 1815.7781,
      "encoder_q-layer.10": 1397.8749,
      "encoder_q-layer.11": 2892.0986,
      "encoder_q-layer.2": 2174.4324,
      "encoder_q-layer.3": 2473.2898,
      "encoder_q-layer.4": 2439.3601,
      "encoder_q-layer.5": 2562.6155,
      "encoder_q-layer.6": 2922.8613,
      "encoder_q-layer.7": 3025.9265,
      "encoder_q-layer.8": 2345.8772,
      "encoder_q-layer.9": 1385.1841,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2242,
      "inbatch_pos_score": 0.8057,
      "learning_rate": 4.2444444444444445e-05,
      "loss": 3.5691,
      "norm_diff": 0.0124,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3425.8516,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2249,
      "query_norm": 1.3945,
      "queue_k_norm": 1.4073,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0805,
      "sent_len_1": 66.7718,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4025,
      "stdk": 0.0477,
      "stdq": 0.0445,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 23600
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5872,
      "doc_norm": 1.4068,
      "encoder_q-embeddings": 1530.6113,
      "encoder_q-layer.0": 997.1699,
      "encoder_q-layer.1": 1145.0341,
      "encoder_q-layer.10": 1497.0886,
      "encoder_q-layer.11": 2880.1621,
      "encoder_q-layer.2": 1351.1687,
      "encoder_q-layer.3": 1434.4486,
      "encoder_q-layer.4": 1521.6628,
      "encoder_q-layer.5": 1361.0128,
      "encoder_q-layer.6": 1434.2611,
      "encoder_q-layer.7": 1475.564,
      "encoder_q-layer.8": 1534.5137,
      "encoder_q-layer.9": 1284.0227,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2294,
      "inbatch_pos_score": 0.8701,
      "learning_rate": 4.238888888888889e-05,
      "loss": 3.5872,
      "norm_diff": 0.0246,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2330.3359,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2297,
      "query_norm": 1.4315,
      "queue_k_norm": 1.4073,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9906,
      "sent_len_1": 66.7462,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.28,
      "stdk": 0.0477,
      "stdq": 0.0455,
      "stdqueue_k": 0.0478,
      "stdqueue_q": 0.0,
      "step": 23700
    },
    {
      "accuracy": 51.0742,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5718,
      "doc_norm": 1.4085,
      "encoder_q-embeddings": 1305.0663,
      "encoder_q-layer.0": 926.1746,
      "encoder_q-layer.1": 959.4102,
      "encoder_q-layer.10": 1394.382,
      "encoder_q-layer.11": 2909.5505,
      "encoder_q-layer.2": 1084.5039,
      "encoder_q-layer.3": 1142.0828,
      "encoder_q-layer.4": 1216.4233,
      "encoder_q-layer.5": 1162.5623,
      "encoder_q-layer.6": 1271.6853,
      "encoder_q-layer.7": 1386.3175,
      "encoder_q-layer.8": 1512.6034,
      "encoder_q-layer.9": 1367.5359,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2302,
      "inbatch_pos_score": 0.8779,
      "learning_rate": 4.233333333333334e-05,
      "loss": 3.5718,
      "norm_diff": 0.038,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2117.3673,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2292,
      "query_norm": 1.4465,
      "queue_k_norm": 1.4077,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7617,
      "sent_len_1": 66.773,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2425,
      "stdk": 0.0478,
      "stdq": 0.046,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 23800
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5719,
      "doc_norm": 1.4078,
      "encoder_q-embeddings": 2297.7986,
      "encoder_q-layer.0": 1484.4125,
      "encoder_q-layer.1": 1737.5536,
      "encoder_q-layer.10": 1234.7692,
      "encoder_q-layer.11": 2938.458,
      "encoder_q-layer.2": 1996.7877,
      "encoder_q-layer.3": 2198.085,
      "encoder_q-layer.4": 2520.0894,
      "encoder_q-layer.5": 2688.6724,
      "encoder_q-layer.6": 2615.0977,
      "encoder_q-layer.7": 2601.2397,
      "encoder_q-layer.8": 2557.99,
      "encoder_q-layer.9": 1725.6864,
      "epoch": 0.23,
      "inbatch_neg_score": 0.2295,
      "inbatch_pos_score": 0.8613,
      "learning_rate": 4.227777777777778e-05,
      "loss": 3.5719,
      "norm_diff": 0.0078,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3378.1722,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2286,
      "query_norm": 1.4066,
      "queue_k_norm": 1.4079,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.667,
      "sent_len_1": 66.6614,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4837,
      "stdk": 0.0478,
      "stdq": 0.045,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 23900
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5713,
      "doc_norm": 1.4032,
      "encoder_q-embeddings": 3459.79,
      "encoder_q-layer.0": 2206.8093,
      "encoder_q-layer.1": 2320.2864,
      "encoder_q-layer.10": 2518.967,
      "encoder_q-layer.11": 6143.6133,
      "encoder_q-layer.2": 2527.5444,
      "encoder_q-layer.3": 2636.3894,
      "encoder_q-layer.4": 2908.0276,
      "encoder_q-layer.5": 2858.7522,
      "encoder_q-layer.6": 3202.9019,
      "encoder_q-layer.7": 3232.6675,
      "encoder_q-layer.8": 3353.6975,
      "encoder_q-layer.9": 2585.4253,
      "epoch": 0.23,
      "inbatch_neg_score": 0.224,
      "inbatch_pos_score": 0.8574,
      "learning_rate": 4.222222222222222e-05,
      "loss": 3.5713,
      "norm_diff": 0.0111,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4963.7615,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.222,
      "query_norm": 1.4104,
      "queue_k_norm": 1.4093,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9125,
      "sent_len_1": 66.886,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2688,
      "stdk": 0.0476,
      "stdq": 0.0453,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24000
    },
    {
      "accuracy": 48.6328,
      "active_queue_size": 16384.0,
      "cl_loss": 3.547,
      "doc_norm": 1.4109,
      "encoder_q-embeddings": 10265.2842,
      "encoder_q-layer.0": 6931.9639,
      "encoder_q-layer.1": 7056.2603,
      "encoder_q-layer.10": 2582.1152,
      "encoder_q-layer.11": 5954.2788,
      "encoder_q-layer.2": 8804.9834,
      "encoder_q-layer.3": 8914.7148,
      "encoder_q-layer.4": 10310.4248,
      "encoder_q-layer.5": 9830.8418,
      "encoder_q-layer.6": 10915.6221,
      "encoder_q-layer.7": 9890.5537,
      "encoder_q-layer.8": 6952.0151,
      "encoder_q-layer.9": 3098.6875,
      "epoch": 0.24,
      "inbatch_neg_score": 0.2213,
      "inbatch_pos_score": 0.8408,
      "learning_rate": 4.216666666666667e-05,
      "loss": 3.547,
      "norm_diff": 0.0195,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12456.7658,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2203,
      "query_norm": 1.3922,
      "queue_k_norm": 1.4092,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0569,
      "sent_len_1": 66.9633,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2575,
      "stdk": 0.0479,
      "stdq": 0.0447,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24100
    },
    {
      "accuracy": 47.2656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.566,
      "doc_norm": 1.4085,
      "encoder_q-embeddings": 8513.6787,
      "encoder_q-layer.0": 6656.6738,
      "encoder_q-layer.1": 6399.8848,
      "encoder_q-layer.10": 1623.7386,
      "encoder_q-layer.11": 3324.2686,
      "encoder_q-layer.2": 6104.1343,
      "encoder_q-layer.3": 6016.7144,
      "encoder_q-layer.4": 5937.2998,
      "encoder_q-layer.5": 5727.7954,
      "encoder_q-layer.6": 4062.3701,
      "encoder_q-layer.7": 2896.2278,
      "encoder_q-layer.8": 2190.9924,
      "encoder_q-layer.9": 1576.1761,
      "epoch": 0.24,
      "inbatch_neg_score": 0.206,
      "inbatch_pos_score": 0.8247,
      "learning_rate": 4.211111111111111e-05,
      "loss": 3.566,
      "norm_diff": 0.014,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7941.0816,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.2064,
      "query_norm": 1.4034,
      "queue_k_norm": 1.4092,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0568,
      "sent_len_1": 66.8877,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0775,
      "stdk": 0.0478,
      "stdq": 0.0458,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 24200
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5659,
      "doc_norm": 1.4125,
      "encoder_q-embeddings": 1705.1709,
      "encoder_q-layer.0": 1154.1046,
      "encoder_q-layer.1": 1259.8827,
      "encoder_q-layer.10": 1307.4722,
      "encoder_q-layer.11": 2806.2351,
      "encoder_q-layer.2": 1404.4512,
      "encoder_q-layer.3": 1311.0679,
      "encoder_q-layer.4": 1433.3123,
      "encoder_q-layer.5": 1523.8708,
      "encoder_q-layer.6": 1675.7119,
      "encoder_q-layer.7": 1644.9497,
      "encoder_q-layer.8": 1620.5249,
      "encoder_q-layer.9": 1287.3176,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1884,
      "inbatch_pos_score": 0.8359,
      "learning_rate": 4.205555555555556e-05,
      "loss": 3.5659,
      "norm_diff": 0.0229,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2361.0204,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1882,
      "query_norm": 1.3896,
      "queue_k_norm": 1.4077,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0042,
      "sent_len_1": 66.7976,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1925,
      "stdk": 0.048,
      "stdq": 0.0457,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24300
    },
    {
      "accuracy": 47.9492,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5891,
      "doc_norm": 1.4082,
      "encoder_q-embeddings": 1310.2006,
      "encoder_q-layer.0": 900.2599,
      "encoder_q-layer.1": 944.9661,
      "encoder_q-layer.10": 1288.2823,
      "encoder_q-layer.11": 2969.2017,
      "encoder_q-layer.2": 1041.3353,
      "encoder_q-layer.3": 1060.3147,
      "encoder_q-layer.4": 1124.5089,
      "encoder_q-layer.5": 1156.1047,
      "encoder_q-layer.6": 1305.1735,
      "encoder_q-layer.7": 1443.01,
      "encoder_q-layer.8": 1549.918,
      "encoder_q-layer.9": 1233.9718,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1854,
      "inbatch_pos_score": 0.7827,
      "learning_rate": 4.2e-05,
      "loss": 3.5891,
      "norm_diff": 0.0581,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2167.5711,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1857,
      "query_norm": 1.3501,
      "queue_k_norm": 1.4071,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9818,
      "sent_len_1": 66.6916,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6262,
      "stdk": 0.0479,
      "stdq": 0.0445,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 24400
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5632,
      "doc_norm": 1.4044,
      "encoder_q-embeddings": 1926.8082,
      "encoder_q-layer.0": 1424.1288,
      "encoder_q-layer.1": 1579.7301,
      "encoder_q-layer.10": 1373.3003,
      "encoder_q-layer.11": 2889.9219,
      "encoder_q-layer.2": 1683.0625,
      "encoder_q-layer.3": 1703.1108,
      "encoder_q-layer.4": 1688.6277,
      "encoder_q-layer.5": 1806.5778,
      "encoder_q-layer.6": 1921.9536,
      "encoder_q-layer.7": 1802.0238,
      "encoder_q-layer.8": 1764.7095,
      "encoder_q-layer.9": 1354.8308,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1963,
      "inbatch_pos_score": 0.8472,
      "learning_rate": 4.194444444444445e-05,
      "loss": 3.5632,
      "norm_diff": 0.0234,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2670.3085,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1963,
      "query_norm": 1.381,
      "queue_k_norm": 1.4059,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0131,
      "sent_len_1": 66.8902,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7413,
      "stdk": 0.0478,
      "stdq": 0.0454,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24500
    },
    {
      "accuracy": 49.4141,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5831,
      "doc_norm": 1.4047,
      "encoder_q-embeddings": 1922.0703,
      "encoder_q-layer.0": 1349.8608,
      "encoder_q-layer.1": 1329.2041,
      "encoder_q-layer.10": 1269.2135,
      "encoder_q-layer.11": 2883.9678,
      "encoder_q-layer.2": 1611.4182,
      "encoder_q-layer.3": 1458.2748,
      "encoder_q-layer.4": 1543.2598,
      "encoder_q-layer.5": 1542.1842,
      "encoder_q-layer.6": 1807.6936,
      "encoder_q-layer.7": 1563.0177,
      "encoder_q-layer.8": 1488.6674,
      "encoder_q-layer.9": 1249.9965,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1957,
      "inbatch_pos_score": 0.7939,
      "learning_rate": 4.188888888888889e-05,
      "loss": 3.5831,
      "norm_diff": 0.0546,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2539.4154,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1957,
      "query_norm": 1.3501,
      "queue_k_norm": 1.4051,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8895,
      "sent_len_1": 66.351,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.19,
      "stdk": 0.0478,
      "stdq": 0.044,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24600
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5733,
      "doc_norm": 1.405,
      "encoder_q-embeddings": 934.355,
      "encoder_q-layer.0": 572.7925,
      "encoder_q-layer.1": 603.4168,
      "encoder_q-layer.10": 622.5825,
      "encoder_q-layer.11": 1333.3092,
      "encoder_q-layer.2": 655.3574,
      "encoder_q-layer.3": 708.8694,
      "encoder_q-layer.4": 684.6823,
      "encoder_q-layer.5": 682.7202,
      "encoder_q-layer.6": 741.801,
      "encoder_q-layer.7": 788.8773,
      "encoder_q-layer.8": 828.6475,
      "encoder_q-layer.9": 638.4784,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1788,
      "inbatch_pos_score": 0.8184,
      "learning_rate": 4.183333333333334e-05,
      "loss": 3.5733,
      "norm_diff": 0.027,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1171.0291,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1792,
      "query_norm": 1.378,
      "queue_k_norm": 1.4063,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8187,
      "sent_len_1": 66.7158,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.71,
      "stdk": 0.0479,
      "stdq": 0.0456,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 24700
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5459,
      "doc_norm": 1.4034,
      "encoder_q-embeddings": 1033.2865,
      "encoder_q-layer.0": 667.8786,
      "encoder_q-layer.1": 755.6133,
      "encoder_q-layer.10": 689.7005,
      "encoder_q-layer.11": 1436.0685,
      "encoder_q-layer.2": 870.5583,
      "encoder_q-layer.3": 900.4341,
      "encoder_q-layer.4": 961.5785,
      "encoder_q-layer.5": 995.0579,
      "encoder_q-layer.6": 973.025,
      "encoder_q-layer.7": 1071.7722,
      "encoder_q-layer.8": 859.6398,
      "encoder_q-layer.9": 630.2768,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1761,
      "inbatch_pos_score": 0.7832,
      "learning_rate": 4.177777777777778e-05,
      "loss": 3.5459,
      "norm_diff": 0.0504,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1386.0328,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1771,
      "query_norm": 1.353,
      "queue_k_norm": 1.4018,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0274,
      "sent_len_1": 66.858,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.75,
      "stdk": 0.0479,
      "stdq": 0.0447,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24800
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5496,
      "doc_norm": 1.4007,
      "encoder_q-embeddings": 882.5779,
      "encoder_q-layer.0": 597.355,
      "encoder_q-layer.1": 694.0412,
      "encoder_q-layer.10": 590.1336,
      "encoder_q-layer.11": 1362.5961,
      "encoder_q-layer.2": 772.8394,
      "encoder_q-layer.3": 822.7858,
      "encoder_q-layer.4": 971.8295,
      "encoder_q-layer.5": 926.5442,
      "encoder_q-layer.6": 1037.9303,
      "encoder_q-layer.7": 1089.0382,
      "encoder_q-layer.8": 1127.6836,
      "encoder_q-layer.9": 787.0945,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1808,
      "inbatch_pos_score": 0.8472,
      "learning_rate": 4.172222222222222e-05,
      "loss": 3.5496,
      "norm_diff": 0.0044,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1360.8163,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.181,
      "query_norm": 1.3989,
      "queue_k_norm": 1.4018,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9363,
      "sent_len_1": 66.7553,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.3713,
      "stdk": 0.0478,
      "stdq": 0.0463,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 24900
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5697,
      "doc_norm": 1.3977,
      "encoder_q-embeddings": 816.1497,
      "encoder_q-layer.0": 561.1924,
      "encoder_q-layer.1": 620.3623,
      "encoder_q-layer.10": 621.9829,
      "encoder_q-layer.11": 1432.0073,
      "encoder_q-layer.2": 713.9332,
      "encoder_q-layer.3": 691.9762,
      "encoder_q-layer.4": 697.6194,
      "encoder_q-layer.5": 711.0004,
      "encoder_q-layer.6": 681.537,
      "encoder_q-layer.7": 680.4737,
      "encoder_q-layer.8": 699.9769,
      "encoder_q-layer.9": 622.3752,
      "epoch": 0.24,
      "inbatch_neg_score": 0.1753,
      "inbatch_pos_score": 0.8062,
      "learning_rate": 4.166666666666667e-05,
      "loss": 3.5697,
      "norm_diff": 0.0316,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1156.4828,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1748,
      "query_norm": 1.3661,
      "queue_k_norm": 1.4016,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.28,
      "sent_len_1": 66.9843,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1625,
      "stdk": 0.0477,
      "stdq": 0.045,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 25000
    },
    {
      "accuracy": 52.9297,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5112,
      "doc_norm": 1.3986,
      "encoder_q-embeddings": 6851.2241,
      "encoder_q-layer.0": 4937.1279,
      "encoder_q-layer.1": 5628.4907,
      "encoder_q-layer.10": 632.2959,
      "encoder_q-layer.11": 1429.4181,
      "encoder_q-layer.2": 6969.8579,
      "encoder_q-layer.3": 7158.3276,
      "encoder_q-layer.4": 8189.5679,
      "encoder_q-layer.5": 8145.7334,
      "encoder_q-layer.6": 8363.2236,
      "encoder_q-layer.7": 9160.4727,
      "encoder_q-layer.8": 4204.3154,
      "encoder_q-layer.9": 1010.9896,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1741,
      "inbatch_pos_score": 0.8037,
      "learning_rate": 4.1611111111111114e-05,
      "loss": 3.5112,
      "norm_diff": 0.0273,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9431.4961,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.1735,
      "query_norm": 1.3713,
      "queue_k_norm": 1.4025,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8112,
      "sent_len_1": 66.8377,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9938,
      "stdk": 0.0478,
      "stdq": 0.0453,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 25100
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5497,
      "doc_norm": 1.4029,
      "encoder_q-embeddings": 788.2953,
      "encoder_q-layer.0": 525.5163,
      "encoder_q-layer.1": 597.5987,
      "encoder_q-layer.10": 579.0491,
      "encoder_q-layer.11": 1321.7043,
      "encoder_q-layer.2": 630.2535,
      "encoder_q-layer.3": 649.7164,
      "encoder_q-layer.4": 649.5215,
      "encoder_q-layer.5": 603.5771,
      "encoder_q-layer.6": 657.5274,
      "encoder_q-layer.7": 720.9599,
      "encoder_q-layer.8": 746.6126,
      "encoder_q-layer.9": 631.1224,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1782,
      "inbatch_pos_score": 0.8364,
      "learning_rate": 4.155555555555556e-05,
      "loss": 3.5497,
      "norm_diff": 0.007,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1089.1405,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1774,
      "query_norm": 1.3991,
      "queue_k_norm": 1.3997,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.013,
      "sent_len_1": 66.7933,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.225,
      "stdk": 0.048,
      "stdq": 0.046,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 25200
    },
    {
      "accuracy": 47.168,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5143,
      "doc_norm": 1.3998,
      "encoder_q-embeddings": 1690.8929,
      "encoder_q-layer.0": 1224.5664,
      "encoder_q-layer.1": 1393.4215,
      "encoder_q-layer.10": 655.324,
      "encoder_q-layer.11": 1431.0282,
      "encoder_q-layer.2": 1676.6263,
      "encoder_q-layer.3": 1694.9559,
      "encoder_q-layer.4": 1643.9779,
      "encoder_q-layer.5": 1625.0046,
      "encoder_q-layer.6": 1617.3436,
      "encoder_q-layer.7": 1331.283,
      "encoder_q-layer.8": 903.6763,
      "encoder_q-layer.9": 647.5179,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1669,
      "inbatch_pos_score": 0.7578,
      "learning_rate": 4.15e-05,
      "loss": 3.5143,
      "norm_diff": 0.0331,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2098.1811,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1671,
      "query_norm": 1.3667,
      "queue_k_norm": 1.3989,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9039,
      "sent_len_1": 66.9086,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7937,
      "stdk": 0.0479,
      "stdq": 0.0449,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 25300
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5273,
      "doc_norm": 1.3944,
      "encoder_q-embeddings": 653.9332,
      "encoder_q-layer.0": 441.3697,
      "encoder_q-layer.1": 484.5655,
      "encoder_q-layer.10": 667.628,
      "encoder_q-layer.11": 1353.0951,
      "encoder_q-layer.2": 571.1337,
      "encoder_q-layer.3": 636.2188,
      "encoder_q-layer.4": 754.7802,
      "encoder_q-layer.5": 709.5286,
      "encoder_q-layer.6": 682.3445,
      "encoder_q-layer.7": 725.6863,
      "encoder_q-layer.8": 773.0938,
      "encoder_q-layer.9": 686.7794,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1608,
      "inbatch_pos_score": 0.7554,
      "learning_rate": 4.144444444444445e-05,
      "loss": 3.5273,
      "norm_diff": 0.0377,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1087.5457,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1599,
      "query_norm": 1.3568,
      "queue_k_norm": 1.3998,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9387,
      "sent_len_1": 66.8254,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7775,
      "stdk": 0.0477,
      "stdq": 0.0447,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 25400
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5518,
      "doc_norm": 1.3994,
      "encoder_q-embeddings": 607.7343,
      "encoder_q-layer.0": 399.4394,
      "encoder_q-layer.1": 422.9692,
      "encoder_q-layer.10": 594.6741,
      "encoder_q-layer.11": 1324.2738,
      "encoder_q-layer.2": 459.2203,
      "encoder_q-layer.3": 462.6089,
      "encoder_q-layer.4": 519.4867,
      "encoder_q-layer.5": 520.1624,
      "encoder_q-layer.6": 578.7879,
      "encoder_q-layer.7": 624.2927,
      "encoder_q-layer.8": 705.2097,
      "encoder_q-layer.9": 600.8412,
      "epoch": 0.25,
      "inbatch_neg_score": 0.164,
      "inbatch_pos_score": 0.812,
      "learning_rate": 4.138888888888889e-05,
      "loss": 3.5518,
      "norm_diff": 0.0234,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 967.6744,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1642,
      "query_norm": 1.3761,
      "queue_k_norm": 1.3968,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9204,
      "sent_len_1": 66.5848,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5275,
      "stdk": 0.0479,
      "stdq": 0.0454,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 25500
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5581,
      "doc_norm": 1.3931,
      "encoder_q-embeddings": 1234.7136,
      "encoder_q-layer.0": 887.1579,
      "encoder_q-layer.1": 1085.578,
      "encoder_q-layer.10": 611.2844,
      "encoder_q-layer.11": 1422.4746,
      "encoder_q-layer.2": 1228.5321,
      "encoder_q-layer.3": 1385.6758,
      "encoder_q-layer.4": 1379.2295,
      "encoder_q-layer.5": 1512.0835,
      "encoder_q-layer.6": 1446.1877,
      "encoder_q-layer.7": 1173.692,
      "encoder_q-layer.8": 884.6814,
      "encoder_q-layer.9": 667.8317,
      "epoch": 0.25,
      "inbatch_neg_score": 0.151,
      "inbatch_pos_score": 0.79,
      "learning_rate": 4.133333333333333e-05,
      "loss": 3.5581,
      "norm_diff": 0.0315,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1734.5822,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1517,
      "query_norm": 1.3616,
      "queue_k_norm": 1.3965,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9664,
      "sent_len_1": 66.6213,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8063,
      "stdk": 0.0478,
      "stdq": 0.0452,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 25600
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5375,
      "doc_norm": 1.3983,
      "encoder_q-embeddings": 671.6667,
      "encoder_q-layer.0": 464.782,
      "encoder_q-layer.1": 487.2064,
      "encoder_q-layer.10": 674.0069,
      "encoder_q-layer.11": 1403.1613,
      "encoder_q-layer.2": 541.2934,
      "encoder_q-layer.3": 544.9177,
      "encoder_q-layer.4": 608.1513,
      "encoder_q-layer.5": 600.2668,
      "encoder_q-layer.6": 699.0875,
      "encoder_q-layer.7": 746.1172,
      "encoder_q-layer.8": 803.3029,
      "encoder_q-layer.9": 656.0577,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1585,
      "inbatch_pos_score": 0.7832,
      "learning_rate": 4.127777777777778e-05,
      "loss": 3.5375,
      "norm_diff": 0.0574,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1078.8687,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1589,
      "query_norm": 1.3409,
      "queue_k_norm": 1.393,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9656,
      "sent_len_1": 66.5649,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8,
      "stdk": 0.048,
      "stdq": 0.0445,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 25700
    },
    {
      "accuracy": 50.4883,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5403,
      "doc_norm": 1.3972,
      "encoder_q-embeddings": 908.3321,
      "encoder_q-layer.0": 682.1144,
      "encoder_q-layer.1": 686.8648,
      "encoder_q-layer.10": 648.6836,
      "encoder_q-layer.11": 1399.6937,
      "encoder_q-layer.2": 758.0508,
      "encoder_q-layer.3": 749.5003,
      "encoder_q-layer.4": 738.66,
      "encoder_q-layer.5": 680.2478,
      "encoder_q-layer.6": 728.2859,
      "encoder_q-layer.7": 751.4062,
      "encoder_q-layer.8": 798.1392,
      "encoder_q-layer.9": 681.0893,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1629,
      "inbatch_pos_score": 0.7783,
      "learning_rate": 4.1222222222222224e-05,
      "loss": 3.5403,
      "norm_diff": 0.0497,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1225.7888,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1625,
      "query_norm": 1.3475,
      "queue_k_norm": 1.393,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9617,
      "sent_len_1": 66.7989,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6037,
      "stdk": 0.048,
      "stdq": 0.0448,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 25800
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5307,
      "doc_norm": 1.3923,
      "encoder_q-embeddings": 1150.6343,
      "encoder_q-layer.0": 750.9628,
      "encoder_q-layer.1": 877.5085,
      "encoder_q-layer.10": 636.1991,
      "encoder_q-layer.11": 1359.3192,
      "encoder_q-layer.2": 964.3879,
      "encoder_q-layer.3": 1047.7778,
      "encoder_q-layer.4": 1176.2489,
      "encoder_q-layer.5": 1202.72,
      "encoder_q-layer.6": 1317.64,
      "encoder_q-layer.7": 1151.1555,
      "encoder_q-layer.8": 890.9929,
      "encoder_q-layer.9": 693.3546,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1617,
      "inbatch_pos_score": 0.8242,
      "learning_rate": 4.116666666666667e-05,
      "loss": 3.5307,
      "norm_diff": 0.0312,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1564.3005,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1606,
      "query_norm": 1.3611,
      "queue_k_norm": 1.3925,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9629,
      "sent_len_1": 66.7961,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9787,
      "stdk": 0.0478,
      "stdq": 0.0452,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 25900
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5221,
      "doc_norm": 1.3913,
      "encoder_q-embeddings": 834.9116,
      "encoder_q-layer.0": 574.2308,
      "encoder_q-layer.1": 631.3254,
      "encoder_q-layer.10": 608.8398,
      "encoder_q-layer.11": 1285.1926,
      "encoder_q-layer.2": 687.5097,
      "encoder_q-layer.3": 701.9085,
      "encoder_q-layer.4": 754.2084,
      "encoder_q-layer.5": 735.2666,
      "encoder_q-layer.6": 820.6118,
      "encoder_q-layer.7": 791.1491,
      "encoder_q-layer.8": 807.9592,
      "encoder_q-layer.9": 656.1193,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1579,
      "inbatch_pos_score": 0.8101,
      "learning_rate": 4.111111111111111e-05,
      "loss": 3.5221,
      "norm_diff": 0.0284,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1171.2132,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1572,
      "query_norm": 1.3629,
      "queue_k_norm": 1.3921,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9771,
      "sent_len_1": 66.8276,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4863,
      "stdk": 0.0478,
      "stdq": 0.0455,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26000
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5376,
      "doc_norm": 1.3916,
      "encoder_q-embeddings": 1018.6212,
      "encoder_q-layer.0": 676.7032,
      "encoder_q-layer.1": 713.3555,
      "encoder_q-layer.10": 689.7332,
      "encoder_q-layer.11": 1308.3473,
      "encoder_q-layer.2": 770.7841,
      "encoder_q-layer.3": 837.4434,
      "encoder_q-layer.4": 894.8958,
      "encoder_q-layer.5": 836.6804,
      "encoder_q-layer.6": 934.0964,
      "encoder_q-layer.7": 911.0344,
      "encoder_q-layer.8": 1053.7948,
      "encoder_q-layer.9": 751.9384,
      "epoch": 0.25,
      "inbatch_neg_score": 0.1622,
      "inbatch_pos_score": 0.8071,
      "learning_rate": 4.105555555555556e-05,
      "loss": 3.5376,
      "norm_diff": 0.0396,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1340.5905,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1626,
      "query_norm": 1.352,
      "queue_k_norm": 1.3918,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9682,
      "sent_len_1": 66.6759,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1538,
      "stdk": 0.0479,
      "stdq": 0.0452,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26100
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5361,
      "doc_norm": 1.3917,
      "encoder_q-embeddings": 1201.308,
      "encoder_q-layer.0": 809.2301,
      "encoder_q-layer.1": 784.938,
      "encoder_q-layer.10": 608.929,
      "encoder_q-layer.11": 1336.6559,
      "encoder_q-layer.2": 842.9503,
      "encoder_q-layer.3": 867.3347,
      "encoder_q-layer.4": 880.0499,
      "encoder_q-layer.5": 861.022,
      "encoder_q-layer.6": 987.8536,
      "encoder_q-layer.7": 1099.7781,
      "encoder_q-layer.8": 1221.1567,
      "encoder_q-layer.9": 942.696,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1623,
      "inbatch_pos_score": 0.7759,
      "learning_rate": 4.1e-05,
      "loss": 3.5361,
      "norm_diff": 0.0416,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1489.5752,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1615,
      "query_norm": 1.3501,
      "queue_k_norm": 1.3912,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6091,
      "sent_len_1": 66.5303,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8325,
      "stdk": 0.0479,
      "stdq": 0.045,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26200
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5181,
      "doc_norm": 1.3955,
      "encoder_q-embeddings": 822.0681,
      "encoder_q-layer.0": 520.69,
      "encoder_q-layer.1": 568.8849,
      "encoder_q-layer.10": 657.7006,
      "encoder_q-layer.11": 1342.9023,
      "encoder_q-layer.2": 658.5743,
      "encoder_q-layer.3": 730.9957,
      "encoder_q-layer.4": 874.441,
      "encoder_q-layer.5": 812.2322,
      "encoder_q-layer.6": 869.7023,
      "encoder_q-layer.7": 951.3959,
      "encoder_q-layer.8": 990.9709,
      "encoder_q-layer.9": 762.0366,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1554,
      "inbatch_pos_score": 0.7988,
      "learning_rate": 4.094444444444445e-05,
      "loss": 3.5181,
      "norm_diff": 0.0378,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1227.6262,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1558,
      "query_norm": 1.3577,
      "queue_k_norm": 1.3918,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9338,
      "sent_len_1": 66.7686,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.175,
      "stdk": 0.0481,
      "stdq": 0.0451,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26300
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5474,
      "doc_norm": 1.3925,
      "encoder_q-embeddings": 434.8495,
      "encoder_q-layer.0": 275.7167,
      "encoder_q-layer.1": 291.1348,
      "encoder_q-layer.10": 299.6149,
      "encoder_q-layer.11": 671.0661,
      "encoder_q-layer.2": 341.0104,
      "encoder_q-layer.3": 346.9167,
      "encoder_q-layer.4": 369.6541,
      "encoder_q-layer.5": 394.1975,
      "encoder_q-layer.6": 431.6562,
      "encoder_q-layer.7": 443.7527,
      "encoder_q-layer.8": 409.6902,
      "encoder_q-layer.9": 323.8975,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1695,
      "inbatch_pos_score": 0.7949,
      "learning_rate": 4.088888888888889e-05,
      "loss": 3.5474,
      "norm_diff": 0.0316,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 593.6864,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1689,
      "query_norm": 1.3609,
      "queue_k_norm": 1.3906,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0958,
      "sent_len_1": 66.8328,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7438,
      "stdk": 0.0479,
      "stdq": 0.045,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26400
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5248,
      "doc_norm": 1.3958,
      "encoder_q-embeddings": 731.1332,
      "encoder_q-layer.0": 528.668,
      "encoder_q-layer.1": 526.669,
      "encoder_q-layer.10": 365.2702,
      "encoder_q-layer.11": 734.7773,
      "encoder_q-layer.2": 653.6866,
      "encoder_q-layer.3": 650.3055,
      "encoder_q-layer.4": 704.3323,
      "encoder_q-layer.5": 864.8784,
      "encoder_q-layer.6": 759.0584,
      "encoder_q-layer.7": 774.3522,
      "encoder_q-layer.8": 663.5508,
      "encoder_q-layer.9": 372.2983,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1708,
      "inbatch_pos_score": 0.7656,
      "learning_rate": 4.0833333333333334e-05,
      "loss": 3.5248,
      "norm_diff": 0.0657,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 978.9363,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1697,
      "query_norm": 1.3301,
      "queue_k_norm": 1.3918,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.972,
      "sent_len_1": 66.67,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.3562,
      "stdk": 0.0481,
      "stdq": 0.044,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26500
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5085,
      "doc_norm": 1.3974,
      "encoder_q-embeddings": 2590.9241,
      "encoder_q-layer.0": 1648.0104,
      "encoder_q-layer.1": 2022.7517,
      "encoder_q-layer.10": 318.1642,
      "encoder_q-layer.11": 650.1288,
      "encoder_q-layer.2": 2254.6875,
      "encoder_q-layer.3": 2482.2295,
      "encoder_q-layer.4": 2882.1702,
      "encoder_q-layer.5": 2908.5691,
      "encoder_q-layer.6": 2847.249,
      "encoder_q-layer.7": 2696.9863,
      "encoder_q-layer.8": 1828.0314,
      "encoder_q-layer.9": 494.6945,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1702,
      "inbatch_pos_score": 0.8232,
      "learning_rate": 4.0777777777777783e-05,
      "loss": 3.5085,
      "norm_diff": 0.0285,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3278.4848,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1703,
      "query_norm": 1.3688,
      "queue_k_norm": 1.3902,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.926,
      "sent_len_1": 66.7146,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9025,
      "stdk": 0.0482,
      "stdq": 0.0452,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26600
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4935,
      "doc_norm": 1.392,
      "encoder_q-embeddings": 294.8237,
      "encoder_q-layer.0": 189.7005,
      "encoder_q-layer.1": 200.5098,
      "encoder_q-layer.10": 294.4138,
      "encoder_q-layer.11": 637.7573,
      "encoder_q-layer.2": 222.725,
      "encoder_q-layer.3": 235.5561,
      "encoder_q-layer.4": 249.1794,
      "encoder_q-layer.5": 255.2377,
      "encoder_q-layer.6": 287.7271,
      "encoder_q-layer.7": 313.5806,
      "encoder_q-layer.8": 353.6754,
      "encoder_q-layer.9": 301.3709,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1744,
      "inbatch_pos_score": 0.793,
      "learning_rate": 4.0722222222222226e-05,
      "loss": 3.4935,
      "norm_diff": 0.0458,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 471.8889,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1749,
      "query_norm": 1.3462,
      "queue_k_norm": 1.3904,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8871,
      "sent_len_1": 66.6826,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.375,
      "stdk": 0.0479,
      "stdq": 0.0444,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26700
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5497,
      "doc_norm": 1.3913,
      "encoder_q-embeddings": 346.0812,
      "encoder_q-layer.0": 230.4144,
      "encoder_q-layer.1": 247.5842,
      "encoder_q-layer.10": 304.1868,
      "encoder_q-layer.11": 661.2893,
      "encoder_q-layer.2": 280.2791,
      "encoder_q-layer.3": 289.0876,
      "encoder_q-layer.4": 314.1222,
      "encoder_q-layer.5": 316.5557,
      "encoder_q-layer.6": 357.868,
      "encoder_q-layer.7": 369.4095,
      "encoder_q-layer.8": 389.0192,
      "encoder_q-layer.9": 327.0127,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1704,
      "inbatch_pos_score": 0.7646,
      "learning_rate": 4.066666666666667e-05,
      "loss": 3.5497,
      "norm_diff": 0.0419,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 526.4968,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1693,
      "query_norm": 1.3494,
      "queue_k_norm": 1.3919,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1841,
      "sent_len_1": 66.3981,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.99,
      "stdk": 0.0479,
      "stdq": 0.0448,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26800
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4855,
      "doc_norm": 1.395,
      "encoder_q-embeddings": 415.6185,
      "encoder_q-layer.0": 262.8745,
      "encoder_q-layer.1": 287.9299,
      "encoder_q-layer.10": 323.6875,
      "encoder_q-layer.11": 714.2651,
      "encoder_q-layer.2": 340.694,
      "encoder_q-layer.3": 377.3113,
      "encoder_q-layer.4": 388.7909,
      "encoder_q-layer.5": 366.0081,
      "encoder_q-layer.6": 406.4124,
      "encoder_q-layer.7": 453.693,
      "encoder_q-layer.8": 509.3709,
      "encoder_q-layer.9": 395.7092,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1764,
      "inbatch_pos_score": 0.7886,
      "learning_rate": 4.061111111111111e-05,
      "loss": 3.4855,
      "norm_diff": 0.027,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 620.4213,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1763,
      "query_norm": 1.368,
      "queue_k_norm": 1.392,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9759,
      "sent_len_1": 66.8957,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.78,
      "stdk": 0.0481,
      "stdq": 0.0451,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 26900
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5075,
      "doc_norm": 1.3987,
      "encoder_q-embeddings": 422.9335,
      "encoder_q-layer.0": 283.9524,
      "encoder_q-layer.1": 294.6614,
      "encoder_q-layer.10": 344.2137,
      "encoder_q-layer.11": 693.5967,
      "encoder_q-layer.2": 334.8974,
      "encoder_q-layer.3": 353.0682,
      "encoder_q-layer.4": 389.657,
      "encoder_q-layer.5": 374.6926,
      "encoder_q-layer.6": 418.3292,
      "encoder_q-layer.7": 438.0147,
      "encoder_q-layer.8": 440.9618,
      "encoder_q-layer.9": 358.6366,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1757,
      "inbatch_pos_score": 0.8105,
      "learning_rate": 4.055555555555556e-05,
      "loss": 3.5075,
      "norm_diff": 0.0257,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 600.4656,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.176,
      "query_norm": 1.373,
      "queue_k_norm": 1.3922,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8242,
      "sent_len_1": 66.7009,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3975,
      "stdk": 0.0482,
      "stdq": 0.0453,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27000
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4817,
      "doc_norm": 1.3936,
      "encoder_q-embeddings": 342.4071,
      "encoder_q-layer.0": 228.2847,
      "encoder_q-layer.1": 253.4028,
      "encoder_q-layer.10": 316.7568,
      "encoder_q-layer.11": 676.728,
      "encoder_q-layer.2": 276.8078,
      "encoder_q-layer.3": 304.1387,
      "encoder_q-layer.4": 331.607,
      "encoder_q-layer.5": 346.9325,
      "encoder_q-layer.6": 370.0833,
      "encoder_q-layer.7": 391.8976,
      "encoder_q-layer.8": 420.2132,
      "encoder_q-layer.9": 332.3729,
      "epoch": 0.26,
      "inbatch_neg_score": 0.1788,
      "inbatch_pos_score": 0.8315,
      "learning_rate": 4.05e-05,
      "loss": 3.4817,
      "norm_diff": 0.0108,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 541.6928,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1799,
      "query_norm": 1.4009,
      "queue_k_norm": 1.3924,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0406,
      "sent_len_1": 66.815,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.79,
      "stdk": 0.048,
      "stdq": 0.0463,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27100
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5016,
      "doc_norm": 1.3923,
      "encoder_q-embeddings": 302.4758,
      "encoder_q-layer.0": 189.554,
      "encoder_q-layer.1": 206.2597,
      "encoder_q-layer.10": 305.8005,
      "encoder_q-layer.11": 644.6493,
      "encoder_q-layer.2": 232.7827,
      "encoder_q-layer.3": 249.6786,
      "encoder_q-layer.4": 258.4534,
      "encoder_q-layer.5": 261.7056,
      "encoder_q-layer.6": 298.1148,
      "encoder_q-layer.7": 328.8524,
      "encoder_q-layer.8": 356.2351,
      "encoder_q-layer.9": 312.4963,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1703,
      "inbatch_pos_score": 0.835,
      "learning_rate": 4.0444444444444444e-05,
      "loss": 3.5016,
      "norm_diff": 0.0096,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 479.4482,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1714,
      "query_norm": 1.3949,
      "queue_k_norm": 1.3915,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1094,
      "sent_len_1": 66.8766,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8625,
      "stdk": 0.048,
      "stdq": 0.0462,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27200
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4781,
      "doc_norm": 1.3904,
      "encoder_q-embeddings": 375.3645,
      "encoder_q-layer.0": 250.8079,
      "encoder_q-layer.1": 277.5989,
      "encoder_q-layer.10": 309.9203,
      "encoder_q-layer.11": 673.5418,
      "encoder_q-layer.2": 321.6376,
      "encoder_q-layer.3": 336.8546,
      "encoder_q-layer.4": 360.3401,
      "encoder_q-layer.5": 375.1649,
      "encoder_q-layer.6": 407.0685,
      "encoder_q-layer.7": 485.9859,
      "encoder_q-layer.8": 492.8849,
      "encoder_q-layer.9": 376.9204,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1685,
      "inbatch_pos_score": 0.7993,
      "learning_rate": 4.038888888888889e-05,
      "loss": 3.4781,
      "norm_diff": 0.0222,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 593.4435,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1687,
      "query_norm": 1.3709,
      "queue_k_norm": 1.3914,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9587,
      "sent_len_1": 66.9557,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.435,
      "stdk": 0.0479,
      "stdq": 0.0452,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27300
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4856,
      "doc_norm": 1.4018,
      "encoder_q-embeddings": 353.5304,
      "encoder_q-layer.0": 222.2991,
      "encoder_q-layer.1": 236.6762,
      "encoder_q-layer.10": 296.3898,
      "encoder_q-layer.11": 617.7743,
      "encoder_q-layer.2": 290.8299,
      "encoder_q-layer.3": 318.4363,
      "encoder_q-layer.4": 350.7294,
      "encoder_q-layer.5": 317.9522,
      "encoder_q-layer.6": 313.2959,
      "encoder_q-layer.7": 325.9909,
      "encoder_q-layer.8": 347.3604,
      "encoder_q-layer.9": 291.961,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1659,
      "inbatch_pos_score": 0.8545,
      "learning_rate": 4.0333333333333336e-05,
      "loss": 3.4856,
      "norm_diff": 0.0166,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 511.8589,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1659,
      "query_norm": 1.3852,
      "queue_k_norm": 1.3906,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8918,
      "sent_len_1": 67.0359,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4038,
      "stdk": 0.0483,
      "stdq": 0.046,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27400
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.465,
      "doc_norm": 1.3867,
      "encoder_q-embeddings": 601.6387,
      "encoder_q-layer.0": 402.3464,
      "encoder_q-layer.1": 462.7727,
      "encoder_q-layer.10": 333.6154,
      "encoder_q-layer.11": 673.7175,
      "encoder_q-layer.2": 541.8199,
      "encoder_q-layer.3": 632.8795,
      "encoder_q-layer.4": 666.3254,
      "encoder_q-layer.5": 629.3623,
      "encoder_q-layer.6": 626.1603,
      "encoder_q-layer.7": 617.8134,
      "encoder_q-layer.8": 605.3142,
      "encoder_q-layer.9": 419.7522,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1666,
      "inbatch_pos_score": 0.8027,
      "learning_rate": 4.027777777777778e-05,
      "loss": 3.465,
      "norm_diff": 0.0151,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 833.7951,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1659,
      "query_norm": 1.3752,
      "queue_k_norm": 1.3914,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8513,
      "sent_len_1": 66.747,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8275,
      "stdk": 0.0477,
      "stdq": 0.0453,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27500
    },
    {
      "accuracy": 47.8516,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4982,
      "doc_norm": 1.3872,
      "encoder_q-embeddings": 346.261,
      "encoder_q-layer.0": 230.2973,
      "encoder_q-layer.1": 243.5777,
      "encoder_q-layer.10": 317.7522,
      "encoder_q-layer.11": 650.7831,
      "encoder_q-layer.2": 279.2197,
      "encoder_q-layer.3": 289.3611,
      "encoder_q-layer.4": 307.4388,
      "encoder_q-layer.5": 325.4483,
      "encoder_q-layer.6": 379.2369,
      "encoder_q-layer.7": 389.2622,
      "encoder_q-layer.8": 405.872,
      "encoder_q-layer.9": 331.5341,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1668,
      "inbatch_pos_score": 0.7686,
      "learning_rate": 4.022222222222222e-05,
      "loss": 3.4982,
      "norm_diff": 0.0137,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 530.6624,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.166,
      "query_norm": 1.378,
      "queue_k_norm": 1.3911,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2192,
      "sent_len_1": 66.7863,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9663,
      "stdk": 0.0478,
      "stdq": 0.0452,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27600
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4639,
      "doc_norm": 1.3907,
      "encoder_q-embeddings": 298.5613,
      "encoder_q-layer.0": 192.212,
      "encoder_q-layer.1": 202.6815,
      "encoder_q-layer.10": 306.3604,
      "encoder_q-layer.11": 683.1004,
      "encoder_q-layer.2": 225.2671,
      "encoder_q-layer.3": 231.7056,
      "encoder_q-layer.4": 250.2779,
      "encoder_q-layer.5": 260.3818,
      "encoder_q-layer.6": 290.6517,
      "encoder_q-layer.7": 328.9333,
      "encoder_q-layer.8": 384.2188,
      "encoder_q-layer.9": 327.0975,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1712,
      "inbatch_pos_score": 0.8047,
      "learning_rate": 4.016666666666667e-05,
      "loss": 3.4639,
      "norm_diff": 0.0068,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 487.6429,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1707,
      "query_norm": 1.3863,
      "queue_k_norm": 1.3902,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0202,
      "sent_len_1": 66.9073,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1312,
      "stdk": 0.0479,
      "stdq": 0.0455,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27700
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5119,
      "doc_norm": 1.3952,
      "encoder_q-embeddings": 374.5482,
      "encoder_q-layer.0": 240.9914,
      "encoder_q-layer.1": 265.3622,
      "encoder_q-layer.10": 315.3367,
      "encoder_q-layer.11": 668.0703,
      "encoder_q-layer.2": 282.7263,
      "encoder_q-layer.3": 278.5398,
      "encoder_q-layer.4": 278.1819,
      "encoder_q-layer.5": 283.6913,
      "encoder_q-layer.6": 305.9671,
      "encoder_q-layer.7": 324.252,
      "encoder_q-layer.8": 359.6732,
      "encoder_q-layer.9": 319.6258,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1746,
      "inbatch_pos_score": 0.8281,
      "learning_rate": 4.011111111111111e-05,
      "loss": 3.5119,
      "norm_diff": 0.0216,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 520.9467,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1749,
      "query_norm": 1.3736,
      "queue_k_norm": 1.3894,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9542,
      "sent_len_1": 66.8313,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6525,
      "stdk": 0.0481,
      "stdq": 0.0449,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 27800
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4944,
      "doc_norm": 1.3905,
      "encoder_q-embeddings": 346.4968,
      "encoder_q-layer.0": 217.0985,
      "encoder_q-layer.1": 239.3578,
      "encoder_q-layer.10": 302.6161,
      "encoder_q-layer.11": 666.8978,
      "encoder_q-layer.2": 262.0471,
      "encoder_q-layer.3": 262.0487,
      "encoder_q-layer.4": 263.7651,
      "encoder_q-layer.5": 274.7829,
      "encoder_q-layer.6": 306.566,
      "encoder_q-layer.7": 359.2784,
      "encoder_q-layer.8": 376.7516,
      "encoder_q-layer.9": 311.2013,
      "epoch": 0.27,
      "inbatch_neg_score": 0.184,
      "inbatch_pos_score": 0.8296,
      "learning_rate": 4.0055555555555554e-05,
      "loss": 3.4944,
      "norm_diff": 0.0084,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 516.9232,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1833,
      "query_norm": 1.3841,
      "queue_k_norm": 1.3922,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0764,
      "sent_len_1": 66.92,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8887,
      "stdk": 0.0479,
      "stdq": 0.045,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 27900
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.5004,
      "doc_norm": 1.3894,
      "encoder_q-embeddings": 343.2004,
      "encoder_q-layer.0": 221.8574,
      "encoder_q-layer.1": 252.1581,
      "encoder_q-layer.10": 296.596,
      "encoder_q-layer.11": 649.8074,
      "encoder_q-layer.2": 299.2414,
      "encoder_q-layer.3": 322.8567,
      "encoder_q-layer.4": 318.3813,
      "encoder_q-layer.5": 334.9,
      "encoder_q-layer.6": 344.2604,
      "encoder_q-layer.7": 360.5254,
      "encoder_q-layer.8": 355.1637,
      "encoder_q-layer.9": 310.8172,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1905,
      "inbatch_pos_score": 0.8203,
      "learning_rate": 4e-05,
      "loss": 3.5004,
      "norm_diff": 0.0126,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 522.2304,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1907,
      "query_norm": 1.3959,
      "queue_k_norm": 1.3891,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1867,
      "sent_len_1": 66.7977,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.055,
      "stdk": 0.0479,
      "stdq": 0.0451,
      "stdqueue_k": 0.0479,
      "stdqueue_q": 0.0,
      "step": 28000
    },
    {
      "accuracy": 52.2461,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4727,
      "doc_norm": 1.3981,
      "encoder_q-embeddings": 420.0651,
      "encoder_q-layer.0": 285.7755,
      "encoder_q-layer.1": 299.4875,
      "encoder_q-layer.10": 313.2478,
      "encoder_q-layer.11": 677.5815,
      "encoder_q-layer.2": 335.0786,
      "encoder_q-layer.3": 348.6338,
      "encoder_q-layer.4": 357.3847,
      "encoder_q-layer.5": 346.0302,
      "encoder_q-layer.6": 401.2935,
      "encoder_q-layer.7": 374.0186,
      "encoder_q-layer.8": 387.6236,
      "encoder_q-layer.9": 312.1217,
      "epoch": 0.27,
      "inbatch_neg_score": 0.1982,
      "inbatch_pos_score": 0.8403,
      "learning_rate": 3.9944444444444446e-05,
      "loss": 3.4727,
      "norm_diff": 0.0183,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 583.6885,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1968,
      "query_norm": 1.4163,
      "queue_k_norm": 1.3918,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9572,
      "sent_len_1": 67.1188,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1138,
      "stdk": 0.0481,
      "stdq": 0.0458,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 28100
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4711,
      "doc_norm": 1.3929,
      "encoder_q-embeddings": 347.2731,
      "encoder_q-layer.0": 221.7388,
      "encoder_q-layer.1": 233.8527,
      "encoder_q-layer.10": 317.85,
      "encoder_q-layer.11": 730.4453,
      "encoder_q-layer.2": 266.7227,
      "encoder_q-layer.3": 297.9767,
      "encoder_q-layer.4": 315.1057,
      "encoder_q-layer.5": 311.1672,
      "encoder_q-layer.6": 359.6851,
      "encoder_q-layer.7": 392.1165,
      "encoder_q-layer.8": 416.3395,
      "encoder_q-layer.9": 334.5261,
      "epoch": 0.28,
      "inbatch_neg_score": 0.197,
      "inbatch_pos_score": 0.8213,
      "learning_rate": 3.9888888888888895e-05,
      "loss": 3.4711,
      "norm_diff": 0.0076,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 554.1314,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1968,
      "query_norm": 1.3947,
      "queue_k_norm": 1.3934,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9461,
      "sent_len_1": 66.7811,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.3338,
      "stdk": 0.0479,
      "stdq": 0.045,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 28200
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4927,
      "doc_norm": 1.3934,
      "encoder_q-embeddings": 857.4952,
      "encoder_q-layer.0": 595.813,
      "encoder_q-layer.1": 712.618,
      "encoder_q-layer.10": 317.4207,
      "encoder_q-layer.11": 679.4676,
      "encoder_q-layer.2": 819.0377,
      "encoder_q-layer.3": 842.0222,
      "encoder_q-layer.4": 905.3379,
      "encoder_q-layer.5": 840.8854,
      "encoder_q-layer.6": 867.9388,
      "encoder_q-layer.7": 803.2452,
      "encoder_q-layer.8": 699.1875,
      "encoder_q-layer.9": 407.1569,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1925,
      "inbatch_pos_score": 0.8345,
      "learning_rate": 3.983333333333333e-05,
      "loss": 3.4927,
      "norm_diff": 0.0118,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1102.284,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1925,
      "query_norm": 1.398,
      "queue_k_norm": 1.3921,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1289,
      "sent_len_1": 66.8193,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2763,
      "stdk": 0.0479,
      "stdq": 0.0458,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 28300
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4847,
      "doc_norm": 1.3902,
      "encoder_q-embeddings": 3135.0586,
      "encoder_q-layer.0": 2246.8845,
      "encoder_q-layer.1": 2670.4297,
      "encoder_q-layer.10": 713.8407,
      "encoder_q-layer.11": 1416.9149,
      "encoder_q-layer.2": 3422.9207,
      "encoder_q-layer.3": 3409.24,
      "encoder_q-layer.4": 3381.0413,
      "encoder_q-layer.5": 3364.0869,
      "encoder_q-layer.6": 3218.1396,
      "encoder_q-layer.7": 3735.9138,
      "encoder_q-layer.8": 2933.6494,
      "encoder_q-layer.9": 1365.4336,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1797,
      "inbatch_pos_score": 0.8198,
      "learning_rate": 3.977777777777778e-05,
      "loss": 3.4847,
      "norm_diff": 0.0174,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4235.2498,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1797,
      "query_norm": 1.3825,
      "queue_k_norm": 1.3946,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1203,
      "sent_len_1": 66.9696,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.1763,
      "stdk": 0.0478,
      "stdq": 0.0456,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 28400
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4563,
      "doc_norm": 1.3898,
      "encoder_q-embeddings": 598.5494,
      "encoder_q-layer.0": 383.8035,
      "encoder_q-layer.1": 419.5444,
      "encoder_q-layer.10": 596.1348,
      "encoder_q-layer.11": 1339.6506,
      "encoder_q-layer.2": 478.8961,
      "encoder_q-layer.3": 493.0173,
      "encoder_q-layer.4": 540.1423,
      "encoder_q-layer.5": 527.1987,
      "encoder_q-layer.6": 573.7214,
      "encoder_q-layer.7": 645.0833,
      "encoder_q-layer.8": 707.3516,
      "encoder_q-layer.9": 628.204,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1698,
      "inbatch_pos_score": 0.811,
      "learning_rate": 3.972222222222222e-05,
      "loss": 3.4563,
      "norm_diff": 0.0262,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 977.867,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1696,
      "query_norm": 1.3673,
      "queue_k_norm": 1.3928,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8099,
      "sent_len_1": 66.6719,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6275,
      "stdk": 0.0478,
      "stdq": 0.0454,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 28500
    },
    {
      "accuracy": 49.0234,
      "active_queue_size": 16384.0,
      "cl_loss": 3.481,
      "doc_norm": 1.3925,
      "encoder_q-embeddings": 784.5367,
      "encoder_q-layer.0": 529.4267,
      "encoder_q-layer.1": 598.5735,
      "encoder_q-layer.10": 599.4656,
      "encoder_q-layer.11": 1400.8477,
      "encoder_q-layer.2": 677.0276,
      "encoder_q-layer.3": 708.3558,
      "encoder_q-layer.4": 744.191,
      "encoder_q-layer.5": 733.0649,
      "encoder_q-layer.6": 774.0969,
      "encoder_q-layer.7": 884.2452,
      "encoder_q-layer.8": 983.4915,
      "encoder_q-layer.9": 750.6884,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1678,
      "inbatch_pos_score": 0.7861,
      "learning_rate": 3.966666666666667e-05,
      "loss": 3.481,
      "norm_diff": 0.0265,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1211.8104,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1682,
      "query_norm": 1.3681,
      "queue_k_norm": 1.3947,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8048,
      "sent_len_1": 66.754,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8438,
      "stdk": 0.0479,
      "stdq": 0.0456,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 28600
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4683,
      "doc_norm": 1.3824,
      "encoder_q-embeddings": 731.1481,
      "encoder_q-layer.0": 487.1625,
      "encoder_q-layer.1": 530.7601,
      "encoder_q-layer.10": 598.067,
      "encoder_q-layer.11": 1432.0481,
      "encoder_q-layer.2": 610.6943,
      "encoder_q-layer.3": 665.2195,
      "encoder_q-layer.4": 779.403,
      "encoder_q-layer.5": 765.603,
      "encoder_q-layer.6": 897.1092,
      "encoder_q-layer.7": 886.6963,
      "encoder_q-layer.8": 808.048,
      "encoder_q-layer.9": 662.9845,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1629,
      "inbatch_pos_score": 0.8052,
      "learning_rate": 3.961111111111111e-05,
      "loss": 3.4683,
      "norm_diff": 0.0177,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1177.5492,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1631,
      "query_norm": 1.3647,
      "queue_k_norm": 1.3941,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9124,
      "sent_len_1": 66.7057,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6488,
      "stdk": 0.0475,
      "stdq": 0.0454,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 28700
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.468,
      "doc_norm": 1.3916,
      "encoder_q-embeddings": 794.1321,
      "encoder_q-layer.0": 561.8221,
      "encoder_q-layer.1": 612.3479,
      "encoder_q-layer.10": 572.9445,
      "encoder_q-layer.11": 1308.3022,
      "encoder_q-layer.2": 679.9037,
      "encoder_q-layer.3": 695.8387,
      "encoder_q-layer.4": 730.5717,
      "encoder_q-layer.5": 748.0062,
      "encoder_q-layer.6": 785.1478,
      "encoder_q-layer.7": 883.6098,
      "encoder_q-layer.8": 843.8471,
      "encoder_q-layer.9": 605.913,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1551,
      "inbatch_pos_score": 0.8203,
      "learning_rate": 3.9555555555555556e-05,
      "loss": 3.468,
      "norm_diff": 0.0389,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1158.7019,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1548,
      "query_norm": 1.3527,
      "queue_k_norm": 1.3937,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9285,
      "sent_len_1": 66.7019,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7237,
      "stdk": 0.0479,
      "stdq": 0.0452,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 28800
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.489,
      "doc_norm": 1.3882,
      "encoder_q-embeddings": 663.7396,
      "encoder_q-layer.0": 446.0032,
      "encoder_q-layer.1": 486.0212,
      "encoder_q-layer.10": 618.6823,
      "encoder_q-layer.11": 1298.6927,
      "encoder_q-layer.2": 545.3929,
      "encoder_q-layer.3": 595.8863,
      "encoder_q-layer.4": 615.9988,
      "encoder_q-layer.5": 649.7404,
      "encoder_q-layer.6": 731.676,
      "encoder_q-layer.7": 858.8658,
      "encoder_q-layer.8": 782.895,
      "encoder_q-layer.9": 632.9567,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1519,
      "inbatch_pos_score": 0.7959,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 3.489,
      "norm_diff": 0.0355,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1062.4589,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1517,
      "query_norm": 1.3527,
      "queue_k_norm": 1.3916,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0138,
      "sent_len_1": 66.5179,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.4175,
      "stdk": 0.0478,
      "stdq": 0.0453,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 28900
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.433,
      "doc_norm": 1.3906,
      "encoder_q-embeddings": 593.0281,
      "encoder_q-layer.0": 368.2375,
      "encoder_q-layer.1": 376.9623,
      "encoder_q-layer.10": 644.426,
      "encoder_q-layer.11": 1375.1624,
      "encoder_q-layer.2": 400.4505,
      "encoder_q-layer.3": 426.1071,
      "encoder_q-layer.4": 455.6331,
      "encoder_q-layer.5": 464.2518,
      "encoder_q-layer.6": 539.9892,
      "encoder_q-layer.7": 643.7612,
      "encoder_q-layer.8": 709.8179,
      "encoder_q-layer.9": 610.7318,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1431,
      "inbatch_pos_score": 0.7734,
      "learning_rate": 3.944444444444445e-05,
      "loss": 3.433,
      "norm_diff": 0.0658,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 957.2409,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1429,
      "query_norm": 1.3248,
      "queue_k_norm": 1.3912,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7859,
      "sent_len_1": 66.7305,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8512,
      "stdk": 0.0479,
      "stdq": 0.0445,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 29000
    },
    {
      "accuracy": 50.0977,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4714,
      "doc_norm": 1.3887,
      "encoder_q-embeddings": 1627.2789,
      "encoder_q-layer.0": 1162.1538,
      "encoder_q-layer.1": 1252.764,
      "encoder_q-layer.10": 581.7797,
      "encoder_q-layer.11": 1317.1597,
      "encoder_q-layer.2": 1412.2751,
      "encoder_q-layer.3": 1608.6775,
      "encoder_q-layer.4": 1751.0814,
      "encoder_q-layer.5": 1972.1448,
      "encoder_q-layer.6": 2176.2786,
      "encoder_q-layer.7": 2580.6807,
      "encoder_q-layer.8": 1565.8423,
      "encoder_q-layer.9": 640.9696,
      "epoch": 0.28,
      "inbatch_neg_score": 0.1501,
      "inbatch_pos_score": 0.7598,
      "learning_rate": 3.938888888888889e-05,
      "loss": 3.4714,
      "norm_diff": 0.0553,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2427.2139,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1499,
      "query_norm": 1.3334,
      "queue_k_norm": 1.3896,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7663,
      "sent_len_1": 66.8876,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.58,
      "stdk": 0.0479,
      "stdq": 0.0445,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 29100
    },
    {
      "accuracy": 50.4883,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4799,
      "doc_norm": 1.3943,
      "encoder_q-embeddings": 751.0612,
      "encoder_q-layer.0": 530.5705,
      "encoder_q-layer.1": 568.816,
      "encoder_q-layer.10": 626.2019,
      "encoder_q-layer.11": 1446.153,
      "encoder_q-layer.2": 651.2855,
      "encoder_q-layer.3": 638.7267,
      "encoder_q-layer.4": 649.0042,
      "encoder_q-layer.5": 651.2932,
      "encoder_q-layer.6": 689.9622,
      "encoder_q-layer.7": 732.9716,
      "encoder_q-layer.8": 733.8787,
      "encoder_q-layer.9": 602.8826,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1485,
      "inbatch_pos_score": 0.7666,
      "learning_rate": 3.933333333333333e-05,
      "loss": 3.4799,
      "norm_diff": 0.057,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1140.2813,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1487,
      "query_norm": 1.3373,
      "queue_k_norm": 1.3912,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1043,
      "sent_len_1": 66.9059,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.145,
      "stdk": 0.0481,
      "stdq": 0.0446,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 29200
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4361,
      "doc_norm": 1.3888,
      "encoder_q-embeddings": 726.0633,
      "encoder_q-layer.0": 485.118,
      "encoder_q-layer.1": 511.9567,
      "encoder_q-layer.10": 655.6023,
      "encoder_q-layer.11": 1334.6447,
      "encoder_q-layer.2": 541.5114,
      "encoder_q-layer.3": 601.9948,
      "encoder_q-layer.4": 616.5939,
      "encoder_q-layer.5": 631.6752,
      "encoder_q-layer.6": 700.851,
      "encoder_q-layer.7": 793.55,
      "encoder_q-layer.8": 806.941,
      "encoder_q-layer.9": 642.6925,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1445,
      "inbatch_pos_score": 0.7871,
      "learning_rate": 3.927777777777778e-05,
      "loss": 3.4361,
      "norm_diff": 0.0191,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1082.0113,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1444,
      "query_norm": 1.3696,
      "queue_k_norm": 1.3898,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9393,
      "sent_len_1": 66.939,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7237,
      "stdk": 0.048,
      "stdq": 0.0459,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 29300
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4679,
      "doc_norm": 1.3838,
      "encoder_q-embeddings": 655.6889,
      "encoder_q-layer.0": 426.4973,
      "encoder_q-layer.1": 471.1734,
      "encoder_q-layer.10": 629.3854,
      "encoder_q-layer.11": 1416.5835,
      "encoder_q-layer.2": 549.6725,
      "encoder_q-layer.3": 588.0253,
      "encoder_q-layer.4": 615.4957,
      "encoder_q-layer.5": 671.6792,
      "encoder_q-layer.6": 746.4229,
      "encoder_q-layer.7": 806.4343,
      "encoder_q-layer.8": 896.5707,
      "encoder_q-layer.9": 677.0565,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1338,
      "inbatch_pos_score": 0.7559,
      "learning_rate": 3.922222222222223e-05,
      "loss": 3.4679,
      "norm_diff": 0.0589,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1110.5559,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1338,
      "query_norm": 1.3249,
      "queue_k_norm": 1.3881,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0486,
      "sent_len_1": 66.599,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.66,
      "stdk": 0.0478,
      "stdq": 0.0445,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 29400
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4478,
      "doc_norm": 1.3855,
      "encoder_q-embeddings": 1402.8934,
      "encoder_q-layer.0": 984.7115,
      "encoder_q-layer.1": 1171.4819,
      "encoder_q-layer.10": 579.3382,
      "encoder_q-layer.11": 1273.7118,
      "encoder_q-layer.2": 1389.9948,
      "encoder_q-layer.3": 1529.9806,
      "encoder_q-layer.4": 1624.176,
      "encoder_q-layer.5": 1571.2291,
      "encoder_q-layer.6": 1624.928,
      "encoder_q-layer.7": 1605.8124,
      "encoder_q-layer.8": 1354.942,
      "encoder_q-layer.9": 810.9839,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1232,
      "inbatch_pos_score": 0.7441,
      "learning_rate": 3.9166666666666665e-05,
      "loss": 3.4478,
      "norm_diff": 0.0546,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1983.6659,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.123,
      "query_norm": 1.3309,
      "queue_k_norm": 1.386,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8987,
      "sent_len_1": 66.5796,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6312,
      "stdk": 0.0479,
      "stdq": 0.0449,
      "stdqueue_k": 0.048,
      "stdqueue_q": 0.0,
      "step": 29500
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4611,
      "doc_norm": 1.389,
      "encoder_q-embeddings": 1400.8192,
      "encoder_q-layer.0": 925.5343,
      "encoder_q-layer.1": 987.3005,
      "encoder_q-layer.10": 640.984,
      "encoder_q-layer.11": 1326.5619,
      "encoder_q-layer.2": 1079.6589,
      "encoder_q-layer.3": 1164.7699,
      "encoder_q-layer.4": 1259.8539,
      "encoder_q-layer.5": 1133.5531,
      "encoder_q-layer.6": 1290.8114,
      "encoder_q-layer.7": 1047.1473,
      "encoder_q-layer.8": 967.5212,
      "encoder_q-layer.9": 701.904,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1234,
      "inbatch_pos_score": 0.8291,
      "learning_rate": 3.9111111111111115e-05,
      "loss": 3.4611,
      "norm_diff": 0.0145,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1656.8074,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1227,
      "query_norm": 1.3954,
      "queue_k_norm": 1.3873,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0617,
      "sent_len_1": 67.123,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0012,
      "stdk": 0.0481,
      "stdq": 0.0472,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 29600
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4514,
      "doc_norm": 1.3807,
      "encoder_q-embeddings": 285.1942,
      "encoder_q-layer.0": 185.1839,
      "encoder_q-layer.1": 195.8015,
      "encoder_q-layer.10": 305.9779,
      "encoder_q-layer.11": 656.3795,
      "encoder_q-layer.2": 221.0107,
      "encoder_q-layer.3": 226.7665,
      "encoder_q-layer.4": 245.8741,
      "encoder_q-layer.5": 239.0941,
      "encoder_q-layer.6": 266.4803,
      "encoder_q-layer.7": 313.1242,
      "encoder_q-layer.8": 345.7126,
      "encoder_q-layer.9": 309.1768,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1178,
      "inbatch_pos_score": 0.7793,
      "learning_rate": 3.905555555555556e-05,
      "loss": 3.4514,
      "norm_diff": 0.0188,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 463.5834,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1172,
      "query_norm": 1.3623,
      "queue_k_norm": 1.3885,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0198,
      "sent_len_1": 66.9426,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1987,
      "stdk": 0.0478,
      "stdq": 0.0458,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 29700
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4614,
      "doc_norm": 1.3843,
      "encoder_q-embeddings": 442.89,
      "encoder_q-layer.0": 291.0706,
      "encoder_q-layer.1": 311.4517,
      "encoder_q-layer.10": 330.1254,
      "encoder_q-layer.11": 672.4741,
      "encoder_q-layer.2": 373.0616,
      "encoder_q-layer.3": 392.0553,
      "encoder_q-layer.4": 449.9597,
      "encoder_q-layer.5": 474.0485,
      "encoder_q-layer.6": 499.3018,
      "encoder_q-layer.7": 441.2113,
      "encoder_q-layer.8": 417.8773,
      "encoder_q-layer.9": 331.8708,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1191,
      "inbatch_pos_score": 0.7588,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 3.4614,
      "norm_diff": 0.0651,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 634.4898,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.119,
      "query_norm": 1.3192,
      "queue_k_norm": 1.3857,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0186,
      "sent_len_1": 67.0511,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.395,
      "stdk": 0.048,
      "stdq": 0.044,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 29800
    },
    {
      "accuracy": 49.3164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4574,
      "doc_norm": 1.3805,
      "encoder_q-embeddings": 317.4329,
      "encoder_q-layer.0": 191.2176,
      "encoder_q-layer.1": 201.3423,
      "encoder_q-layer.10": 320.4917,
      "encoder_q-layer.11": 652.6097,
      "encoder_q-layer.2": 219.5969,
      "encoder_q-layer.3": 233.0954,
      "encoder_q-layer.4": 255.7206,
      "encoder_q-layer.5": 261.2845,
      "encoder_q-layer.6": 311.1333,
      "encoder_q-layer.7": 355.8026,
      "encoder_q-layer.8": 381.3768,
      "encoder_q-layer.9": 317.1201,
      "epoch": 0.29,
      "inbatch_neg_score": 0.111,
      "inbatch_pos_score": 0.729,
      "learning_rate": 3.894444444444444e-05,
      "loss": 3.4574,
      "norm_diff": 0.0198,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 492.603,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1116,
      "query_norm": 1.3607,
      "queue_k_norm": 1.3848,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8698,
      "sent_len_1": 66.9204,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6937,
      "stdk": 0.0479,
      "stdq": 0.0454,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 29900
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4547,
      "doc_norm": 1.3883,
      "encoder_q-embeddings": 605.2714,
      "encoder_q-layer.0": 402.2841,
      "encoder_q-layer.1": 467.2162,
      "encoder_q-layer.10": 288.9337,
      "encoder_q-layer.11": 626.6219,
      "encoder_q-layer.2": 604.9283,
      "encoder_q-layer.3": 554.6036,
      "encoder_q-layer.4": 570.2275,
      "encoder_q-layer.5": 541.5269,
      "encoder_q-layer.6": 586.2702,
      "encoder_q-layer.7": 531.697,
      "encoder_q-layer.8": 573.3337,
      "encoder_q-layer.9": 356.4286,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1085,
      "inbatch_pos_score": 0.7568,
      "learning_rate": 3.888888888888889e-05,
      "loss": 3.4547,
      "norm_diff": 0.0118,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 798.7838,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1072,
      "query_norm": 1.3764,
      "queue_k_norm": 1.3822,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7597,
      "sent_len_1": 66.7155,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3025,
      "stdk": 0.0482,
      "stdq": 0.0456,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 30000
    },
    {
      "dev_runtime": 38.1229,
      "dev_samples_per_second": 1.679,
      "dev_steps_per_second": 0.026,
      "epoch": 0.29,
      "step": 30000,
      "test_accuracy": 94.0185546875,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3498751223087311,
      "test_doc_norm": 1.380068063735962,
      "test_inbatch_neg_score": 0.46064990758895874,
      "test_inbatch_pos_score": 1.4007467031478882,
      "test_loss": 0.3498751223087311,
      "test_loss_align": 1.2858573198318481,
      "test_loss_unif": 3.963012218475342,
      "test_loss_unif_q@queue": 3.963012218475342,
      "test_norm_diff": 0.08707223832607269,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.08252152800559998,
      "test_query_norm": 1.4671404361724854,
      "test_queue_k_norm": 1.3822989463806152,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04247192665934563,
      "test_stdq": 0.04322770610451698,
      "test_stdqueue_k": 0.04808451980352402,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 38.1229,
      "dev_samples_per_second": 1.679,
      "dev_steps_per_second": 0.026,
      "epoch": 0.29,
      "eval_beir-arguana_ndcg@10": 0.37221,
      "eval_beir-arguana_recall@10": 0.62518,
      "eval_beir-arguana_recall@100": 0.90967,
      "eval_beir-arguana_recall@20": 0.75818,
      "eval_beir-avg_ndcg@10": 0.3373613333333333,
      "eval_beir-avg_recall@10": 0.402616,
      "eval_beir-avg_recall@100": 0.578604,
      "eval_beir-avg_recall@20": 0.4599897499999999,
      "eval_beir-cqadupstack_ndcg@10": 0.2407933333333334,
      "eval_beir-cqadupstack_recall@10": 0.32663000000000003,
      "eval_beir-cqadupstack_recall@100": 0.55051,
      "eval_beir-cqadupstack_recall@20": 0.39045749999999996,
      "eval_beir-fiqa_ndcg@10": 0.16827,
      "eval_beir-fiqa_recall@10": 0.21612,
      "eval_beir-fiqa_recall@100": 0.45689,
      "eval_beir-fiqa_recall@20": 0.28102,
      "eval_beir-nfcorpus_ndcg@10": 0.2608,
      "eval_beir-nfcorpus_recall@10": 0.12924,
      "eval_beir-nfcorpus_recall@100": 0.25164,
      "eval_beir-nfcorpus_recall@20": 0.15714,
      "eval_beir-nq_ndcg@10": 0.2449,
      "eval_beir-nq_recall@10": 0.4052,
      "eval_beir-nq_recall@100": 0.75135,
      "eval_beir-nq_recall@20": 0.51972,
      "eval_beir-quora_ndcg@10": 0.791,
      "eval_beir-quora_recall@10": 0.89303,
      "eval_beir-quora_recall@100": 0.97892,
      "eval_beir-quora_recall@20": 0.93155,
      "eval_beir-scidocs_ndcg@10": 0.12623,
      "eval_beir-scidocs_recall@10": 0.13348,
      "eval_beir-scidocs_recall@100": 0.32103,
      "eval_beir-scidocs_recall@20": 0.18533,
      "eval_beir-scifact_ndcg@10": 0.59499,
      "eval_beir-scifact_recall@10": 0.73833,
      "eval_beir-scifact_recall@100": 0.89189,
      "eval_beir-scifact_recall@20": 0.78833,
      "eval_beir-trec-covid_ndcg@10": 0.40669,
      "eval_beir-trec-covid_recall@10": 0.424,
      "eval_beir-trec-covid_recall@100": 0.3042,
      "eval_beir-trec-covid_recall@20": 0.39,
      "eval_beir-webis-touche2020_ndcg@10": 0.16773,
      "eval_beir-webis-touche2020_recall@10": 0.13495,
      "eval_beir-webis-touche2020_recall@100": 0.36994,
      "eval_beir-webis-touche2020_recall@20": 0.19817,
      "eval_senteval-avg_sts": 0.7341376697602899,
      "eval_senteval-sickr_spearman": 0.7004668032127727,
      "eval_senteval-stsb_spearman": 0.7678085363078071,
      "step": 30000,
      "test_accuracy": 94.0185546875,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3498751223087311,
      "test_doc_norm": 1.380068063735962,
      "test_inbatch_neg_score": 0.46064990758895874,
      "test_inbatch_pos_score": 1.4007467031478882,
      "test_loss": 0.3498751223087311,
      "test_loss_align": 1.2858573198318481,
      "test_loss_unif": 3.963012218475342,
      "test_loss_unif_q@queue": 3.963012218475342,
      "test_norm_diff": 0.08707223832607269,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.08252152800559998,
      "test_query_norm": 1.4671404361724854,
      "test_queue_k_norm": 1.3822989463806152,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04247192665934563,
      "test_stdq": 0.04322770610451698,
      "test_stdqueue_k": 0.04808451980352402,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.454,
      "doc_norm": 1.3866,
      "encoder_q-embeddings": 280.5482,
      "encoder_q-layer.0": 178.5174,
      "encoder_q-layer.1": 194.6327,
      "encoder_q-layer.10": 374.2943,
      "encoder_q-layer.11": 668.9407,
      "encoder_q-layer.2": 218.0716,
      "encoder_q-layer.3": 231.2326,
      "encoder_q-layer.4": 245.5285,
      "encoder_q-layer.5": 254.2705,
      "encoder_q-layer.6": 303.0859,
      "encoder_q-layer.7": 342.4812,
      "encoder_q-layer.8": 381.1265,
      "encoder_q-layer.9": 351.576,
      "epoch": 0.29,
      "inbatch_neg_score": 0.0937,
      "inbatch_pos_score": 0.7788,
      "learning_rate": 3.883333333333333e-05,
      "loss": 3.454,
      "norm_diff": 0.0499,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 482.87,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.0936,
      "query_norm": 1.4364,
      "queue_k_norm": 1.3843,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0609,
      "sent_len_1": 66.9788,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2587,
      "stdk": 0.0482,
      "stdq": 0.0473,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 30100
    },
    {
      "accuracy": 51.0742,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4633,
      "doc_norm": 1.3817,
      "encoder_q-embeddings": 368.3238,
      "encoder_q-layer.0": 249.5048,
      "encoder_q-layer.1": 272.9344,
      "encoder_q-layer.10": 334.1044,
      "encoder_q-layer.11": 654.7311,
      "encoder_q-layer.2": 311.6733,
      "encoder_q-layer.3": 336.2488,
      "encoder_q-layer.4": 361.3743,
      "encoder_q-layer.5": 370.3396,
      "encoder_q-layer.6": 394.8345,
      "encoder_q-layer.7": 388.946,
      "encoder_q-layer.8": 422.5164,
      "encoder_q-layer.9": 341.915,
      "epoch": 0.29,
      "inbatch_neg_score": 0.1006,
      "inbatch_pos_score": 0.7207,
      "learning_rate": 3.877777777777778e-05,
      "loss": 3.4633,
      "norm_diff": 0.006,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 560.2296,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1001,
      "query_norm": 1.3803,
      "queue_k_norm": 1.3807,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6236,
      "sent_len_1": 66.7555,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4625,
      "stdk": 0.0481,
      "stdq": 0.0455,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 30200
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4371,
      "doc_norm": 1.3832,
      "encoder_q-embeddings": 341.3515,
      "encoder_q-layer.0": 233.5754,
      "encoder_q-layer.1": 245.5852,
      "encoder_q-layer.10": 288.7607,
      "encoder_q-layer.11": 613.0705,
      "encoder_q-layer.2": 278.2746,
      "encoder_q-layer.3": 289.2029,
      "encoder_q-layer.4": 329.1838,
      "encoder_q-layer.5": 336.2146,
      "encoder_q-layer.6": 388.6993,
      "encoder_q-layer.7": 400.1411,
      "encoder_q-layer.8": 398.3086,
      "encoder_q-layer.9": 301.0706,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1121,
      "inbatch_pos_score": 0.791,
      "learning_rate": 3.8722222222222225e-05,
      "loss": 3.4371,
      "norm_diff": 0.0076,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 526.5611,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1111,
      "query_norm": 1.3885,
      "queue_k_norm": 1.3807,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9858,
      "sent_len_1": 66.7527,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1337,
      "stdk": 0.0482,
      "stdq": 0.0461,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 30300
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4571,
      "doc_norm": 1.3788,
      "encoder_q-embeddings": 474.1507,
      "encoder_q-layer.0": 343.4873,
      "encoder_q-layer.1": 385.6761,
      "encoder_q-layer.10": 303.8468,
      "encoder_q-layer.11": 672.322,
      "encoder_q-layer.2": 472.5108,
      "encoder_q-layer.3": 508.7935,
      "encoder_q-layer.4": 565.8071,
      "encoder_q-layer.5": 611.9758,
      "encoder_q-layer.6": 634.4288,
      "encoder_q-layer.7": 569.3544,
      "encoder_q-layer.8": 475.5955,
      "encoder_q-layer.9": 337.5597,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1108,
      "inbatch_pos_score": 0.7661,
      "learning_rate": 3.866666666666667e-05,
      "loss": 3.4571,
      "norm_diff": 0.0087,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 731.7751,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1103,
      "query_norm": 1.3743,
      "queue_k_norm": 1.3788,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9136,
      "sent_len_1": 66.8388,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.795,
      "stdk": 0.048,
      "stdq": 0.0459,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 30400
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.433,
      "doc_norm": 1.371,
      "encoder_q-embeddings": 296.9214,
      "encoder_q-layer.0": 192.3117,
      "encoder_q-layer.1": 213.067,
      "encoder_q-layer.10": 277.5073,
      "encoder_q-layer.11": 630.6231,
      "encoder_q-layer.2": 232.1652,
      "encoder_q-layer.3": 253.053,
      "encoder_q-layer.4": 284.0829,
      "encoder_q-layer.5": 282.1805,
      "encoder_q-layer.6": 313.1195,
      "encoder_q-layer.7": 369.0649,
      "encoder_q-layer.8": 399.6806,
      "encoder_q-layer.9": 322.2369,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1137,
      "inbatch_pos_score": 0.7627,
      "learning_rate": 3.8611111111111116e-05,
      "loss": 3.433,
      "norm_diff": 0.0234,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 491.4719,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1135,
      "query_norm": 1.3476,
      "queue_k_norm": 1.3798,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0483,
      "sent_len_1": 66.9817,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9075,
      "stdk": 0.0478,
      "stdq": 0.0454,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 30500
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4395,
      "doc_norm": 1.3794,
      "encoder_q-embeddings": 278.8021,
      "encoder_q-layer.0": 181.3715,
      "encoder_q-layer.1": 193.0242,
      "encoder_q-layer.10": 293.9362,
      "encoder_q-layer.11": 667.9576,
      "encoder_q-layer.2": 212.4348,
      "encoder_q-layer.3": 227.884,
      "encoder_q-layer.4": 233.6788,
      "encoder_q-layer.5": 249.8927,
      "encoder_q-layer.6": 278.4362,
      "encoder_q-layer.7": 315.1787,
      "encoder_q-layer.8": 361.655,
      "encoder_q-layer.9": 314.8848,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1203,
      "inbatch_pos_score": 0.772,
      "learning_rate": 3.855555555555556e-05,
      "loss": 3.4395,
      "norm_diff": 0.0329,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 471.6485,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1202,
      "query_norm": 1.3465,
      "queue_k_norm": 1.3814,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8753,
      "sent_len_1": 66.7127,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.475,
      "stdk": 0.0481,
      "stdq": 0.0455,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 30600
    },
    {
      "accuracy": 52.9297,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4249,
      "doc_norm": 1.3824,
      "encoder_q-embeddings": 366.9658,
      "encoder_q-layer.0": 244.4004,
      "encoder_q-layer.1": 260.4187,
      "encoder_q-layer.10": 301.1602,
      "encoder_q-layer.11": 620.4222,
      "encoder_q-layer.2": 294.6967,
      "encoder_q-layer.3": 296.3588,
      "encoder_q-layer.4": 316.1601,
      "encoder_q-layer.5": 315.5422,
      "encoder_q-layer.6": 350.979,
      "encoder_q-layer.7": 362.6467,
      "encoder_q-layer.8": 410.5288,
      "encoder_q-layer.9": 325.3217,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1153,
      "inbatch_pos_score": 0.7446,
      "learning_rate": 3.85e-05,
      "loss": 3.4249,
      "norm_diff": 0.0546,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 529.2165,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1146,
      "query_norm": 1.3279,
      "queue_k_norm": 1.3798,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9558,
      "sent_len_1": 66.9005,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2163,
      "stdk": 0.0482,
      "stdq": 0.0449,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 30700
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4597,
      "doc_norm": 1.3776,
      "encoder_q-embeddings": 320.7915,
      "encoder_q-layer.0": 203.5895,
      "encoder_q-layer.1": 226.0952,
      "encoder_q-layer.10": 286.535,
      "encoder_q-layer.11": 619.6965,
      "encoder_q-layer.2": 250.7513,
      "encoder_q-layer.3": 261.0804,
      "encoder_q-layer.4": 275.5554,
      "encoder_q-layer.5": 273.5303,
      "encoder_q-layer.6": 313.9961,
      "encoder_q-layer.7": 345.7357,
      "encoder_q-layer.8": 390.0846,
      "encoder_q-layer.9": 310.047,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1116,
      "inbatch_pos_score": 0.7461,
      "learning_rate": 3.844444444444444e-05,
      "loss": 3.4597,
      "norm_diff": 0.0601,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 493.7567,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.113,
      "query_norm": 1.3175,
      "queue_k_norm": 1.3779,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7002,
      "sent_len_1": 66.6415,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1863,
      "stdk": 0.0481,
      "stdq": 0.0447,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 30800
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4338,
      "doc_norm": 1.3863,
      "encoder_q-embeddings": 272.305,
      "encoder_q-layer.0": 171.3681,
      "encoder_q-layer.1": 174.0996,
      "encoder_q-layer.10": 300.1933,
      "encoder_q-layer.11": 636.0707,
      "encoder_q-layer.2": 197.1947,
      "encoder_q-layer.3": 203.5813,
      "encoder_q-layer.4": 223.611,
      "encoder_q-layer.5": 230.891,
      "encoder_q-layer.6": 262.9292,
      "encoder_q-layer.7": 305.1797,
      "encoder_q-layer.8": 345.8303,
      "encoder_q-layer.9": 312.8163,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1151,
      "inbatch_pos_score": 0.7905,
      "learning_rate": 3.838888888888889e-05,
      "loss": 3.4338,
      "norm_diff": 0.0229,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 450.5826,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1149,
      "query_norm": 1.3661,
      "queue_k_norm": 1.3771,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0196,
      "sent_len_1": 66.5207,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.16,
      "stdk": 0.0484,
      "stdq": 0.0464,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 30900
    },
    {
      "accuracy": 48.9258,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4217,
      "doc_norm": 1.3789,
      "encoder_q-embeddings": 289.8737,
      "encoder_q-layer.0": 184.4675,
      "encoder_q-layer.1": 194.3885,
      "encoder_q-layer.10": 393.9234,
      "encoder_q-layer.11": 721.4958,
      "encoder_q-layer.2": 214.0194,
      "encoder_q-layer.3": 221.8728,
      "encoder_q-layer.4": 239.303,
      "encoder_q-layer.5": 242.6071,
      "encoder_q-layer.6": 288.9218,
      "encoder_q-layer.7": 341.1543,
      "encoder_q-layer.8": 420.8401,
      "encoder_q-layer.9": 401.449,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1179,
      "inbatch_pos_score": 0.7251,
      "learning_rate": 3.8333333333333334e-05,
      "loss": 3.4217,
      "norm_diff": 0.0295,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 504.1768,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1176,
      "query_norm": 1.3493,
      "queue_k_norm": 1.3762,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9068,
      "sent_len_1": 66.6287,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3325,
      "stdk": 0.0481,
      "stdq": 0.0458,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 31000
    },
    {
      "accuracy": 50.1953,
      "active_queue_size": 16384.0,
      "cl_loss": 3.424,
      "doc_norm": 1.3722,
      "encoder_q-embeddings": 349.7294,
      "encoder_q-layer.0": 225.4463,
      "encoder_q-layer.1": 241.314,
      "encoder_q-layer.10": 294.541,
      "encoder_q-layer.11": 658.2355,
      "encoder_q-layer.2": 265.9107,
      "encoder_q-layer.3": 288.6017,
      "encoder_q-layer.4": 320.9901,
      "encoder_q-layer.5": 323.0116,
      "encoder_q-layer.6": 340.0059,
      "encoder_q-layer.7": 379.2144,
      "encoder_q-layer.8": 386.954,
      "encoder_q-layer.9": 304.0169,
      "epoch": 0.3,
      "inbatch_neg_score": 0.1181,
      "inbatch_pos_score": 0.7368,
      "learning_rate": 3.827777777777778e-05,
      "loss": 3.424,
      "norm_diff": 0.041,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 522.7045,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1189,
      "query_norm": 1.3311,
      "queue_k_norm": 1.3758,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6905,
      "sent_len_1": 66.8089,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6725,
      "stdk": 0.0479,
      "stdq": 0.0451,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 31100
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4188,
      "doc_norm": 1.3803,
      "encoder_q-embeddings": 301.32,
      "encoder_q-layer.0": 200.8216,
      "encoder_q-layer.1": 207.226,
      "encoder_q-layer.10": 339.6295,
      "encoder_q-layer.11": 680.9681,
      "encoder_q-layer.2": 235.76,
      "encoder_q-layer.3": 255.7731,
      "encoder_q-layer.4": 272.9095,
      "encoder_q-layer.5": 274.7479,
      "encoder_q-layer.6": 304.4538,
      "encoder_q-layer.7": 340.4616,
      "encoder_q-layer.8": 395.2029,
      "encoder_q-layer.9": 353.8735,
      "epoch": 0.3,
      "inbatch_neg_score": 0.126,
      "inbatch_pos_score": 0.7603,
      "learning_rate": 3.8222222222222226e-05,
      "loss": 3.4188,
      "norm_diff": 0.0574,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 496.2139,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.125,
      "query_norm": 1.3229,
      "queue_k_norm": 1.3761,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9287,
      "sent_len_1": 66.7848,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4187,
      "stdk": 0.0482,
      "stdq": 0.0445,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 31200
    },
    {
      "accuracy": 47.4609,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4404,
      "doc_norm": 1.3811,
      "encoder_q-embeddings": 855.4653,
      "encoder_q-layer.0": 584.805,
      "encoder_q-layer.1": 696.1826,
      "encoder_q-layer.10": 327.0185,
      "encoder_q-layer.11": 696.358,
      "encoder_q-layer.2": 755.5555,
      "encoder_q-layer.3": 870.929,
      "encoder_q-layer.4": 917.5845,
      "encoder_q-layer.5": 968.691,
      "encoder_q-layer.6": 997.0345,
      "encoder_q-layer.7": 1005.5869,
      "encoder_q-layer.8": 899.2949,
      "encoder_q-layer.9": 446.4858,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1296,
      "inbatch_pos_score": 0.7285,
      "learning_rate": 3.816666666666667e-05,
      "loss": 3.4404,
      "norm_diff": 0.0411,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1166.6489,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1284,
      "query_norm": 1.3399,
      "queue_k_norm": 1.3764,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1473,
      "sent_len_1": 66.7842,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1475,
      "stdk": 0.0482,
      "stdq": 0.0451,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 31300
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4255,
      "doc_norm": 1.3705,
      "encoder_q-embeddings": 334.2563,
      "encoder_q-layer.0": 229.487,
      "encoder_q-layer.1": 249.5126,
      "encoder_q-layer.10": 286.0422,
      "encoder_q-layer.11": 627.2976,
      "encoder_q-layer.2": 282.0812,
      "encoder_q-layer.3": 308.979,
      "encoder_q-layer.4": 323.1936,
      "encoder_q-layer.5": 316.5303,
      "encoder_q-layer.6": 310.8006,
      "encoder_q-layer.7": 322.1045,
      "encoder_q-layer.8": 376.8978,
      "encoder_q-layer.9": 295.084,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1292,
      "inbatch_pos_score": 0.7832,
      "learning_rate": 3.811111111111112e-05,
      "loss": 3.4255,
      "norm_diff": 0.0135,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 504.6346,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1278,
      "query_norm": 1.3582,
      "queue_k_norm": 1.3786,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9701,
      "sent_len_1": 66.8726,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4425,
      "stdk": 0.0478,
      "stdq": 0.0457,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 31400
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4274,
      "doc_norm": 1.3755,
      "encoder_q-embeddings": 299.8251,
      "encoder_q-layer.0": 187.3497,
      "encoder_q-layer.1": 191.9148,
      "encoder_q-layer.10": 296.9844,
      "encoder_q-layer.11": 614.5362,
      "encoder_q-layer.2": 218.1567,
      "encoder_q-layer.3": 231.1966,
      "encoder_q-layer.4": 247.8283,
      "encoder_q-layer.5": 254.4945,
      "encoder_q-layer.6": 288.0428,
      "encoder_q-layer.7": 307.4443,
      "encoder_q-layer.8": 342.6272,
      "encoder_q-layer.9": 290.8919,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1316,
      "inbatch_pos_score": 0.7671,
      "learning_rate": 3.805555555555555e-05,
      "loss": 3.4274,
      "norm_diff": 0.0421,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 463.6692,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1315,
      "query_norm": 1.3333,
      "queue_k_norm": 1.3789,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0824,
      "sent_len_1": 67.1225,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6188,
      "stdk": 0.048,
      "stdq": 0.0448,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 31500
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4064,
      "doc_norm": 1.3767,
      "encoder_q-embeddings": 502.3304,
      "encoder_q-layer.0": 313.8199,
      "encoder_q-layer.1": 324.5991,
      "encoder_q-layer.10": 309.5508,
      "encoder_q-layer.11": 677.4636,
      "encoder_q-layer.2": 372.6641,
      "encoder_q-layer.3": 426.833,
      "encoder_q-layer.4": 462.7313,
      "encoder_q-layer.5": 485.2301,
      "encoder_q-layer.6": 512.4251,
      "encoder_q-layer.7": 561.4656,
      "encoder_q-layer.8": 501.5811,
      "encoder_q-layer.9": 331.5261,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1344,
      "inbatch_pos_score": 0.7939,
      "learning_rate": 3.8e-05,
      "loss": 3.4064,
      "norm_diff": 0.0182,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 676.3165,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1343,
      "query_norm": 1.3602,
      "queue_k_norm": 1.3797,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8437,
      "sent_len_1": 66.7214,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2887,
      "stdk": 0.048,
      "stdq": 0.0458,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 31600
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4312,
      "doc_norm": 1.3846,
      "encoder_q-embeddings": 697.2238,
      "encoder_q-layer.0": 451.4753,
      "encoder_q-layer.1": 471.8613,
      "encoder_q-layer.10": 600.5973,
      "encoder_q-layer.11": 1204.605,
      "encoder_q-layer.2": 564.4144,
      "encoder_q-layer.3": 606.4318,
      "encoder_q-layer.4": 700.127,
      "encoder_q-layer.5": 724.9402,
      "encoder_q-layer.6": 804.2742,
      "encoder_q-layer.7": 875.3527,
      "encoder_q-layer.8": 943.1266,
      "encoder_q-layer.9": 757.0526,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1345,
      "inbatch_pos_score": 0.7769,
      "learning_rate": 3.7944444444444444e-05,
      "loss": 3.4312,
      "norm_diff": 0.0413,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1115.4323,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.135,
      "query_norm": 1.3434,
      "queue_k_norm": 1.3768,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7887,
      "sent_len_1": 66.7368,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8237,
      "stdk": 0.0483,
      "stdq": 0.045,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 31700
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4042,
      "doc_norm": 1.3751,
      "encoder_q-embeddings": 641.7238,
      "encoder_q-layer.0": 425.4843,
      "encoder_q-layer.1": 460.3211,
      "encoder_q-layer.10": 609.5402,
      "encoder_q-layer.11": 1383.5685,
      "encoder_q-layer.2": 507.6791,
      "encoder_q-layer.3": 565.1183,
      "encoder_q-layer.4": 611.7704,
      "encoder_q-layer.5": 616.5628,
      "encoder_q-layer.6": 708.8625,
      "encoder_q-layer.7": 743.7507,
      "encoder_q-layer.8": 818.7391,
      "encoder_q-layer.9": 683.9813,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1406,
      "inbatch_pos_score": 0.8027,
      "learning_rate": 3.7888888888888894e-05,
      "loss": 3.4042,
      "norm_diff": 0.01,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1055.8361,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1411,
      "query_norm": 1.3763,
      "queue_k_norm": 1.3778,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.876,
      "sent_len_1": 66.6571,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0062,
      "stdk": 0.0479,
      "stdq": 0.046,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 31800
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4346,
      "doc_norm": 1.374,
      "encoder_q-embeddings": 536.9716,
      "encoder_q-layer.0": 355.7304,
      "encoder_q-layer.1": 371.3701,
      "encoder_q-layer.10": 579.4551,
      "encoder_q-layer.11": 1261.6665,
      "encoder_q-layer.2": 415.0086,
      "encoder_q-layer.3": 438.7644,
      "encoder_q-layer.4": 464.7516,
      "encoder_q-layer.5": 484.9456,
      "encoder_q-layer.6": 561.824,
      "encoder_q-layer.7": 619.3617,
      "encoder_q-layer.8": 709.6611,
      "encoder_q-layer.9": 596.8021,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1454,
      "inbatch_pos_score": 0.7832,
      "learning_rate": 3.7833333333333336e-05,
      "loss": 3.4346,
      "norm_diff": 0.0208,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 902.8982,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1443,
      "query_norm": 1.3532,
      "queue_k_norm": 1.3818,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9206,
      "sent_len_1": 66.8313,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5012,
      "stdk": 0.0479,
      "stdq": 0.045,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 31900
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4246,
      "doc_norm": 1.3774,
      "encoder_q-embeddings": 995.9264,
      "encoder_q-layer.0": 622.9302,
      "encoder_q-layer.1": 673.9819,
      "encoder_q-layer.10": 675.8527,
      "encoder_q-layer.11": 1338.1176,
      "encoder_q-layer.2": 789.348,
      "encoder_q-layer.3": 887.179,
      "encoder_q-layer.4": 985.7158,
      "encoder_q-layer.5": 1081.3486,
      "encoder_q-layer.6": 1290.0127,
      "encoder_q-layer.7": 1856.6559,
      "encoder_q-layer.8": 2312.2422,
      "encoder_q-layer.9": 1499.7362,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1435,
      "inbatch_pos_score": 0.7603,
      "learning_rate": 3.777777777777778e-05,
      "loss": 3.4246,
      "norm_diff": 0.0147,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1889.2727,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1444,
      "query_norm": 1.3627,
      "queue_k_norm": 1.3787,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7285,
      "sent_len_1": 66.7858,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.2375,
      "stdk": 0.048,
      "stdq": 0.0453,
      "stdqueue_k": 0.0481,
      "stdqueue_q": 0.0,
      "step": 32000
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4246,
      "doc_norm": 1.3754,
      "encoder_q-embeddings": 538.1599,
      "encoder_q-layer.0": 344.7924,
      "encoder_q-layer.1": 364.9943,
      "encoder_q-layer.10": 623.8983,
      "encoder_q-layer.11": 1267.0154,
      "encoder_q-layer.2": 403.6741,
      "encoder_q-layer.3": 422.6486,
      "encoder_q-layer.4": 452.7309,
      "encoder_q-layer.5": 446.3288,
      "encoder_q-layer.6": 506.3129,
      "encoder_q-layer.7": 614.0193,
      "encoder_q-layer.8": 703.0637,
      "encoder_q-layer.9": 640.0541,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1469,
      "inbatch_pos_score": 0.8374,
      "learning_rate": 3.772222222222223e-05,
      "loss": 3.4246,
      "norm_diff": 0.0193,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 885.3088,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1478,
      "query_norm": 1.3924,
      "queue_k_norm": 1.3811,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9603,
      "sent_len_1": 66.7738,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.59,
      "stdk": 0.0479,
      "stdq": 0.0463,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 32100
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3798,
      "doc_norm": 1.3848,
      "encoder_q-embeddings": 781.974,
      "encoder_q-layer.0": 509.3951,
      "encoder_q-layer.1": 579.3201,
      "encoder_q-layer.10": 582.0287,
      "encoder_q-layer.11": 1336.6405,
      "encoder_q-layer.2": 643.1027,
      "encoder_q-layer.3": 706.2607,
      "encoder_q-layer.4": 697.3978,
      "encoder_q-layer.5": 769.8038,
      "encoder_q-layer.6": 889.5179,
      "encoder_q-layer.7": 915.5873,
      "encoder_q-layer.8": 946.3196,
      "encoder_q-layer.9": 633.7072,
      "epoch": 0.31,
      "inbatch_neg_score": 0.1503,
      "inbatch_pos_score": 0.75,
      "learning_rate": 3.766666666666667e-05,
      "loss": 3.3798,
      "norm_diff": 0.0403,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1201.5055,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1487,
      "query_norm": 1.3445,
      "queue_k_norm": 1.3824,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9209,
      "sent_len_1": 67.0209,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6562,
      "stdk": 0.0482,
      "stdq": 0.0447,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 32200
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4026,
      "doc_norm": 1.3838,
      "encoder_q-embeddings": 1045.9771,
      "encoder_q-layer.0": 715.9929,
      "encoder_q-layer.1": 860.4742,
      "encoder_q-layer.10": 580.7891,
      "encoder_q-layer.11": 1304.5306,
      "encoder_q-layer.2": 1056.8632,
      "encoder_q-layer.3": 1207.7693,
      "encoder_q-layer.4": 1300.9337,
      "encoder_q-layer.5": 1431.6687,
      "encoder_q-layer.6": 1316.4879,
      "encoder_q-layer.7": 1222.7627,
      "encoder_q-layer.8": 835.1311,
      "encoder_q-layer.9": 626.16,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1518,
      "inbatch_pos_score": 0.8047,
      "learning_rate": 3.761111111111111e-05,
      "loss": 3.4026,
      "norm_diff": 0.0149,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1577.8123,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1508,
      "query_norm": 1.3754,
      "queue_k_norm": 1.3821,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6932,
      "sent_len_1": 66.7401,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1712,
      "stdk": 0.0482,
      "stdq": 0.0456,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 32300
    },
    {
      "accuracy": 51.7578,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3901,
      "doc_norm": 1.3888,
      "encoder_q-embeddings": 858.9704,
      "encoder_q-layer.0": 579.0088,
      "encoder_q-layer.1": 634.9401,
      "encoder_q-layer.10": 578.2537,
      "encoder_q-layer.11": 1265.2333,
      "encoder_q-layer.2": 731.7856,
      "encoder_q-layer.3": 849.2943,
      "encoder_q-layer.4": 884.3734,
      "encoder_q-layer.5": 878.6885,
      "encoder_q-layer.6": 1017.0172,
      "encoder_q-layer.7": 964.452,
      "encoder_q-layer.8": 891.7581,
      "encoder_q-layer.9": 656.6426,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1558,
      "inbatch_pos_score": 0.7954,
      "learning_rate": 3.7555555555555554e-05,
      "loss": 3.3901,
      "norm_diff": 0.0103,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1247.8641,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1558,
      "query_norm": 1.3828,
      "queue_k_norm": 1.3829,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8021,
      "sent_len_1": 66.6921,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0037,
      "stdk": 0.0483,
      "stdq": 0.0453,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 32400
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4174,
      "doc_norm": 1.384,
      "encoder_q-embeddings": 565.011,
      "encoder_q-layer.0": 366.8632,
      "encoder_q-layer.1": 394.1293,
      "encoder_q-layer.10": 553.558,
      "encoder_q-layer.11": 1219.5487,
      "encoder_q-layer.2": 452.4711,
      "encoder_q-layer.3": 464.6905,
      "encoder_q-layer.4": 489.0354,
      "encoder_q-layer.5": 506.0476,
      "encoder_q-layer.6": 551.5798,
      "encoder_q-layer.7": 605.963,
      "encoder_q-layer.8": 682.475,
      "encoder_q-layer.9": 583.0072,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1678,
      "inbatch_pos_score": 0.8218,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 3.4174,
      "norm_diff": 0.0168,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 899.5454,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1677,
      "query_norm": 1.3988,
      "queue_k_norm": 1.3844,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9416,
      "sent_len_1": 66.7314,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.475,
      "stdk": 0.0481,
      "stdq": 0.0455,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 32500
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.383,
      "doc_norm": 1.3932,
      "encoder_q-embeddings": 1228.4341,
      "encoder_q-layer.0": 831.584,
      "encoder_q-layer.1": 818.6694,
      "encoder_q-layer.10": 574.4443,
      "encoder_q-layer.11": 1265.1036,
      "encoder_q-layer.2": 961.3421,
      "encoder_q-layer.3": 994.0351,
      "encoder_q-layer.4": 1023.3383,
      "encoder_q-layer.5": 934.7885,
      "encoder_q-layer.6": 1017.669,
      "encoder_q-layer.7": 1017.4345,
      "encoder_q-layer.8": 912.4229,
      "encoder_q-layer.9": 588.221,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1826,
      "inbatch_pos_score": 0.8374,
      "learning_rate": 3.7444444444444446e-05,
      "loss": 3.383,
      "norm_diff": 0.0238,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1462.1262,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1819,
      "query_norm": 1.4165,
      "queue_k_norm": 1.3872,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0163,
      "sent_len_1": 66.9566,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3587,
      "stdk": 0.0484,
      "stdq": 0.0453,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 32600
    },
    {
      "accuracy": 48.3398,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4147,
      "doc_norm": 1.3867,
      "encoder_q-embeddings": 629.8568,
      "encoder_q-layer.0": 401.3384,
      "encoder_q-layer.1": 437.7686,
      "encoder_q-layer.10": 669.0426,
      "encoder_q-layer.11": 1382.1389,
      "encoder_q-layer.2": 491.6976,
      "encoder_q-layer.3": 495.2484,
      "encoder_q-layer.4": 510.4324,
      "encoder_q-layer.5": 530.4768,
      "encoder_q-layer.6": 620.1228,
      "encoder_q-layer.7": 673.8422,
      "encoder_q-layer.8": 796.6137,
      "encoder_q-layer.9": 652.9197,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1898,
      "inbatch_pos_score": 0.8145,
      "learning_rate": 3.738888888888889e-05,
      "loss": 3.4147,
      "norm_diff": 0.0312,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 995.748,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1882,
      "query_norm": 1.4179,
      "queue_k_norm": 1.3875,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7862,
      "sent_len_1": 66.907,
      "sent_max_len_0": 127.9938,
      "sent_max_len_1": 190.8887,
      "stdk": 0.0481,
      "stdq": 0.0454,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 32700
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3845,
      "doc_norm": 1.3872,
      "encoder_q-embeddings": 1094.9276,
      "encoder_q-layer.0": 811.7403,
      "encoder_q-layer.1": 983.8295,
      "encoder_q-layer.10": 610.3517,
      "encoder_q-layer.11": 1298.7789,
      "encoder_q-layer.2": 1190.8508,
      "encoder_q-layer.3": 1091.5809,
      "encoder_q-layer.4": 1107.7832,
      "encoder_q-layer.5": 1004.8062,
      "encoder_q-layer.6": 1085.4924,
      "encoder_q-layer.7": 1025.4967,
      "encoder_q-layer.8": 888.8201,
      "encoder_q-layer.9": 657.7963,
      "epoch": 0.32,
      "inbatch_neg_score": 0.2013,
      "inbatch_pos_score": 0.8462,
      "learning_rate": 3.733333333333334e-05,
      "loss": 3.3845,
      "norm_diff": 0.0491,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1528.0442,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2004,
      "query_norm": 1.4363,
      "queue_k_norm": 1.3903,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7262,
      "sent_len_1": 66.6586,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8525,
      "stdk": 0.0481,
      "stdq": 0.0457,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 32800
    },
    {
      "accuracy": 51.7578,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4436,
      "doc_norm": 1.3905,
      "encoder_q-embeddings": 824.6462,
      "encoder_q-layer.0": 571.1013,
      "encoder_q-layer.1": 627.9854,
      "encoder_q-layer.10": 603.557,
      "encoder_q-layer.11": 1296.4722,
      "encoder_q-layer.2": 739.9105,
      "encoder_q-layer.3": 830.4653,
      "encoder_q-layer.4": 810.952,
      "encoder_q-layer.5": 770.6436,
      "encoder_q-layer.6": 845.1725,
      "encoder_q-layer.7": 742.8347,
      "encoder_q-layer.8": 721.063,
      "encoder_q-layer.9": 616.3286,
      "epoch": 0.32,
      "inbatch_neg_score": 0.2056,
      "inbatch_pos_score": 0.832,
      "learning_rate": 3.727777777777778e-05,
      "loss": 3.4436,
      "norm_diff": 0.0175,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1173.4032,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2041,
      "query_norm": 1.406,
      "queue_k_norm": 1.3912,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9396,
      "sent_len_1": 66.4789,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6375,
      "stdk": 0.0481,
      "stdq": 0.0449,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 32900
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4076,
      "doc_norm": 1.3872,
      "encoder_q-embeddings": 1339.8234,
      "encoder_q-layer.0": 930.6437,
      "encoder_q-layer.1": 964.3276,
      "encoder_q-layer.10": 663.6592,
      "encoder_q-layer.11": 1372.3993,
      "encoder_q-layer.2": 1114.6703,
      "encoder_q-layer.3": 1273.1814,
      "encoder_q-layer.4": 1411.0192,
      "encoder_q-layer.5": 1454.6106,
      "encoder_q-layer.6": 1895.8619,
      "encoder_q-layer.7": 2132.7385,
      "encoder_q-layer.8": 3090.1213,
      "encoder_q-layer.9": 1960.2994,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1973,
      "inbatch_pos_score": 0.8574,
      "learning_rate": 3.722222222222222e-05,
      "loss": 3.4076,
      "norm_diff": 0.0239,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2522.7582,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1973,
      "query_norm": 1.4111,
      "queue_k_norm": 1.3919,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9517,
      "sent_len_1": 66.5999,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.465,
      "stdk": 0.048,
      "stdq": 0.0459,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 33000
    },
    {
      "accuracy": 49.2188,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3998,
      "doc_norm": 1.3898,
      "encoder_q-embeddings": 1680.1329,
      "encoder_q-layer.0": 1136.8539,
      "encoder_q-layer.1": 1149.072,
      "encoder_q-layer.10": 588.304,
      "encoder_q-layer.11": 1349.6948,
      "encoder_q-layer.2": 1255.5181,
      "encoder_q-layer.3": 1265.749,
      "encoder_q-layer.4": 1305.041,
      "encoder_q-layer.5": 1422.4856,
      "encoder_q-layer.6": 1430.2257,
      "encoder_q-layer.7": 1395.866,
      "encoder_q-layer.8": 1109.7488,
      "encoder_q-layer.9": 684.7908,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1993,
      "inbatch_pos_score": 0.8188,
      "learning_rate": 3.7166666666666664e-05,
      "loss": 3.3998,
      "norm_diff": 0.0067,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1912.7182,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1987,
      "query_norm": 1.391,
      "queue_k_norm": 1.394,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7963,
      "sent_len_1": 66.921,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0888,
      "stdk": 0.048,
      "stdq": 0.0454,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 33100
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4186,
      "doc_norm": 1.3959,
      "encoder_q-embeddings": 2693.9114,
      "encoder_q-layer.0": 2053.9436,
      "encoder_q-layer.1": 2604.7986,
      "encoder_q-layer.10": 666.9109,
      "encoder_q-layer.11": 1258.5303,
      "encoder_q-layer.2": 3317.668,
      "encoder_q-layer.3": 3238.4756,
      "encoder_q-layer.4": 3115.3601,
      "encoder_q-layer.5": 2785.9653,
      "encoder_q-layer.6": 2264.1565,
      "encoder_q-layer.7": 1893.5862,
      "encoder_q-layer.8": 1781.5763,
      "encoder_q-layer.9": 1126.2532,
      "epoch": 0.32,
      "inbatch_neg_score": 0.1885,
      "inbatch_pos_score": 0.8203,
      "learning_rate": 3.7111111111111113e-05,
      "loss": 3.4186,
      "norm_diff": 0.0471,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3691.5725,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1895,
      "query_norm": 1.3489,
      "queue_k_norm": 1.3962,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7748,
      "sent_len_1": 66.7442,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5938,
      "stdk": 0.0482,
      "stdq": 0.0444,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 33200
    },
    {
      "accuracy": 50.1953,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4166,
      "doc_norm": 1.3963,
      "encoder_q-embeddings": 545.7358,
      "encoder_q-layer.0": 337.8994,
      "encoder_q-layer.1": 356.7563,
      "encoder_q-layer.10": 634.9484,
      "encoder_q-layer.11": 1346.0634,
      "encoder_q-layer.2": 384.1034,
      "encoder_q-layer.3": 411.9646,
      "encoder_q-layer.4": 435.1906,
      "encoder_q-layer.5": 446.2536,
      "encoder_q-layer.6": 509.8382,
      "encoder_q-layer.7": 590.6722,
      "encoder_q-layer.8": 709.6586,
      "encoder_q-layer.9": 631.4519,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1775,
      "inbatch_pos_score": 0.7988,
      "learning_rate": 3.705555555555556e-05,
      "loss": 3.4166,
      "norm_diff": 0.0387,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 912.8118,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1782,
      "query_norm": 1.3576,
      "queue_k_norm": 1.3964,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8148,
      "sent_len_1": 66.697,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.445,
      "stdk": 0.0482,
      "stdq": 0.0452,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 33300
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4213,
      "doc_norm": 1.4027,
      "encoder_q-embeddings": 1639.1564,
      "encoder_q-layer.0": 1075.3419,
      "encoder_q-layer.1": 1113.1862,
      "encoder_q-layer.10": 619.6888,
      "encoder_q-layer.11": 1239.8955,
      "encoder_q-layer.2": 1220.4054,
      "encoder_q-layer.3": 1145.183,
      "encoder_q-layer.4": 1257.3303,
      "encoder_q-layer.5": 1184.8674,
      "encoder_q-layer.6": 1190.0043,
      "encoder_q-layer.7": 1232.2415,
      "encoder_q-layer.8": 1012.4897,
      "encoder_q-layer.9": 634.3809,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1697,
      "inbatch_pos_score": 0.8477,
      "learning_rate": 3.7e-05,
      "loss": 3.4213,
      "norm_diff": 0.0416,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1776.3753,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1697,
      "query_norm": 1.3611,
      "queue_k_norm": 1.3988,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9616,
      "sent_len_1": 66.6542,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7988,
      "stdk": 0.0484,
      "stdq": 0.0456,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 33400
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4268,
      "doc_norm": 1.396,
      "encoder_q-embeddings": 922.4612,
      "encoder_q-layer.0": 620.6849,
      "encoder_q-layer.1": 700.2497,
      "encoder_q-layer.10": 621.5677,
      "encoder_q-layer.11": 1315.9706,
      "encoder_q-layer.2": 857.8074,
      "encoder_q-layer.3": 929.5607,
      "encoder_q-layer.4": 1019.0856,
      "encoder_q-layer.5": 1040.3019,
      "encoder_q-layer.6": 1114.0416,
      "encoder_q-layer.7": 1188.6166,
      "encoder_q-layer.8": 936.7408,
      "encoder_q-layer.9": 637.7393,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1644,
      "inbatch_pos_score": 0.7783,
      "learning_rate": 3.694444444444445e-05,
      "loss": 3.4268,
      "norm_diff": 0.0759,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1381.5233,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1638,
      "query_norm": 1.3202,
      "queue_k_norm": 1.3973,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.912,
      "sent_len_1": 66.695,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7075,
      "stdk": 0.0482,
      "stdq": 0.0442,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 33500
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3982,
      "doc_norm": 1.4032,
      "encoder_q-embeddings": 560.9611,
      "encoder_q-layer.0": 353.8989,
      "encoder_q-layer.1": 367.1816,
      "encoder_q-layer.10": 653.4078,
      "encoder_q-layer.11": 1396.3828,
      "encoder_q-layer.2": 412.0691,
      "encoder_q-layer.3": 434.0828,
      "encoder_q-layer.4": 463.1119,
      "encoder_q-layer.5": 466.8572,
      "encoder_q-layer.6": 523.8105,
      "encoder_q-layer.7": 615.972,
      "encoder_q-layer.8": 725.4834,
      "encoder_q-layer.9": 666.1911,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1521,
      "inbatch_pos_score": 0.8213,
      "learning_rate": 3.688888888888889e-05,
      "loss": 3.3982,
      "norm_diff": 0.0371,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 942.2325,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1528,
      "query_norm": 1.3661,
      "queue_k_norm": 1.3978,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9515,
      "sent_len_1": 66.6962,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.31,
      "stdk": 0.0485,
      "stdq": 0.0462,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 33600
    },
    {
      "accuracy": 48.8281,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4079,
      "doc_norm": 1.3859,
      "encoder_q-embeddings": 594.8828,
      "encoder_q-layer.0": 377.6842,
      "encoder_q-layer.1": 389.572,
      "encoder_q-layer.10": 666.8322,
      "encoder_q-layer.11": 1377.2577,
      "encoder_q-layer.2": 438.1504,
      "encoder_q-layer.3": 459.1713,
      "encoder_q-layer.4": 475.8571,
      "encoder_q-layer.5": 478.6035,
      "encoder_q-layer.6": 559.501,
      "encoder_q-layer.7": 631.9219,
      "encoder_q-layer.8": 713.808,
      "encoder_q-layer.9": 641.9203,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1435,
      "inbatch_pos_score": 0.7466,
      "learning_rate": 3.683333333333334e-05,
      "loss": 3.4079,
      "norm_diff": 0.0506,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 974.6063,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1427,
      "query_norm": 1.3353,
      "queue_k_norm": 1.3958,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7652,
      "sent_len_1": 66.6305,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6037,
      "stdk": 0.0478,
      "stdq": 0.0452,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 33700
    },
    {
      "accuracy": 60.3516,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3793,
      "doc_norm": 1.4023,
      "encoder_q-embeddings": 648.1539,
      "encoder_q-layer.0": 428.6269,
      "encoder_q-layer.1": 478.5431,
      "encoder_q-layer.10": 570.8197,
      "encoder_q-layer.11": 1235.8611,
      "encoder_q-layer.2": 564.1092,
      "encoder_q-layer.3": 608.7717,
      "encoder_q-layer.4": 584.014,
      "encoder_q-layer.5": 585.4926,
      "encoder_q-layer.6": 658.8486,
      "encoder_q-layer.7": 677.7982,
      "encoder_q-layer.8": 745.0098,
      "encoder_q-layer.9": 655.1516,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1405,
      "inbatch_pos_score": 0.876,
      "learning_rate": 3.677777777777778e-05,
      "loss": 3.3793,
      "norm_diff": 0.0151,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 991.7152,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1407,
      "query_norm": 1.3933,
      "queue_k_norm": 1.3928,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8129,
      "sent_len_1": 66.6719,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.17,
      "stdk": 0.0485,
      "stdq": 0.0472,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 33800
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4057,
      "doc_norm": 1.3945,
      "encoder_q-embeddings": 556.5262,
      "encoder_q-layer.0": 359.4926,
      "encoder_q-layer.1": 391.9268,
      "encoder_q-layer.10": 563.6735,
      "encoder_q-layer.11": 1265.3895,
      "encoder_q-layer.2": 435.2762,
      "encoder_q-layer.3": 467.4,
      "encoder_q-layer.4": 499.6545,
      "encoder_q-layer.5": 504.4809,
      "encoder_q-layer.6": 530.743,
      "encoder_q-layer.7": 604.5628,
      "encoder_q-layer.8": 693.5175,
      "encoder_q-layer.9": 590.9453,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1403,
      "inbatch_pos_score": 0.7935,
      "learning_rate": 3.672222222222222e-05,
      "loss": 3.4057,
      "norm_diff": 0.0509,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 921.1366,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1406,
      "query_norm": 1.3436,
      "queue_k_norm": 1.3925,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8119,
      "sent_len_1": 66.5999,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6912,
      "stdk": 0.0482,
      "stdq": 0.0453,
      "stdqueue_k": 0.0482,
      "stdqueue_q": 0.0,
      "step": 33900
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3831,
      "doc_norm": 1.3969,
      "encoder_q-embeddings": 909.8882,
      "encoder_q-layer.0": 623.6838,
      "encoder_q-layer.1": 624.4474,
      "encoder_q-layer.10": 560.4666,
      "encoder_q-layer.11": 1260.9718,
      "encoder_q-layer.2": 684.6583,
      "encoder_q-layer.3": 683.2859,
      "encoder_q-layer.4": 716.6005,
      "encoder_q-layer.5": 670.4549,
      "encoder_q-layer.6": 728.9794,
      "encoder_q-layer.7": 731.7294,
      "encoder_q-layer.8": 764.7271,
      "encoder_q-layer.9": 610.1914,
      "epoch": 0.33,
      "inbatch_neg_score": 0.136,
      "inbatch_pos_score": 0.8027,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 3.3831,
      "norm_diff": 0.0413,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1144.9833,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1362,
      "query_norm": 1.3556,
      "queue_k_norm": 1.3932,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.763,
      "sent_len_1": 66.7077,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9913,
      "stdk": 0.0483,
      "stdq": 0.0457,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 34000
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4059,
      "doc_norm": 1.3892,
      "encoder_q-embeddings": 1146.7867,
      "encoder_q-layer.0": 802.8483,
      "encoder_q-layer.1": 813.4438,
      "encoder_q-layer.10": 579.0887,
      "encoder_q-layer.11": 1226.0988,
      "encoder_q-layer.2": 999.5176,
      "encoder_q-layer.3": 971.2878,
      "encoder_q-layer.4": 1010.9708,
      "encoder_q-layer.5": 1082.434,
      "encoder_q-layer.6": 1049.0164,
      "encoder_q-layer.7": 836.7924,
      "encoder_q-layer.8": 803.6633,
      "encoder_q-layer.9": 667.7416,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1373,
      "inbatch_pos_score": 0.8135,
      "learning_rate": 3.6611111111111115e-05,
      "loss": 3.4059,
      "norm_diff": 0.0289,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1405.9494,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.137,
      "query_norm": 1.3604,
      "queue_k_norm": 1.393,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8403,
      "sent_len_1": 66.5894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.94,
      "stdk": 0.0481,
      "stdq": 0.0458,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 34100
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4018,
      "doc_norm": 1.3942,
      "encoder_q-embeddings": 1550.0554,
      "encoder_q-layer.0": 1168.3311,
      "encoder_q-layer.1": 1157.5745,
      "encoder_q-layer.10": 650.4166,
      "encoder_q-layer.11": 1296.9515,
      "encoder_q-layer.2": 1452.1879,
      "encoder_q-layer.3": 1500.3707,
      "encoder_q-layer.4": 1556.8416,
      "encoder_q-layer.5": 1691.9148,
      "encoder_q-layer.6": 1888.8674,
      "encoder_q-layer.7": 1725.902,
      "encoder_q-layer.8": 1732.42,
      "encoder_q-layer.9": 1044.017,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1369,
      "inbatch_pos_score": 0.8066,
      "learning_rate": 3.655555555555556e-05,
      "loss": 3.4018,
      "norm_diff": 0.0544,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2201.3351,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1375,
      "query_norm": 1.3398,
      "queue_k_norm": 1.3923,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8875,
      "sent_len_1": 66.5821,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.7475,
      "stdk": 0.0483,
      "stdq": 0.0451,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 34200
    },
    {
      "accuracy": 49.707,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4014,
      "doc_norm": 1.393,
      "encoder_q-embeddings": 2536.0413,
      "encoder_q-layer.0": 1681.2228,
      "encoder_q-layer.1": 1611.538,
      "encoder_q-layer.10": 619.4867,
      "encoder_q-layer.11": 1356.873,
      "encoder_q-layer.2": 2037.8395,
      "encoder_q-layer.3": 1892.2881,
      "encoder_q-layer.4": 2179.3889,
      "encoder_q-layer.5": 2108.5686,
      "encoder_q-layer.6": 2290.4651,
      "encoder_q-layer.7": 2256.0938,
      "encoder_q-layer.8": 2065.5803,
      "encoder_q-layer.9": 1057.7574,
      "epoch": 0.33,
      "inbatch_neg_score": 0.1439,
      "inbatch_pos_score": 0.7397,
      "learning_rate": 3.65e-05,
      "loss": 3.4014,
      "norm_diff": 0.0655,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2950.471,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1444,
      "query_norm": 1.3275,
      "queue_k_norm": 1.3926,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0109,
      "sent_len_1": 66.7444,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6113,
      "stdk": 0.0483,
      "stdq": 0.0444,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 34300
    },
    {
      "accuracy": 51.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3791,
      "doc_norm": 1.3968,
      "encoder_q-embeddings": 2436.189,
      "encoder_q-layer.0": 1703.7983,
      "encoder_q-layer.1": 1867.972,
      "encoder_q-layer.10": 614.3889,
      "encoder_q-layer.11": 1322.6165,
      "encoder_q-layer.2": 2254.2964,
      "encoder_q-layer.3": 2225.7063,
      "encoder_q-layer.4": 2346.5093,
      "encoder_q-layer.5": 2576.4668,
      "encoder_q-layer.6": 2475.2341,
      "encoder_q-layer.7": 2324.4451,
      "encoder_q-layer.8": 1819.9364,
      "encoder_q-layer.9": 764.8059,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1451,
      "inbatch_pos_score": 0.7852,
      "learning_rate": 3.644444444444445e-05,
      "loss": 3.3791,
      "norm_diff": 0.039,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3027.0497,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1443,
      "query_norm": 1.3578,
      "queue_k_norm": 1.393,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0235,
      "sent_len_1": 66.8909,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2325,
      "stdk": 0.0484,
      "stdq": 0.0454,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 34400
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4129,
      "doc_norm": 1.3873,
      "encoder_q-embeddings": 720.245,
      "encoder_q-layer.0": 457.8578,
      "encoder_q-layer.1": 466.8432,
      "encoder_q-layer.10": 634.5436,
      "encoder_q-layer.11": 1351.1941,
      "encoder_q-layer.2": 492.4396,
      "encoder_q-layer.3": 521.1402,
      "encoder_q-layer.4": 639.189,
      "encoder_q-layer.5": 614.4017,
      "encoder_q-layer.6": 659.8909,
      "encoder_q-layer.7": 749.6517,
      "encoder_q-layer.8": 755.2548,
      "encoder_q-layer.9": 634.2066,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1414,
      "inbatch_pos_score": 0.7705,
      "learning_rate": 3.638888888888889e-05,
      "loss": 3.4129,
      "norm_diff": 0.0239,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1051.9361,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.142,
      "query_norm": 1.3634,
      "queue_k_norm": 1.3906,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8546,
      "sent_len_1": 66.7978,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8837,
      "stdk": 0.0481,
      "stdq": 0.0458,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 34500
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3727,
      "doc_norm": 1.3939,
      "encoder_q-embeddings": 1355.166,
      "encoder_q-layer.0": 898.5949,
      "encoder_q-layer.1": 1024.7487,
      "encoder_q-layer.10": 642.9348,
      "encoder_q-layer.11": 1312.17,
      "encoder_q-layer.2": 1191.6672,
      "encoder_q-layer.3": 1318.4622,
      "encoder_q-layer.4": 1320.6699,
      "encoder_q-layer.5": 1198.7212,
      "encoder_q-layer.6": 1216.9335,
      "encoder_q-layer.7": 1328.1207,
      "encoder_q-layer.8": 1080.4872,
      "encoder_q-layer.9": 680.5975,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1439,
      "inbatch_pos_score": 0.8052,
      "learning_rate": 3.633333333333333e-05,
      "loss": 3.3727,
      "norm_diff": 0.0259,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1720.2252,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1432,
      "query_norm": 1.3704,
      "queue_k_norm": 1.3933,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0843,
      "sent_len_1": 66.6471,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.83,
      "stdk": 0.0483,
      "stdq": 0.0458,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 34600
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3597,
      "doc_norm": 1.3924,
      "encoder_q-embeddings": 1777.011,
      "encoder_q-layer.0": 1393.0509,
      "encoder_q-layer.1": 1425.931,
      "encoder_q-layer.10": 574.5323,
      "encoder_q-layer.11": 1289.3749,
      "encoder_q-layer.2": 1448.4294,
      "encoder_q-layer.3": 1179.7618,
      "encoder_q-layer.4": 1138.3582,
      "encoder_q-layer.5": 975.7834,
      "encoder_q-layer.6": 1043.9279,
      "encoder_q-layer.7": 1109.8102,
      "encoder_q-layer.8": 765.6638,
      "encoder_q-layer.9": 595.572,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1419,
      "inbatch_pos_score": 0.7881,
      "learning_rate": 3.6277777777777776e-05,
      "loss": 3.3597,
      "norm_diff": 0.0192,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1845.761,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1418,
      "query_norm": 1.3732,
      "queue_k_norm": 1.391,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1332,
      "sent_len_1": 66.7381,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.265,
      "stdk": 0.0483,
      "stdq": 0.0458,
      "stdqueue_k": 0.0483,
      "stdqueue_q": 0.0,
      "step": 34700
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4114,
      "doc_norm": 1.3881,
      "encoder_q-embeddings": 692.049,
      "encoder_q-layer.0": 445.1454,
      "encoder_q-layer.1": 489.2068,
      "encoder_q-layer.10": 595.0972,
      "encoder_q-layer.11": 1295.5333,
      "encoder_q-layer.2": 555.8281,
      "encoder_q-layer.3": 580.0585,
      "encoder_q-layer.4": 631.0938,
      "encoder_q-layer.5": 601.1688,
      "encoder_q-layer.6": 670.2349,
      "encoder_q-layer.7": 713.8792,
      "encoder_q-layer.8": 721.2831,
      "encoder_q-layer.9": 601.1348,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1362,
      "inbatch_pos_score": 0.7739,
      "learning_rate": 3.6222222222222225e-05,
      "loss": 3.4114,
      "norm_diff": 0.0305,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1032.4639,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1365,
      "query_norm": 1.3577,
      "queue_k_norm": 1.3921,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7111,
      "sent_len_1": 66.9413,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7775,
      "stdk": 0.0482,
      "stdq": 0.0451,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 34800
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3528,
      "doc_norm": 1.3887,
      "encoder_q-embeddings": 1217.3873,
      "encoder_q-layer.0": 823.319,
      "encoder_q-layer.1": 897.4232,
      "encoder_q-layer.10": 592.9701,
      "encoder_q-layer.11": 1307.7307,
      "encoder_q-layer.2": 990.0146,
      "encoder_q-layer.3": 1038.5233,
      "encoder_q-layer.4": 1195.994,
      "encoder_q-layer.5": 1300.6631,
      "encoder_q-layer.6": 1237.2614,
      "encoder_q-layer.7": 1249.8456,
      "encoder_q-layer.8": 1065.754,
      "encoder_q-layer.9": 675.2884,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1372,
      "inbatch_pos_score": 0.7754,
      "learning_rate": 3.6166666666666674e-05,
      "loss": 3.3528,
      "norm_diff": 0.0258,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1609.9695,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1375,
      "query_norm": 1.3629,
      "queue_k_norm": 1.3913,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9903,
      "sent_len_1": 66.834,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.865,
      "stdk": 0.0482,
      "stdq": 0.0452,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 34900
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3933,
      "doc_norm": 1.3952,
      "encoder_q-embeddings": 620.4772,
      "encoder_q-layer.0": 417.0194,
      "encoder_q-layer.1": 441.6933,
      "encoder_q-layer.10": 650.9056,
      "encoder_q-layer.11": 1314.1658,
      "encoder_q-layer.2": 496.8109,
      "encoder_q-layer.3": 529.6899,
      "encoder_q-layer.4": 638.7408,
      "encoder_q-layer.5": 615.7935,
      "encoder_q-layer.6": 666.3781,
      "encoder_q-layer.7": 715.643,
      "encoder_q-layer.8": 739.5762,
      "encoder_q-layer.9": 633.0247,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1374,
      "inbatch_pos_score": 0.7661,
      "learning_rate": 3.611111111111111e-05,
      "loss": 3.3933,
      "norm_diff": 0.0233,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1013.5027,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1372,
      "query_norm": 1.3719,
      "queue_k_norm": 1.3897,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8516,
      "sent_len_1": 66.6141,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.345,
      "stdk": 0.0485,
      "stdq": 0.0456,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35000
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3623,
      "doc_norm": 1.3949,
      "encoder_q-embeddings": 989.9117,
      "encoder_q-layer.0": 653.278,
      "encoder_q-layer.1": 773.6516,
      "encoder_q-layer.10": 567.9959,
      "encoder_q-layer.11": 1262.2476,
      "encoder_q-layer.2": 903.7637,
      "encoder_q-layer.3": 1058.3367,
      "encoder_q-layer.4": 1172.9805,
      "encoder_q-layer.5": 1081.7161,
      "encoder_q-layer.6": 1044.8942,
      "encoder_q-layer.7": 996.5469,
      "encoder_q-layer.8": 813.5123,
      "encoder_q-layer.9": 627.5322,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1382,
      "inbatch_pos_score": 0.8208,
      "learning_rate": 3.605555555555556e-05,
      "loss": 3.3623,
      "norm_diff": 0.0179,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1404.9201,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1384,
      "query_norm": 1.3788,
      "queue_k_norm": 1.3892,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.002,
      "sent_len_1": 66.7369,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8988,
      "stdk": 0.0485,
      "stdq": 0.0458,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35100
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3932,
      "doc_norm": 1.3881,
      "encoder_q-embeddings": 637.3997,
      "encoder_q-layer.0": 408.1852,
      "encoder_q-layer.1": 438.2856,
      "encoder_q-layer.10": 585.7705,
      "encoder_q-layer.11": 1301.5605,
      "encoder_q-layer.2": 496.0449,
      "encoder_q-layer.3": 513.157,
      "encoder_q-layer.4": 566.3279,
      "encoder_q-layer.5": 616.3674,
      "encoder_q-layer.6": 667.5393,
      "encoder_q-layer.7": 688.881,
      "encoder_q-layer.8": 728.4999,
      "encoder_q-layer.9": 643.5395,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1367,
      "inbatch_pos_score": 0.8071,
      "learning_rate": 3.6e-05,
      "loss": 3.3932,
      "norm_diff": 0.0107,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 995.5181,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1372,
      "query_norm": 1.3929,
      "queue_k_norm": 1.3892,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9773,
      "sent_len_1": 66.858,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2675,
      "stdk": 0.0482,
      "stdq": 0.0464,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35200
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3811,
      "doc_norm": 1.3929,
      "encoder_q-embeddings": 598.6394,
      "encoder_q-layer.0": 384.9061,
      "encoder_q-layer.1": 417.3241,
      "encoder_q-layer.10": 644.1778,
      "encoder_q-layer.11": 1329.4602,
      "encoder_q-layer.2": 467.5759,
      "encoder_q-layer.3": 490.3026,
      "encoder_q-layer.4": 525.1432,
      "encoder_q-layer.5": 555.5823,
      "encoder_q-layer.6": 628.8047,
      "encoder_q-layer.7": 747.0192,
      "encoder_q-layer.8": 786.6054,
      "encoder_q-layer.9": 661.5988,
      "epoch": 0.34,
      "inbatch_neg_score": 0.1457,
      "inbatch_pos_score": 0.7954,
      "learning_rate": 3.594444444444445e-05,
      "loss": 3.3811,
      "norm_diff": 0.0125,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 991.4518,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1445,
      "query_norm": 1.3999,
      "queue_k_norm": 1.3904,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9441,
      "sent_len_1": 66.7577,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6125,
      "stdk": 0.0484,
      "stdq": 0.0466,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35300
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3714,
      "doc_norm": 1.3946,
      "encoder_q-embeddings": 611.9994,
      "encoder_q-layer.0": 403.8955,
      "encoder_q-layer.1": 435.6125,
      "encoder_q-layer.10": 581.7361,
      "encoder_q-layer.11": 1306.1599,
      "encoder_q-layer.2": 485.0737,
      "encoder_q-layer.3": 507.5988,
      "encoder_q-layer.4": 532.6588,
      "encoder_q-layer.5": 542.8952,
      "encoder_q-layer.6": 559.9435,
      "encoder_q-layer.7": 620.6491,
      "encoder_q-layer.8": 677.7469,
      "encoder_q-layer.9": 581.038,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1441,
      "inbatch_pos_score": 0.7754,
      "learning_rate": 3.5888888888888886e-05,
      "loss": 3.3714,
      "norm_diff": 0.0399,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 967.682,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1434,
      "query_norm": 1.3546,
      "queue_k_norm": 1.3911,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2641,
      "sent_len_1": 67.2515,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7188,
      "stdk": 0.0485,
      "stdq": 0.0449,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 35400
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3929,
      "doc_norm": 1.39,
      "encoder_q-embeddings": 878.7101,
      "encoder_q-layer.0": 554.7265,
      "encoder_q-layer.1": 653.4938,
      "encoder_q-layer.10": 673.9153,
      "encoder_q-layer.11": 1317.3329,
      "encoder_q-layer.2": 727.6923,
      "encoder_q-layer.3": 801.3364,
      "encoder_q-layer.4": 825.9936,
      "encoder_q-layer.5": 863.0383,
      "encoder_q-layer.6": 890.3412,
      "encoder_q-layer.7": 902.1573,
      "encoder_q-layer.8": 800.3546,
      "encoder_q-layer.9": 665.1306,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1435,
      "inbatch_pos_score": 0.7974,
      "learning_rate": 3.5833333333333335e-05,
      "loss": 3.3929,
      "norm_diff": 0.0147,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1222.9381,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1431,
      "query_norm": 1.3754,
      "queue_k_norm": 1.3908,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8831,
      "sent_len_1": 66.8347,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7763,
      "stdk": 0.0483,
      "stdq": 0.0458,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35500
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4049,
      "doc_norm": 1.387,
      "encoder_q-embeddings": 848.1967,
      "encoder_q-layer.0": 604.9983,
      "encoder_q-layer.1": 671.1117,
      "encoder_q-layer.10": 750.4475,
      "encoder_q-layer.11": 1448.7141,
      "encoder_q-layer.2": 800.2543,
      "encoder_q-layer.3": 771.0775,
      "encoder_q-layer.4": 804.3008,
      "encoder_q-layer.5": 849.6263,
      "encoder_q-layer.6": 942.2329,
      "encoder_q-layer.7": 912.6399,
      "encoder_q-layer.8": 893.541,
      "encoder_q-layer.9": 724.5931,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1463,
      "inbatch_pos_score": 0.7959,
      "learning_rate": 3.577777777777778e-05,
      "loss": 3.4049,
      "norm_diff": 0.0089,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1267.5334,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1464,
      "query_norm": 1.3832,
      "queue_k_norm": 1.3906,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9172,
      "sent_len_1": 66.6619,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9475,
      "stdk": 0.0482,
      "stdq": 0.0464,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35600
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3967,
      "doc_norm": 1.3873,
      "encoder_q-embeddings": 1472.4036,
      "encoder_q-layer.0": 966.7648,
      "encoder_q-layer.1": 1080.423,
      "encoder_q-layer.10": 1246.0884,
      "encoder_q-layer.11": 2621.3645,
      "encoder_q-layer.2": 1271.0217,
      "encoder_q-layer.3": 1383.9537,
      "encoder_q-layer.4": 1540.5616,
      "encoder_q-layer.5": 1629.6245,
      "encoder_q-layer.6": 1676.8323,
      "encoder_q-layer.7": 1732.4371,
      "encoder_q-layer.8": 1485.1727,
      "encoder_q-layer.9": 1291.5747,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1493,
      "inbatch_pos_score": 0.8081,
      "learning_rate": 3.5722222222222226e-05,
      "loss": 3.3967,
      "norm_diff": 0.0339,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2266.2558,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1498,
      "query_norm": 1.3534,
      "queue_k_norm": 1.3902,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0012,
      "sent_len_1": 66.5431,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.255,
      "stdk": 0.0482,
      "stdq": 0.0454,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35700
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3892,
      "doc_norm": 1.3901,
      "encoder_q-embeddings": 1299.1021,
      "encoder_q-layer.0": 841.4662,
      "encoder_q-layer.1": 913.3298,
      "encoder_q-layer.10": 1322.8895,
      "encoder_q-layer.11": 2754.7769,
      "encoder_q-layer.2": 1040.2814,
      "encoder_q-layer.3": 1098.7607,
      "encoder_q-layer.4": 1111.1327,
      "encoder_q-layer.5": 1228.3777,
      "encoder_q-layer.6": 1363.3425,
      "encoder_q-layer.7": 1520.9169,
      "encoder_q-layer.8": 1556.208,
      "encoder_q-layer.9": 1287.0082,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1527,
      "inbatch_pos_score": 0.811,
      "learning_rate": 3.566666666666667e-05,
      "loss": 3.3892,
      "norm_diff": 0.0419,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2088.2683,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1528,
      "query_norm": 1.3482,
      "queue_k_norm": 1.391,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7739,
      "sent_len_1": 66.8155,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7738,
      "stdk": 0.0483,
      "stdq": 0.0452,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35800
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3727,
      "doc_norm": 1.3884,
      "encoder_q-embeddings": 1217.8625,
      "encoder_q-layer.0": 778.105,
      "encoder_q-layer.1": 829.3466,
      "encoder_q-layer.10": 1183.3497,
      "encoder_q-layer.11": 2744.6084,
      "encoder_q-layer.2": 914.1365,
      "encoder_q-layer.3": 951.3224,
      "encoder_q-layer.4": 1011.2867,
      "encoder_q-layer.5": 1056.6832,
      "encoder_q-layer.6": 1170.2551,
      "encoder_q-layer.7": 1351.106,
      "encoder_q-layer.8": 1494.1653,
      "encoder_q-layer.9": 1223.3109,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1625,
      "inbatch_pos_score": 0.8203,
      "learning_rate": 3.561111111111111e-05,
      "loss": 3.3727,
      "norm_diff": 0.03,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1974.9801,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1626,
      "query_norm": 1.3585,
      "queue_k_norm": 1.3916,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6128,
      "sent_len_1": 66.6062,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2237,
      "stdk": 0.0482,
      "stdq": 0.0453,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 35900
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3973,
      "doc_norm": 1.3856,
      "encoder_q-embeddings": 1193.5861,
      "encoder_q-layer.0": 780.3847,
      "encoder_q-layer.1": 832.9117,
      "encoder_q-layer.10": 1163.0331,
      "encoder_q-layer.11": 2532.3857,
      "encoder_q-layer.2": 932.3046,
      "encoder_q-layer.3": 940.0674,
      "encoder_q-layer.4": 978.4357,
      "encoder_q-layer.5": 1004.1297,
      "encoder_q-layer.6": 1126.5623,
      "encoder_q-layer.7": 1196.5848,
      "encoder_q-layer.8": 1403.5001,
      "encoder_q-layer.9": 1233.7643,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1541,
      "inbatch_pos_score": 0.791,
      "learning_rate": 3.555555555555556e-05,
      "loss": 3.3973,
      "norm_diff": 0.0351,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1850.21,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1545,
      "query_norm": 1.3505,
      "queue_k_norm": 1.3926,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1639,
      "sent_len_1": 67.0406,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.575,
      "stdk": 0.0481,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 36000
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.351,
      "doc_norm": 1.3833,
      "encoder_q-embeddings": 1395.9885,
      "encoder_q-layer.0": 940.4857,
      "encoder_q-layer.1": 1081.5204,
      "encoder_q-layer.10": 1150.4878,
      "encoder_q-layer.11": 2544.6316,
      "encoder_q-layer.2": 1234.134,
      "encoder_q-layer.3": 1321.9469,
      "encoder_q-layer.4": 1342.7587,
      "encoder_q-layer.5": 1274.4357,
      "encoder_q-layer.6": 1412.9255,
      "encoder_q-layer.7": 1564.0786,
      "encoder_q-layer.8": 1538.6431,
      "encoder_q-layer.9": 1226.4604,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1593,
      "inbatch_pos_score": 0.8535,
      "learning_rate": 3.55e-05,
      "loss": 3.351,
      "norm_diff": 0.01,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2151.5906,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1602,
      "query_norm": 1.3778,
      "queue_k_norm": 1.3907,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.617,
      "sent_len_1": 66.6937,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.5325,
      "stdk": 0.048,
      "stdq": 0.046,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36100
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3625,
      "doc_norm": 1.392,
      "encoder_q-embeddings": 2235.291,
      "encoder_q-layer.0": 1533.2528,
      "encoder_q-layer.1": 1742.4832,
      "encoder_q-layer.10": 1228.7523,
      "encoder_q-layer.11": 2671.5713,
      "encoder_q-layer.2": 2082.3635,
      "encoder_q-layer.3": 2289.5659,
      "encoder_q-layer.4": 2266.2512,
      "encoder_q-layer.5": 2018.6365,
      "encoder_q-layer.6": 1876.689,
      "encoder_q-layer.7": 1448.6658,
      "encoder_q-layer.8": 1437.2837,
      "encoder_q-layer.9": 1215.7932,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1666,
      "inbatch_pos_score": 0.8096,
      "learning_rate": 3.5444444444444445e-05,
      "loss": 3.3625,
      "norm_diff": 0.0276,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2842.0169,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1682,
      "query_norm": 1.3643,
      "queue_k_norm": 1.3918,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8821,
      "sent_len_1": 66.7437,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6225,
      "stdk": 0.0483,
      "stdq": 0.0455,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36200
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3727,
      "doc_norm": 1.3968,
      "encoder_q-embeddings": 1421.1467,
      "encoder_q-layer.0": 1027.3617,
      "encoder_q-layer.1": 1151.2808,
      "encoder_q-layer.10": 1191.5364,
      "encoder_q-layer.11": 2540.3323,
      "encoder_q-layer.2": 1309.4911,
      "encoder_q-layer.3": 1383.5955,
      "encoder_q-layer.4": 1459.4536,
      "encoder_q-layer.5": 1602.1801,
      "encoder_q-layer.6": 1563.5477,
      "encoder_q-layer.7": 1394.5592,
      "encoder_q-layer.8": 1439.9156,
      "encoder_q-layer.9": 1188.9688,
      "epoch": 0.35,
      "inbatch_neg_score": 0.1588,
      "inbatch_pos_score": 0.7969,
      "learning_rate": 3.538888888888889e-05,
      "loss": 3.3727,
      "norm_diff": 0.0482,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2180.5406,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1595,
      "query_norm": 1.3486,
      "queue_k_norm": 1.3924,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1546,
      "sent_len_1": 66.679,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8325,
      "stdk": 0.0485,
      "stdq": 0.0453,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36300
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3769,
      "doc_norm": 1.394,
      "encoder_q-embeddings": 5620.0376,
      "encoder_q-layer.0": 4099.4043,
      "encoder_q-layer.1": 5281.9531,
      "encoder_q-layer.10": 1284.9463,
      "encoder_q-layer.11": 2384.7742,
      "encoder_q-layer.2": 6747.7456,
      "encoder_q-layer.3": 7265.6392,
      "encoder_q-layer.4": 7309.7495,
      "encoder_q-layer.5": 7476.3276,
      "encoder_q-layer.6": 7609.4556,
      "encoder_q-layer.7": 6979.6953,
      "encoder_q-layer.8": 5373.082,
      "encoder_q-layer.9": 3001.9534,
      "epoch": 0.36,
      "inbatch_neg_score": 0.157,
      "inbatch_pos_score": 0.8296,
      "learning_rate": 3.5333333333333336e-05,
      "loss": 3.3769,
      "norm_diff": 0.0309,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8755.6984,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.1566,
      "query_norm": 1.3631,
      "queue_k_norm": 1.3919,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9887,
      "sent_len_1": 66.5346,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5037,
      "stdk": 0.0483,
      "stdq": 0.0458,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36400
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3593,
      "doc_norm": 1.3977,
      "encoder_q-embeddings": 1278.7057,
      "encoder_q-layer.0": 853.7161,
      "encoder_q-layer.1": 937.1256,
      "encoder_q-layer.10": 1211.6221,
      "encoder_q-layer.11": 2547.2578,
      "encoder_q-layer.2": 1094.8828,
      "encoder_q-layer.3": 1194.8397,
      "encoder_q-layer.4": 1275.5389,
      "encoder_q-layer.5": 1269.1602,
      "encoder_q-layer.6": 1293.538,
      "encoder_q-layer.7": 1402.788,
      "encoder_q-layer.8": 1601.3129,
      "encoder_q-layer.9": 1280.1283,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1553,
      "inbatch_pos_score": 0.835,
      "learning_rate": 3.527777777777778e-05,
      "loss": 3.3593,
      "norm_diff": 0.0365,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2053.0315,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1566,
      "query_norm": 1.3612,
      "queue_k_norm": 1.3941,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8932,
      "sent_len_1": 66.7556,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7663,
      "stdk": 0.0485,
      "stdq": 0.0458,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36500
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.356,
      "doc_norm": 1.3997,
      "encoder_q-embeddings": 1554.4653,
      "encoder_q-layer.0": 1075.8995,
      "encoder_q-layer.1": 1152.5884,
      "encoder_q-layer.10": 1132.1792,
      "encoder_q-layer.11": 2549.5208,
      "encoder_q-layer.2": 1302.8918,
      "encoder_q-layer.3": 1445.0848,
      "encoder_q-layer.4": 1447.0691,
      "encoder_q-layer.5": 1552.8007,
      "encoder_q-layer.6": 1585.234,
      "encoder_q-layer.7": 1723.1438,
      "encoder_q-layer.8": 1481.9688,
      "encoder_q-layer.9": 1250.4227,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1561,
      "inbatch_pos_score": 0.8257,
      "learning_rate": 3.522222222222222e-05,
      "loss": 3.356,
      "norm_diff": 0.0316,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2256.1298,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1564,
      "query_norm": 1.3681,
      "queue_k_norm": 1.3945,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0163,
      "sent_len_1": 67.033,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0987,
      "stdk": 0.0485,
      "stdq": 0.046,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36600
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3728,
      "doc_norm": 1.3946,
      "encoder_q-embeddings": 1143.729,
      "encoder_q-layer.0": 732.0656,
      "encoder_q-layer.1": 755.6237,
      "encoder_q-layer.10": 1205.4191,
      "encoder_q-layer.11": 2484.937,
      "encoder_q-layer.2": 878.4327,
      "encoder_q-layer.3": 893.7133,
      "encoder_q-layer.4": 934.9391,
      "encoder_q-layer.5": 975.4617,
      "encoder_q-layer.6": 1088.2228,
      "encoder_q-layer.7": 1218.7886,
      "encoder_q-layer.8": 1350.819,
      "encoder_q-layer.9": 1224.9618,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1577,
      "inbatch_pos_score": 0.8125,
      "learning_rate": 3.516666666666667e-05,
      "loss": 3.3728,
      "norm_diff": 0.0476,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1824.5662,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1578,
      "query_norm": 1.3471,
      "queue_k_norm": 1.3938,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8197,
      "sent_len_1": 66.5797,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.92,
      "stdk": 0.0483,
      "stdq": 0.0454,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36700
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3743,
      "doc_norm": 1.3821,
      "encoder_q-embeddings": 1468.1963,
      "encoder_q-layer.0": 957.1272,
      "encoder_q-layer.1": 1049.1016,
      "encoder_q-layer.10": 1140.1708,
      "encoder_q-layer.11": 2666.9216,
      "encoder_q-layer.2": 1184.3757,
      "encoder_q-layer.3": 1255.2621,
      "encoder_q-layer.4": 1276.0186,
      "encoder_q-layer.5": 1253.7938,
      "encoder_q-layer.6": 1260.2168,
      "encoder_q-layer.7": 1327.3105,
      "encoder_q-layer.8": 1424.1796,
      "encoder_q-layer.9": 1205.4591,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1603,
      "inbatch_pos_score": 0.7788,
      "learning_rate": 3.511111111111111e-05,
      "loss": 3.3743,
      "norm_diff": 0.0538,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2085.873,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1602,
      "query_norm": 1.3282,
      "queue_k_norm": 1.3942,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8823,
      "sent_len_1": 66.9211,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4412,
      "stdk": 0.0479,
      "stdq": 0.0445,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 36800
    },
    {
      "accuracy": 50.293,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3731,
      "doc_norm": 1.394,
      "encoder_q-embeddings": 1407.6649,
      "encoder_q-layer.0": 916.9797,
      "encoder_q-layer.1": 975.7659,
      "encoder_q-layer.10": 1377.6333,
      "encoder_q-layer.11": 2716.3037,
      "encoder_q-layer.2": 1080.391,
      "encoder_q-layer.3": 1109.2002,
      "encoder_q-layer.4": 1199.3307,
      "encoder_q-layer.5": 1240.8937,
      "encoder_q-layer.6": 1348.7174,
      "encoder_q-layer.7": 1525.8695,
      "encoder_q-layer.8": 1685.5297,
      "encoder_q-layer.9": 1347.9808,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1644,
      "inbatch_pos_score": 0.7812,
      "learning_rate": 3.505555555555556e-05,
      "loss": 3.3731,
      "norm_diff": 0.0373,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2161.6592,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1643,
      "query_norm": 1.3567,
      "queue_k_norm": 1.3955,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0322,
      "sent_len_1": 67.0993,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0925,
      "stdk": 0.0483,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 36900
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3524,
      "doc_norm": 1.3934,
      "encoder_q-embeddings": 2701.4717,
      "encoder_q-layer.0": 2026.978,
      "encoder_q-layer.1": 2145.5005,
      "encoder_q-layer.10": 1223.9055,
      "encoder_q-layer.11": 2582.2344,
      "encoder_q-layer.2": 2528.4521,
      "encoder_q-layer.3": 2544.1775,
      "encoder_q-layer.4": 2643.926,
      "encoder_q-layer.5": 2719.364,
      "encoder_q-layer.6": 2531.0222,
      "encoder_q-layer.7": 2559.0232,
      "encoder_q-layer.8": 2588.9238,
      "encoder_q-layer.9": 1581.9308,
      "epoch": 0.36,
      "inbatch_neg_score": 0.167,
      "inbatch_pos_score": 0.8184,
      "learning_rate": 3.5e-05,
      "loss": 3.3524,
      "norm_diff": 0.0154,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3544.3004,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1671,
      "query_norm": 1.3817,
      "queue_k_norm": 1.3956,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1377,
      "sent_len_1": 67.2378,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.815,
      "stdk": 0.0483,
      "stdq": 0.0462,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37000
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3451,
      "doc_norm": 1.4032,
      "encoder_q-embeddings": 3514.8645,
      "encoder_q-layer.0": 2318.7422,
      "encoder_q-layer.1": 2701.3425,
      "encoder_q-layer.10": 1136.5917,
      "encoder_q-layer.11": 2453.5234,
      "encoder_q-layer.2": 3137.2991,
      "encoder_q-layer.3": 3360.821,
      "encoder_q-layer.4": 3638.2588,
      "encoder_q-layer.5": 3825.6003,
      "encoder_q-layer.6": 4037.2273,
      "encoder_q-layer.7": 4039.6006,
      "encoder_q-layer.8": 3414.8413,
      "encoder_q-layer.9": 2059.2605,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1654,
      "inbatch_pos_score": 0.8374,
      "learning_rate": 3.4944444444444446e-05,
      "loss": 3.3451,
      "norm_diff": 0.0498,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4748.6587,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.166,
      "query_norm": 1.3535,
      "queue_k_norm": 1.3972,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7709,
      "sent_len_1": 66.983,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6425,
      "stdk": 0.0486,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37100
    },
    {
      "accuracy": 51.7578,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3496,
      "doc_norm": 1.4008,
      "encoder_q-embeddings": 1625.7239,
      "encoder_q-layer.0": 1120.3136,
      "encoder_q-layer.1": 1208.7566,
      "encoder_q-layer.10": 1173.677,
      "encoder_q-layer.11": 2647.6619,
      "encoder_q-layer.2": 1465.4745,
      "encoder_q-layer.3": 1476.996,
      "encoder_q-layer.4": 1479.8625,
      "encoder_q-layer.5": 1402.4735,
      "encoder_q-layer.6": 1482.486,
      "encoder_q-layer.7": 1620.1455,
      "encoder_q-layer.8": 1592.7828,
      "encoder_q-layer.9": 1269.2473,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1683,
      "inbatch_pos_score": 0.8218,
      "learning_rate": 3.4888888888888895e-05,
      "loss": 3.3496,
      "norm_diff": 0.0477,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2322.5447,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1683,
      "query_norm": 1.3531,
      "queue_k_norm": 1.3979,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9114,
      "sent_len_1": 66.679,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1163,
      "stdk": 0.0485,
      "stdq": 0.0452,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37200
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.4096,
      "doc_norm": 1.4041,
      "encoder_q-embeddings": 1643.0905,
      "encoder_q-layer.0": 1040.9153,
      "encoder_q-layer.1": 1084.1627,
      "encoder_q-layer.10": 1164.4596,
      "encoder_q-layer.11": 2599.5857,
      "encoder_q-layer.2": 1188.5066,
      "encoder_q-layer.3": 1247.7837,
      "encoder_q-layer.4": 1285.0558,
      "encoder_q-layer.5": 1284.6721,
      "encoder_q-layer.6": 1352.7365,
      "encoder_q-layer.7": 1574.8917,
      "encoder_q-layer.8": 1691.1647,
      "encoder_q-layer.9": 1351.5009,
      "epoch": 0.36,
      "inbatch_neg_score": 0.1677,
      "inbatch_pos_score": 0.7964,
      "learning_rate": 3.483333333333334e-05,
      "loss": 3.4096,
      "norm_diff": 0.0661,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2235.4097,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1681,
      "query_norm": 1.338,
      "queue_k_norm": 1.3961,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0961,
      "sent_len_1": 66.9105,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3025,
      "stdk": 0.0486,
      "stdq": 0.0447,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 37300
    },
    {
      "accuracy": 51.5625,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3694,
      "doc_norm": 1.3934,
      "encoder_q-embeddings": 1356.2687,
      "encoder_q-layer.0": 884.5145,
      "encoder_q-layer.1": 990.5148,
      "encoder_q-layer.10": 1270.7151,
      "encoder_q-layer.11": 2691.9866,
      "encoder_q-layer.2": 1142.3348,
      "encoder_q-layer.3": 1143.1927,
      "encoder_q-layer.4": 1268.4967,
      "encoder_q-layer.5": 1397.0415,
      "encoder_q-layer.6": 1505.5405,
      "encoder_q-layer.7": 1592.048,
      "encoder_q-layer.8": 1840.1895,
      "encoder_q-layer.9": 1443.9324,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1662,
      "inbatch_pos_score": 0.8071,
      "learning_rate": 3.477777777777778e-05,
      "loss": 3.3694,
      "norm_diff": 0.0401,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2222.0372,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1652,
      "query_norm": 1.3534,
      "queue_k_norm": 1.3976,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9361,
      "sent_len_1": 66.7934,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3375,
      "stdk": 0.0482,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37400
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3672,
      "doc_norm": 1.4034,
      "encoder_q-embeddings": 11509.5742,
      "encoder_q-layer.0": 8447.6963,
      "encoder_q-layer.1": 8263.6562,
      "encoder_q-layer.10": 2032.012,
      "encoder_q-layer.11": 2571.0786,
      "encoder_q-layer.2": 9239.083,
      "encoder_q-layer.3": 10938.3672,
      "encoder_q-layer.4": 12707.1621,
      "encoder_q-layer.5": 14960.25,
      "encoder_q-layer.6": 20196.3125,
      "encoder_q-layer.7": 29935.0645,
      "encoder_q-layer.8": Infinity,
      "encoder_q-layer.9": 17496.3809,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1696,
      "inbatch_pos_score": 0.835,
      "learning_rate": 3.472222222222222e-05,
      "loss": 3.3672,
      "norm_diff": 0.0171,
      "norm_loss": 0.0,
      "postclip_grad_norm": NaN,
      "preclip_grad_norm": Infinity,
      "preclip_grad_norm_avg": Infinity,
      "q@queue_neg_score": 0.1697,
      "query_norm": 1.3932,
      "queue_k_norm": 1.3976,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0232,
      "sent_len_1": 66.8528,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7525,
      "stdk": 0.0486,
      "stdq": 0.0467,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 37500
    },
    {
      "accuracy": 57.9102,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3577,
      "doc_norm": 1.4035,
      "encoder_q-embeddings": 669.0736,
      "encoder_q-layer.0": 447.5952,
      "encoder_q-layer.1": 487.2352,
      "encoder_q-layer.10": 550.2194,
      "encoder_q-layer.11": 1221.766,
      "encoder_q-layer.2": 563.0723,
      "encoder_q-layer.3": 599.6226,
      "encoder_q-layer.4": 644.653,
      "encoder_q-layer.5": 631.6823,
      "encoder_q-layer.6": 649.6618,
      "encoder_q-layer.7": 719.478,
      "encoder_q-layer.8": 773.7578,
      "encoder_q-layer.9": 596.5146,
      "epoch": 0.37,
      "inbatch_neg_score": 0.176,
      "inbatch_pos_score": 0.8794,
      "learning_rate": 3.466666666666667e-05,
      "loss": 3.3577,
      "norm_diff": 0.0248,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1022.669,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1758,
      "query_norm": 1.3787,
      "queue_k_norm": 1.3983,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8495,
      "sent_len_1": 66.7225,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.845,
      "stdk": 0.0486,
      "stdq": 0.0461,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37600
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3558,
      "doc_norm": 1.3934,
      "encoder_q-embeddings": 1389.0618,
      "encoder_q-layer.0": 963.3607,
      "encoder_q-layer.1": 1132.3533,
      "encoder_q-layer.10": 653.3033,
      "encoder_q-layer.11": 1422.6047,
      "encoder_q-layer.2": 1382.0798,
      "encoder_q-layer.3": 1517.9768,
      "encoder_q-layer.4": 1650.5132,
      "encoder_q-layer.5": 1772.2921,
      "encoder_q-layer.6": 1621.7524,
      "encoder_q-layer.7": 1624.2843,
      "encoder_q-layer.8": 1057.1666,
      "encoder_q-layer.9": 750.3931,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1743,
      "inbatch_pos_score": 0.8203,
      "learning_rate": 3.4611111111111114e-05,
      "loss": 3.3558,
      "norm_diff": 0.0259,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2022.0634,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1746,
      "query_norm": 1.3689,
      "queue_k_norm": 1.3992,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9609,
      "sent_len_1": 66.6854,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.8738,
      "stdk": 0.0481,
      "stdq": 0.0459,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37700
    },
    {
      "accuracy": 47.0703,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3727,
      "doc_norm": 1.3941,
      "encoder_q-embeddings": 655.854,
      "encoder_q-layer.0": 415.7728,
      "encoder_q-layer.1": 441.4001,
      "encoder_q-layer.10": 609.4597,
      "encoder_q-layer.11": 1376.1709,
      "encoder_q-layer.2": 494.6773,
      "encoder_q-layer.3": 547.4537,
      "encoder_q-layer.4": 573.0698,
      "encoder_q-layer.5": 615.4177,
      "encoder_q-layer.6": 704.8508,
      "encoder_q-layer.7": 783.7787,
      "encoder_q-layer.8": 810.6597,
      "encoder_q-layer.9": 672.3845,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1785,
      "inbatch_pos_score": 0.772,
      "learning_rate": 3.4555555555555556e-05,
      "loss": 3.3727,
      "norm_diff": 0.0556,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1071.2842,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.177,
      "query_norm": 1.3386,
      "queue_k_norm": 1.3998,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7958,
      "sent_len_1": 66.8048,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6838,
      "stdk": 0.0481,
      "stdq": 0.0446,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 37800
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.343,
      "doc_norm": 1.4059,
      "encoder_q-embeddings": 1152.7859,
      "encoder_q-layer.0": 796.3542,
      "encoder_q-layer.1": 852.8484,
      "encoder_q-layer.10": 606.1483,
      "encoder_q-layer.11": 1303.1737,
      "encoder_q-layer.2": 1012.7851,
      "encoder_q-layer.3": 1028.9342,
      "encoder_q-layer.4": 1012.9362,
      "encoder_q-layer.5": 957.5239,
      "encoder_q-layer.6": 922.2787,
      "encoder_q-layer.7": 1007.7191,
      "encoder_q-layer.8": 990.0455,
      "encoder_q-layer.9": 676.4854,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1735,
      "inbatch_pos_score": 0.8179,
      "learning_rate": 3.45e-05,
      "loss": 3.343,
      "norm_diff": 0.0565,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1462.337,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1742,
      "query_norm": 1.3495,
      "queue_k_norm": 1.3995,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9448,
      "sent_len_1": 66.834,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1025,
      "stdk": 0.0486,
      "stdq": 0.045,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 37900
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3704,
      "doc_norm": 1.4043,
      "encoder_q-embeddings": 627.4984,
      "encoder_q-layer.0": 426.1138,
      "encoder_q-layer.1": 442.7548,
      "encoder_q-layer.10": 551.704,
      "encoder_q-layer.11": 1214.9602,
      "encoder_q-layer.2": 505.2847,
      "encoder_q-layer.3": 512.6962,
      "encoder_q-layer.4": 545.5739,
      "encoder_q-layer.5": 551.1111,
      "encoder_q-layer.6": 602.2524,
      "encoder_q-layer.7": 702.2904,
      "encoder_q-layer.8": 710.4362,
      "encoder_q-layer.9": 575.722,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1731,
      "inbatch_pos_score": 0.8525,
      "learning_rate": 3.444444444444445e-05,
      "loss": 3.3704,
      "norm_diff": 0.0437,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 965.783,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1729,
      "query_norm": 1.3606,
      "queue_k_norm": 1.4014,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9389,
      "sent_len_1": 66.9929,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.175,
      "stdk": 0.0485,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38000
    },
    {
      "accuracy": 51.7578,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3599,
      "doc_norm": 1.4029,
      "encoder_q-embeddings": 567.3127,
      "encoder_q-layer.0": 370.2695,
      "encoder_q-layer.1": 396.4028,
      "encoder_q-layer.10": 645.5457,
      "encoder_q-layer.11": 1353.7666,
      "encoder_q-layer.2": 444.5764,
      "encoder_q-layer.3": 480.0852,
      "encoder_q-layer.4": 518.1058,
      "encoder_q-layer.5": 566.8663,
      "encoder_q-layer.6": 625.0043,
      "encoder_q-layer.7": 731.7353,
      "encoder_q-layer.8": 823.8107,
      "encoder_q-layer.9": 691.2753,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1783,
      "inbatch_pos_score": 0.8228,
      "learning_rate": 3.438888888888889e-05,
      "loss": 3.3599,
      "norm_diff": 0.0379,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1008.1262,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1788,
      "query_norm": 1.3649,
      "queue_k_norm": 1.4008,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0181,
      "sent_len_1": 66.8521,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.9013,
      "stdk": 0.0485,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38100
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3264,
      "doc_norm": 1.4008,
      "encoder_q-embeddings": 578.1523,
      "encoder_q-layer.0": 356.1647,
      "encoder_q-layer.1": 371.0636,
      "encoder_q-layer.10": 604.4282,
      "encoder_q-layer.11": 1393.3163,
      "encoder_q-layer.2": 407.1126,
      "encoder_q-layer.3": 447.41,
      "encoder_q-layer.4": 473.8751,
      "encoder_q-layer.5": 482.5728,
      "encoder_q-layer.6": 572.4805,
      "encoder_q-layer.7": 657.5682,
      "encoder_q-layer.8": 767.1038,
      "encoder_q-layer.9": 636.5831,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1768,
      "inbatch_pos_score": 0.8066,
      "learning_rate": 3.433333333333333e-05,
      "loss": 3.3264,
      "norm_diff": 0.0624,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 963.566,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1777,
      "query_norm": 1.3384,
      "queue_k_norm": 1.4001,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0683,
      "sent_len_1": 66.9894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7562,
      "stdk": 0.0484,
      "stdq": 0.0445,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 38200
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3809,
      "doc_norm": 1.4059,
      "encoder_q-embeddings": 3069.3687,
      "encoder_q-layer.0": 2096.8442,
      "encoder_q-layer.1": 2468.627,
      "encoder_q-layer.10": 561.3721,
      "encoder_q-layer.11": 1172.9507,
      "encoder_q-layer.2": 2876.323,
      "encoder_q-layer.3": 2932.1504,
      "encoder_q-layer.4": 3052.8401,
      "encoder_q-layer.5": 3077.189,
      "encoder_q-layer.6": 3328.6929,
      "encoder_q-layer.7": 3320.8752,
      "encoder_q-layer.8": 2786.1758,
      "encoder_q-layer.9": 996.637,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1787,
      "inbatch_pos_score": 0.8477,
      "learning_rate": 3.427777777777778e-05,
      "loss": 3.3809,
      "norm_diff": 0.0392,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3989.5027,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.177,
      "query_norm": 1.3666,
      "queue_k_norm": 1.4011,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1834,
      "sent_len_1": 66.6941,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9075,
      "stdk": 0.0485,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38300
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3489,
      "doc_norm": 1.3992,
      "encoder_q-embeddings": 711.3175,
      "encoder_q-layer.0": 497.7701,
      "encoder_q-layer.1": 566.0946,
      "encoder_q-layer.10": 577.7365,
      "encoder_q-layer.11": 1219.6321,
      "encoder_q-layer.2": 686.2197,
      "encoder_q-layer.3": 747.7931,
      "encoder_q-layer.4": 722.5842,
      "encoder_q-layer.5": 682.689,
      "encoder_q-layer.6": 728.0964,
      "encoder_q-layer.7": 798.6857,
      "encoder_q-layer.8": 770.2348,
      "encoder_q-layer.9": 608.45,
      "epoch": 0.37,
      "inbatch_neg_score": 0.1824,
      "inbatch_pos_score": 0.8682,
      "learning_rate": 3.4222222222222224e-05,
      "loss": 3.3489,
      "norm_diff": 0.011,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1091.3866,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1824,
      "query_norm": 1.393,
      "queue_k_norm": 1.4012,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8102,
      "sent_len_1": 66.6387,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1712,
      "stdk": 0.0483,
      "stdq": 0.0462,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38400
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3354,
      "doc_norm": 1.4054,
      "encoder_q-embeddings": 1226.1396,
      "encoder_q-layer.0": 947.7741,
      "encoder_q-layer.1": 1121.4462,
      "encoder_q-layer.10": 591.4812,
      "encoder_q-layer.11": 1244.5713,
      "encoder_q-layer.2": 1264.2577,
      "encoder_q-layer.3": 1281.4403,
      "encoder_q-layer.4": 1502.3201,
      "encoder_q-layer.5": 1513.1021,
      "encoder_q-layer.6": 1736.5763,
      "encoder_q-layer.7": 1477.4039,
      "encoder_q-layer.8": 1559.1649,
      "encoder_q-layer.9": 1011.7104,
      "epoch": 0.38,
      "inbatch_neg_score": 0.1901,
      "inbatch_pos_score": 0.8379,
      "learning_rate": 3.4166666666666666e-05,
      "loss": 3.3354,
      "norm_diff": 0.0229,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1957.3813,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1895,
      "query_norm": 1.3825,
      "queue_k_norm": 1.4044,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0108,
      "sent_len_1": 66.6671,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.12,
      "stdk": 0.0485,
      "stdq": 0.0457,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38500
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3391,
      "doc_norm": 1.4026,
      "encoder_q-embeddings": 2252.5225,
      "encoder_q-layer.0": 1587.0424,
      "encoder_q-layer.1": 1882.9913,
      "encoder_q-layer.10": 632.8958,
      "encoder_q-layer.11": 1290.4493,
      "encoder_q-layer.2": 2389.4355,
      "encoder_q-layer.3": 2915.0642,
      "encoder_q-layer.4": 3456.1443,
      "encoder_q-layer.5": 4018.9807,
      "encoder_q-layer.6": 2517.394,
      "encoder_q-layer.7": 1783.557,
      "encoder_q-layer.8": 1543.5857,
      "encoder_q-layer.9": 925.7364,
      "epoch": 0.38,
      "inbatch_neg_score": 0.1963,
      "inbatch_pos_score": 0.875,
      "learning_rate": 3.411111111111111e-05,
      "loss": 3.3391,
      "norm_diff": 0.0192,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3347.3626,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1957,
      "query_norm": 1.4149,
      "queue_k_norm": 1.4026,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0125,
      "sent_len_1": 66.6625,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7388,
      "stdk": 0.0484,
      "stdq": 0.0465,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38600
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3618,
      "doc_norm": 1.4031,
      "encoder_q-embeddings": 1291.7607,
      "encoder_q-layer.0": 868.8148,
      "encoder_q-layer.1": 1099.7252,
      "encoder_q-layer.10": 644.2998,
      "encoder_q-layer.11": 1315.7524,
      "encoder_q-layer.2": 1228.726,
      "encoder_q-layer.3": 1275.1582,
      "encoder_q-layer.4": 1254.2858,
      "encoder_q-layer.5": 1196.467,
      "encoder_q-layer.6": 1355.7427,
      "encoder_q-layer.7": 1523.1218,
      "encoder_q-layer.8": 2188.8145,
      "encoder_q-layer.9": 1357.9227,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2015,
      "inbatch_pos_score": 0.8384,
      "learning_rate": 3.405555555555556e-05,
      "loss": 3.3618,
      "norm_diff": 0.0163,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1999.4109,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2007,
      "query_norm": 1.3872,
      "queue_k_norm": 1.4017,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.931,
      "sent_len_1": 66.7699,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6525,
      "stdk": 0.0483,
      "stdq": 0.0455,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 38700
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3687,
      "doc_norm": 1.3955,
      "encoder_q-embeddings": 603.1904,
      "encoder_q-layer.0": 386.6649,
      "encoder_q-layer.1": 414.2968,
      "encoder_q-layer.10": 599.7798,
      "encoder_q-layer.11": 1284.0865,
      "encoder_q-layer.2": 469.7938,
      "encoder_q-layer.3": 498.8611,
      "encoder_q-layer.4": 538.4988,
      "encoder_q-layer.5": 589.4727,
      "encoder_q-layer.6": 604.4183,
      "encoder_q-layer.7": 714.5781,
      "encoder_q-layer.8": 779.7006,
      "encoder_q-layer.9": 668.3961,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2024,
      "inbatch_pos_score": 0.8682,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 3.3687,
      "norm_diff": 0.0107,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 986.555,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2017,
      "query_norm": 1.3995,
      "queue_k_norm": 1.404,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0373,
      "sent_len_1": 66.8136,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6513,
      "stdk": 0.048,
      "stdq": 0.0458,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38800
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.37,
      "doc_norm": 1.4062,
      "encoder_q-embeddings": 571.4634,
      "encoder_q-layer.0": 374.6999,
      "encoder_q-layer.1": 384.1917,
      "encoder_q-layer.10": 582.7562,
      "encoder_q-layer.11": 1312.4373,
      "encoder_q-layer.2": 425.2076,
      "encoder_q-layer.3": 453.1045,
      "encoder_q-layer.4": 473.7238,
      "encoder_q-layer.5": 512.3858,
      "encoder_q-layer.6": 548.1695,
      "encoder_q-layer.7": 616.3169,
      "encoder_q-layer.8": 689.299,
      "encoder_q-layer.9": 587.3215,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2108,
      "inbatch_pos_score": 0.8647,
      "learning_rate": 3.394444444444444e-05,
      "loss": 3.37,
      "norm_diff": 0.0337,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 945.1627,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2111,
      "query_norm": 1.3725,
      "queue_k_norm": 1.405,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8132,
      "sent_len_1": 66.8597,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4,
      "stdk": 0.0484,
      "stdq": 0.0446,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 38900
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.354,
      "doc_norm": 1.4063,
      "encoder_q-embeddings": 790.3129,
      "encoder_q-layer.0": 527.9492,
      "encoder_q-layer.1": 594.5428,
      "encoder_q-layer.10": 606.434,
      "encoder_q-layer.11": 1264.8159,
      "encoder_q-layer.2": 740.6779,
      "encoder_q-layer.3": 797.2274,
      "encoder_q-layer.4": 830.6208,
      "encoder_q-layer.5": 859.8293,
      "encoder_q-layer.6": 891.1483,
      "encoder_q-layer.7": 1025.2632,
      "encoder_q-layer.8": 1086.4337,
      "encoder_q-layer.9": 779.0574,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2159,
      "inbatch_pos_score": 0.9023,
      "learning_rate": 3.388888888888889e-05,
      "loss": 3.354,
      "norm_diff": 0.0137,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1270.3952,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2153,
      "query_norm": 1.4005,
      "queue_k_norm": 1.4056,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8772,
      "sent_len_1": 66.6243,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0938,
      "stdk": 0.0484,
      "stdq": 0.0459,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 39000
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3483,
      "doc_norm": 1.4094,
      "encoder_q-embeddings": 706.5385,
      "encoder_q-layer.0": 479.006,
      "encoder_q-layer.1": 488.6994,
      "encoder_q-layer.10": 581.1353,
      "encoder_q-layer.11": 1328.0575,
      "encoder_q-layer.2": 574.3455,
      "encoder_q-layer.3": 644.5504,
      "encoder_q-layer.4": 669.93,
      "encoder_q-layer.5": 709.689,
      "encoder_q-layer.6": 783.1199,
      "encoder_q-layer.7": 768.435,
      "encoder_q-layer.8": 721.5502,
      "encoder_q-layer.9": 610.3312,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2213,
      "inbatch_pos_score": 0.8799,
      "learning_rate": 3.3833333333333334e-05,
      "loss": 3.3483,
      "norm_diff": 0.0079,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1095.1661,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2197,
      "query_norm": 1.4015,
      "queue_k_norm": 1.4079,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8289,
      "sent_len_1": 66.6242,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7413,
      "stdk": 0.0484,
      "stdq": 0.046,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39100
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3474,
      "doc_norm": 1.4031,
      "encoder_q-embeddings": 2467.8459,
      "encoder_q-layer.0": 1707.8228,
      "encoder_q-layer.1": 1668.4395,
      "encoder_q-layer.10": 653.9602,
      "encoder_q-layer.11": 1398.0372,
      "encoder_q-layer.2": 2076.1956,
      "encoder_q-layer.3": 2235.8005,
      "encoder_q-layer.4": 2250.3477,
      "encoder_q-layer.5": 2207.2896,
      "encoder_q-layer.6": 1859.1075,
      "encoder_q-layer.7": 1970.1234,
      "encoder_q-layer.8": 1474.5088,
      "encoder_q-layer.9": 757.8829,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2228,
      "inbatch_pos_score": 0.8579,
      "learning_rate": 3.377777777777778e-05,
      "loss": 3.3474,
      "norm_diff": 0.0221,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2823.6359,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2227,
      "query_norm": 1.381,
      "queue_k_norm": 1.4082,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2848,
      "sent_len_1": 66.8116,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7962,
      "stdk": 0.0481,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39200
    },
    {
      "accuracy": 50.6836,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3685,
      "doc_norm": 1.4094,
      "encoder_q-embeddings": 1938.427,
      "encoder_q-layer.0": 1378.1002,
      "encoder_q-layer.1": 1603.515,
      "encoder_q-layer.10": 622.4032,
      "encoder_q-layer.11": 1359.4736,
      "encoder_q-layer.2": 1923.1278,
      "encoder_q-layer.3": 2019.0319,
      "encoder_q-layer.4": 1965.865,
      "encoder_q-layer.5": 1972.4304,
      "encoder_q-layer.6": 1912.1078,
      "encoder_q-layer.7": 2039.6169,
      "encoder_q-layer.8": 1310.3932,
      "encoder_q-layer.9": 717.5338,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2195,
      "inbatch_pos_score": 0.8501,
      "learning_rate": 3.3722222222222225e-05,
      "loss": 3.3685,
      "norm_diff": 0.0463,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2524.7761,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2191,
      "query_norm": 1.3631,
      "queue_k_norm": 1.4092,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9994,
      "sent_len_1": 66.4936,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.4525,
      "stdk": 0.0483,
      "stdq": 0.0448,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 39300
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3299,
      "doc_norm": 1.4127,
      "encoder_q-embeddings": 757.2884,
      "encoder_q-layer.0": 496.2839,
      "encoder_q-layer.1": 511.6798,
      "encoder_q-layer.10": 584.9152,
      "encoder_q-layer.11": 1340.7513,
      "encoder_q-layer.2": 589.9082,
      "encoder_q-layer.3": 628.8683,
      "encoder_q-layer.4": 664.424,
      "encoder_q-layer.5": 702.7582,
      "encoder_q-layer.6": 714.1782,
      "encoder_q-layer.7": 697.1252,
      "encoder_q-layer.8": 732.8589,
      "encoder_q-layer.9": 607.4174,
      "epoch": 0.38,
      "inbatch_neg_score": 0.2175,
      "inbatch_pos_score": 0.874,
      "learning_rate": 3.366666666666667e-05,
      "loss": 3.3299,
      "norm_diff": 0.0312,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1084.9338,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2173,
      "query_norm": 1.3815,
      "queue_k_norm": 1.4108,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9318,
      "sent_len_1": 66.6555,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9663,
      "stdk": 0.0484,
      "stdq": 0.0456,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39400
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3419,
      "doc_norm": 1.4129,
      "encoder_q-embeddings": 576.2877,
      "encoder_q-layer.0": 354.9731,
      "encoder_q-layer.1": 364.5914,
      "encoder_q-layer.10": 620.0936,
      "encoder_q-layer.11": 1336.3499,
      "encoder_q-layer.2": 401.4544,
      "encoder_q-layer.3": 420.8601,
      "encoder_q-layer.4": 456.3818,
      "encoder_q-layer.5": 471.2786,
      "encoder_q-layer.6": 524.0115,
      "encoder_q-layer.7": 625.4122,
      "encoder_q-layer.8": 717.5291,
      "encoder_q-layer.9": 633.0168,
      "epoch": 0.39,
      "inbatch_neg_score": 0.2161,
      "inbatch_pos_score": 0.8643,
      "learning_rate": 3.3611111111111116e-05,
      "loss": 3.3419,
      "norm_diff": 0.0548,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 932.3785,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2151,
      "query_norm": 1.3581,
      "queue_k_norm": 1.4093,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8663,
      "sent_len_1": 66.6706,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7637,
      "stdk": 0.0484,
      "stdq": 0.0447,
      "stdqueue_k": 0.0484,
      "stdqueue_q": 0.0,
      "step": 39500
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3499,
      "doc_norm": 1.4145,
      "encoder_q-embeddings": 1200.2175,
      "encoder_q-layer.0": 789.0743,
      "encoder_q-layer.1": 873.2488,
      "encoder_q-layer.10": 1362.0583,
      "encoder_q-layer.11": 2638.896,
      "encoder_q-layer.2": 970.1713,
      "encoder_q-layer.3": 1034.9135,
      "encoder_q-layer.4": 1087.2527,
      "encoder_q-layer.5": 1120.6365,
      "encoder_q-layer.6": 1201.6266,
      "encoder_q-layer.7": 1299.5336,
      "encoder_q-layer.8": 1453.8374,
      "encoder_q-layer.9": 1245.3826,
      "epoch": 0.39,
      "inbatch_neg_score": 0.2117,
      "inbatch_pos_score": 0.8779,
      "learning_rate": 3.355555555555556e-05,
      "loss": 3.3499,
      "norm_diff": 0.0205,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1983.5373,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2122,
      "query_norm": 1.394,
      "queue_k_norm": 1.4131,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.095,
      "sent_len_1": 66.7927,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6763,
      "stdk": 0.0485,
      "stdq": 0.0462,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39600
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3832,
      "doc_norm": 1.4123,
      "encoder_q-embeddings": 1148.5396,
      "encoder_q-layer.0": 727.2261,
      "encoder_q-layer.1": 756.4799,
      "encoder_q-layer.10": 1129.6239,
      "encoder_q-layer.11": 2590.2024,
      "encoder_q-layer.2": 856.0329,
      "encoder_q-layer.3": 883.1859,
      "encoder_q-layer.4": 930.1815,
      "encoder_q-layer.5": 934.0659,
      "encoder_q-layer.6": 1024.8186,
      "encoder_q-layer.7": 1193.8624,
      "encoder_q-layer.8": 1351.6421,
      "encoder_q-layer.9": 1183.5057,
      "epoch": 0.39,
      "inbatch_neg_score": 0.2088,
      "inbatch_pos_score": 0.8931,
      "learning_rate": 3.35e-05,
      "loss": 3.3832,
      "norm_diff": 0.0183,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1830.8072,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2094,
      "query_norm": 1.394,
      "queue_k_norm": 1.4125,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.839,
      "sent_len_1": 66.8073,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9487,
      "stdk": 0.0484,
      "stdq": 0.0463,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39700
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.378,
      "doc_norm": 1.4117,
      "encoder_q-embeddings": 1549.6987,
      "encoder_q-layer.0": 1000.3274,
      "encoder_q-layer.1": 1121.3597,
      "encoder_q-layer.10": 1315.4952,
      "encoder_q-layer.11": 2628.6235,
      "encoder_q-layer.2": 1297.7834,
      "encoder_q-layer.3": 1416.7506,
      "encoder_q-layer.4": 1580.7021,
      "encoder_q-layer.5": 1724.7223,
      "encoder_q-layer.6": 1839.4146,
      "encoder_q-layer.7": 1893.8639,
      "encoder_q-layer.8": 1692.6525,
      "encoder_q-layer.9": 1259.5851,
      "epoch": 0.39,
      "inbatch_neg_score": 0.204,
      "inbatch_pos_score": 0.8789,
      "learning_rate": 3.3444444444444443e-05,
      "loss": 3.378,
      "norm_diff": 0.0243,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2417.2484,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2054,
      "query_norm": 1.3873,
      "queue_k_norm": 1.4135,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0303,
      "sent_len_1": 66.6961,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1025,
      "stdk": 0.0484,
      "stdq": 0.0461,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39800
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3403,
      "doc_norm": 1.4116,
      "encoder_q-embeddings": 2308.1023,
      "encoder_q-layer.0": 1552.54,
      "encoder_q-layer.1": 1704.9979,
      "encoder_q-layer.10": 1322.7167,
      "encoder_q-layer.11": 2664.7085,
      "encoder_q-layer.2": 1801.5869,
      "encoder_q-layer.3": 1954.4158,
      "encoder_q-layer.4": 2145.2529,
      "encoder_q-layer.5": 2204.6934,
      "encoder_q-layer.6": 2251.2336,
      "encoder_q-layer.7": 1993.2275,
      "encoder_q-layer.8": 1663.2617,
      "encoder_q-layer.9": 1340.5607,
      "epoch": 0.39,
      "inbatch_neg_score": 0.2035,
      "inbatch_pos_score": 0.8428,
      "learning_rate": 3.338888888888889e-05,
      "loss": 3.3403,
      "norm_diff": 0.0643,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2932.819,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2032,
      "query_norm": 1.3473,
      "queue_k_norm": 1.413,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1949,
      "sent_len_1": 66.7856,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7688,
      "stdk": 0.0483,
      "stdq": 0.0447,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 39900
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3639,
      "doc_norm": 1.4198,
      "encoder_q-embeddings": 1712.2888,
      "encoder_q-layer.0": 1112.1698,
      "encoder_q-layer.1": 1183.1207,
      "encoder_q-layer.10": 1124.5503,
      "encoder_q-layer.11": 2543.2493,
      "encoder_q-layer.2": 1286.7252,
      "encoder_q-layer.3": 1405.7896,
      "encoder_q-layer.4": 1580.1936,
      "encoder_q-layer.5": 1619.6897,
      "encoder_q-layer.6": 1450.9484,
      "encoder_q-layer.7": 1641.8936,
      "encoder_q-layer.8": 1520.4315,
      "encoder_q-layer.9": 1170.9102,
      "epoch": 0.39,
      "inbatch_neg_score": 0.1958,
      "inbatch_pos_score": 0.8613,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 3.3639,
      "norm_diff": 0.0622,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2317.4195,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1974,
      "query_norm": 1.3576,
      "queue_k_norm": 1.413,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1821,
      "sent_len_1": 66.6918,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0213,
      "stdk": 0.0487,
      "stdq": 0.0452,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40000
    },
    {
      "dev_runtime": 37.3792,
      "dev_samples_per_second": 1.712,
      "dev_steps_per_second": 0.027,
      "epoch": 0.39,
      "step": 40000,
      "test_accuracy": 93.9697265625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3441445231437683,
      "test_doc_norm": 1.415613055229187,
      "test_inbatch_neg_score": 0.5627651214599609,
      "test_inbatch_pos_score": 1.512505292892456,
      "test_loss": 0.3441445231437683,
      "test_loss_align": 1.0954631567001343,
      "test_loss_unif": 3.9128928184509277,
      "test_loss_unif_q@queue": 3.9128928184509277,
      "test_norm_diff": 0.02933865785598755,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.18375235795974731,
      "test_query_norm": 1.4449516534805298,
      "test_queue_k_norm": 1.4133580923080444,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.0429464727640152,
      "test_stdq": 0.042964886873960495,
      "test_stdqueue_k": 0.048496000468730927,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 37.3792,
      "dev_samples_per_second": 1.712,
      "dev_steps_per_second": 0.027,
      "epoch": 0.39,
      "eval_beir-arguana_ndcg@10": 0.37345,
      "eval_beir-arguana_recall@10": 0.64011,
      "eval_beir-arguana_recall@100": 0.92817,
      "eval_beir-arguana_recall@20": 0.76956,
      "eval_beir-avg_ndcg@10": 0.3658995,
      "eval_beir-avg_recall@10": 0.43282233333333336,
      "eval_beir-avg_recall@100": 0.6096421666666667,
      "eval_beir-avg_recall@20": 0.4901745833333333,
      "eval_beir-cqadupstack_ndcg@10": 0.25832499999999997,
      "eval_beir-cqadupstack_recall@10": 0.34714333333333336,
      "eval_beir-cqadupstack_recall@100": 0.5682416666666666,
      "eval_beir-cqadupstack_recall@20": 0.41124583333333337,
      "eval_beir-fiqa_ndcg@10": 0.2009,
      "eval_beir-fiqa_recall@10": 0.25622,
      "eval_beir-fiqa_recall@100": 0.50335,
      "eval_beir-fiqa_recall@20": 0.31686,
      "eval_beir-nfcorpus_ndcg@10": 0.27813,
      "eval_beir-nfcorpus_recall@10": 0.13619,
      "eval_beir-nfcorpus_recall@100": 0.25987,
      "eval_beir-nfcorpus_recall@20": 0.16341,
      "eval_beir-nq_ndcg@10": 0.26477,
      "eval_beir-nq_recall@10": 0.4383,
      "eval_beir-nq_recall@100": 0.77622,
      "eval_beir-nq_recall@20": 0.55808,
      "eval_beir-quora_ndcg@10": 0.81506,
      "eval_beir-quora_recall@10": 0.91213,
      "eval_beir-quora_recall@100": 0.98424,
      "eval_beir-quora_recall@20": 0.94741,
      "eval_beir-scidocs_ndcg@10": 0.13649,
      "eval_beir-scidocs_recall@10": 0.14498,
      "eval_beir-scidocs_recall@100": 0.3353,
      "eval_beir-scidocs_recall@20": 0.19777,
      "eval_beir-scifact_ndcg@10": 0.59783,
      "eval_beir-scifact_recall@10": 0.75017,
      "eval_beir-scifact_recall@100": 0.91311,
      "eval_beir-scifact_recall@20": 0.8125,
      "eval_beir-trec-covid_ndcg@10": 0.5112,
      "eval_beir-trec-covid_recall@10": 0.544,
      "eval_beir-trec-covid_recall@100": 0.3836,
      "eval_beir-trec-covid_recall@20": 0.504,
      "eval_beir-webis-touche2020_ndcg@10": 0.22284,
      "eval_beir-webis-touche2020_recall@10": 0.15898,
      "eval_beir-webis-touche2020_recall@100": 0.44432,
      "eval_beir-webis-touche2020_recall@20": 0.22091,
      "eval_senteval-avg_sts": 0.7385578406967059,
      "eval_senteval-sickr_spearman": 0.7056836532643718,
      "eval_senteval-stsb_spearman": 0.77143202812904,
      "step": 40000,
      "test_accuracy": 93.9697265625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3441445231437683,
      "test_doc_norm": 1.415613055229187,
      "test_inbatch_neg_score": 0.5627651214599609,
      "test_inbatch_pos_score": 1.512505292892456,
      "test_loss": 0.3441445231437683,
      "test_loss_align": 1.0954631567001343,
      "test_loss_unif": 3.9128928184509277,
      "test_loss_unif_q@queue": 3.9128928184509277,
      "test_norm_diff": 0.02933865785598755,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.18375235795974731,
      "test_query_norm": 1.4449516534805298,
      "test_queue_k_norm": 1.4133580923080444,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.0429464727640152,
      "test_stdq": 0.042964886873960495,
      "test_stdqueue_k": 0.048496000468730927,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.333,
      "doc_norm": 1.417,
      "encoder_q-embeddings": 3233.7769,
      "encoder_q-layer.0": 2166.4883,
      "encoder_q-layer.1": 2200.9351,
      "encoder_q-layer.10": 1134.8048,
      "encoder_q-layer.11": 2597.5698,
      "encoder_q-layer.2": 2550.4905,
      "encoder_q-layer.3": 2852.7803,
      "encoder_q-layer.4": 3008.3757,
      "encoder_q-layer.5": 2995.5044,
      "encoder_q-layer.6": 2683.9834,
      "encoder_q-layer.7": 2649.6411,
      "encoder_q-layer.8": 2316.0928,
      "encoder_q-layer.9": 1440.4847,
      "epoch": 0.39,
      "inbatch_neg_score": 0.1923,
      "inbatch_pos_score": 0.8301,
      "learning_rate": 3.327777777777778e-05,
      "loss": 3.333,
      "norm_diff": 0.0711,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3843.158,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1914,
      "query_norm": 1.3459,
      "queue_k_norm": 1.414,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.978,
      "sent_len_1": 66.9881,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2612,
      "stdk": 0.0485,
      "stdq": 0.0448,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40100
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3555,
      "doc_norm": 1.4154,
      "encoder_q-embeddings": 1312.0577,
      "encoder_q-layer.0": 792.7803,
      "encoder_q-layer.1": 841.4208,
      "encoder_q-layer.10": 1285.903,
      "encoder_q-layer.11": 2595.6926,
      "encoder_q-layer.2": 972.9858,
      "encoder_q-layer.3": 1014.6182,
      "encoder_q-layer.4": 1085.9027,
      "encoder_q-layer.5": 1185.4161,
      "encoder_q-layer.6": 1236.2623,
      "encoder_q-layer.7": 1489.2904,
      "encoder_q-layer.8": 1492.4631,
      "encoder_q-layer.9": 1355.6554,
      "epoch": 0.39,
      "inbatch_neg_score": 0.1856,
      "inbatch_pos_score": 0.8711,
      "learning_rate": 3.322222222222222e-05,
      "loss": 3.3555,
      "norm_diff": 0.0231,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1995.4671,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1851,
      "query_norm": 1.3923,
      "queue_k_norm": 1.413,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.949,
      "sent_len_1": 66.8894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6525,
      "stdk": 0.0485,
      "stdq": 0.0467,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40200
    },
    {
      "accuracy": 51.5625,
      "active_queue_size": 16384.0,
      "cl_loss": 3.353,
      "doc_norm": 1.4148,
      "encoder_q-embeddings": 1295.3986,
      "encoder_q-layer.0": 893.2507,
      "encoder_q-layer.1": 926.0757,
      "encoder_q-layer.10": 1212.9961,
      "encoder_q-layer.11": 2667.6152,
      "encoder_q-layer.2": 1073.6384,
      "encoder_q-layer.3": 1106.9098,
      "encoder_q-layer.4": 1318.5913,
      "encoder_q-layer.5": 1296.5729,
      "encoder_q-layer.6": 1467.442,
      "encoder_q-layer.7": 1540.2766,
      "encoder_q-layer.8": 1585.2925,
      "encoder_q-layer.9": 1259.6162,
      "epoch": 0.39,
      "inbatch_neg_score": 0.1961,
      "inbatch_pos_score": 0.8359,
      "learning_rate": 3.316666666666667e-05,
      "loss": 3.353,
      "norm_diff": 0.0477,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2148.3447,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1971,
      "query_norm": 1.3672,
      "queue_k_norm": 1.4126,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8375,
      "sent_len_1": 66.7594,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.1438,
      "stdk": 0.0485,
      "stdq": 0.0456,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40300
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3157,
      "doc_norm": 1.407,
      "encoder_q-embeddings": 1499.9557,
      "encoder_q-layer.0": 945.1443,
      "encoder_q-layer.1": 1025.5975,
      "encoder_q-layer.10": 1235.222,
      "encoder_q-layer.11": 2687.2605,
      "encoder_q-layer.2": 1163.568,
      "encoder_q-layer.3": 1280.6097,
      "encoder_q-layer.4": 1359.3488,
      "encoder_q-layer.5": 1302.4471,
      "encoder_q-layer.6": 1506.0605,
      "encoder_q-layer.7": 1759.688,
      "encoder_q-layer.8": 1744.7697,
      "encoder_q-layer.9": 1358.0847,
      "epoch": 0.39,
      "inbatch_neg_score": 0.1938,
      "inbatch_pos_score": 0.8223,
      "learning_rate": 3.311111111111112e-05,
      "loss": 3.3157,
      "norm_diff": 0.054,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2279.0202,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1941,
      "query_norm": 1.353,
      "queue_k_norm": 1.4133,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1128,
      "sent_len_1": 66.8896,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9325,
      "stdk": 0.0482,
      "stdq": 0.045,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40400
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3395,
      "doc_norm": 1.415,
      "encoder_q-embeddings": 1482.4994,
      "encoder_q-layer.0": 1078.2573,
      "encoder_q-layer.1": 1100.6616,
      "encoder_q-layer.10": 1059.0797,
      "encoder_q-layer.11": 2365.8459,
      "encoder_q-layer.2": 1254.5537,
      "encoder_q-layer.3": 1342.7682,
      "encoder_q-layer.4": 1418.6234,
      "encoder_q-layer.5": 1410.4874,
      "encoder_q-layer.6": 1386.5547,
      "encoder_q-layer.7": 1478.4985,
      "encoder_q-layer.8": 1501.1417,
      "encoder_q-layer.9": 1201.2959,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1989,
      "inbatch_pos_score": 0.8779,
      "learning_rate": 3.3055555555555553e-05,
      "loss": 3.3395,
      "norm_diff": 0.0332,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2148.7766,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1996,
      "query_norm": 1.3818,
      "queue_k_norm": 1.413,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9781,
      "sent_len_1": 67.1243,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4938,
      "stdk": 0.0485,
      "stdq": 0.0459,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40500
    },
    {
      "accuracy": 58.4961,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3667,
      "doc_norm": 1.4157,
      "encoder_q-embeddings": 5215.4824,
      "encoder_q-layer.0": 3746.6145,
      "encoder_q-layer.1": 3617.563,
      "encoder_q-layer.10": 1191.9928,
      "encoder_q-layer.11": 2419.4578,
      "encoder_q-layer.2": 4326.498,
      "encoder_q-layer.3": 4866.5679,
      "encoder_q-layer.4": 5382.4336,
      "encoder_q-layer.5": 4525.0581,
      "encoder_q-layer.6": 4216.376,
      "encoder_q-layer.7": 3911.5842,
      "encoder_q-layer.8": 2579.3855,
      "encoder_q-layer.9": 1243.4432,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1973,
      "inbatch_pos_score": 0.8955,
      "learning_rate": 3.3e-05,
      "loss": 3.3667,
      "norm_diff": 0.0423,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5978.8301,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.1981,
      "query_norm": 1.3735,
      "queue_k_norm": 1.4114,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9598,
      "sent_len_1": 66.6363,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.025,
      "stdk": 0.0486,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40600
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3321,
      "doc_norm": 1.4174,
      "encoder_q-embeddings": 1146.5979,
      "encoder_q-layer.0": 726.337,
      "encoder_q-layer.1": 759.117,
      "encoder_q-layer.10": 1282.8011,
      "encoder_q-layer.11": 2601.7727,
      "encoder_q-layer.2": 844.0895,
      "encoder_q-layer.3": 896.6704,
      "encoder_q-layer.4": 1015.1094,
      "encoder_q-layer.5": 1046.1499,
      "encoder_q-layer.6": 1147.4677,
      "encoder_q-layer.7": 1204.4374,
      "encoder_q-layer.8": 1408.2594,
      "encoder_q-layer.9": 1231.9224,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1914,
      "inbatch_pos_score": 0.856,
      "learning_rate": 3.2944444444444445e-05,
      "loss": 3.3321,
      "norm_diff": 0.0359,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1909.4298,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1912,
      "query_norm": 1.3815,
      "queue_k_norm": 1.4144,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9772,
      "sent_len_1": 66.9108,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.78,
      "stdk": 0.0486,
      "stdq": 0.046,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 40700
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.35,
      "doc_norm": 1.4146,
      "encoder_q-embeddings": 5310.4268,
      "encoder_q-layer.0": 3904.5044,
      "encoder_q-layer.1": 3830.0532,
      "encoder_q-layer.10": 1200.8978,
      "encoder_q-layer.11": 2596.0166,
      "encoder_q-layer.2": 3857.1084,
      "encoder_q-layer.3": 3654.3933,
      "encoder_q-layer.4": 4104.5225,
      "encoder_q-layer.5": 4191.1934,
      "encoder_q-layer.6": 4490.9702,
      "encoder_q-layer.7": 4800.6152,
      "encoder_q-layer.8": 5841.5347,
      "encoder_q-layer.9": 2988.4592,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1849,
      "inbatch_pos_score": 0.8213,
      "learning_rate": 3.2888888888888894e-05,
      "loss": 3.35,
      "norm_diff": 0.0525,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6474.436,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.1852,
      "query_norm": 1.3621,
      "queue_k_norm": 1.4121,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8642,
      "sent_len_1": 66.8761,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9737,
      "stdk": 0.0485,
      "stdq": 0.0453,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40800
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3298,
      "doc_norm": 1.4108,
      "encoder_q-embeddings": 1585.4557,
      "encoder_q-layer.0": 1083.2382,
      "encoder_q-layer.1": 1204.5563,
      "encoder_q-layer.10": 1216.8109,
      "encoder_q-layer.11": 2544.6008,
      "encoder_q-layer.2": 1388.3041,
      "encoder_q-layer.3": 1431.5133,
      "encoder_q-layer.4": 1442.0792,
      "encoder_q-layer.5": 1497.8756,
      "encoder_q-layer.6": 1708.5956,
      "encoder_q-layer.7": 1760.917,
      "encoder_q-layer.8": 1708.4431,
      "encoder_q-layer.9": 1353.2325,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1847,
      "inbatch_pos_score": 0.8369,
      "learning_rate": 3.283333333333333e-05,
      "loss": 3.3298,
      "norm_diff": 0.0495,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2353.67,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1849,
      "query_norm": 1.3612,
      "queue_k_norm": 1.4114,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8692,
      "sent_len_1": 66.8079,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7363,
      "stdk": 0.0484,
      "stdq": 0.0452,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 40900
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3231,
      "doc_norm": 1.4149,
      "encoder_q-embeddings": 1676.5778,
      "encoder_q-layer.0": 1227.3209,
      "encoder_q-layer.1": 1347.6742,
      "encoder_q-layer.10": 1154.902,
      "encoder_q-layer.11": 2362.2046,
      "encoder_q-layer.2": 1612.9594,
      "encoder_q-layer.3": 1774.3479,
      "encoder_q-layer.4": 2004.0876,
      "encoder_q-layer.5": 2022.5013,
      "encoder_q-layer.6": 2109.1221,
      "encoder_q-layer.7": 2312.6245,
      "encoder_q-layer.8": 2431.5188,
      "encoder_q-layer.9": 1453.7163,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1846,
      "inbatch_pos_score": 0.8589,
      "learning_rate": 3.277777777777778e-05,
      "loss": 3.3231,
      "norm_diff": 0.0385,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2715.7326,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1844,
      "query_norm": 1.3764,
      "queue_k_norm": 1.4113,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0024,
      "sent_len_1": 66.7786,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9112,
      "stdk": 0.0485,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41000
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3638,
      "doc_norm": 1.4071,
      "encoder_q-embeddings": 1596.5911,
      "encoder_q-layer.0": 1084.8865,
      "encoder_q-layer.1": 1173.3629,
      "encoder_q-layer.10": 1214.7642,
      "encoder_q-layer.11": 2570.7539,
      "encoder_q-layer.2": 1345.5277,
      "encoder_q-layer.3": 1478.833,
      "encoder_q-layer.4": 1576.0919,
      "encoder_q-layer.5": 1676.6656,
      "encoder_q-layer.6": 1778.6035,
      "encoder_q-layer.7": 1941.8195,
      "encoder_q-layer.8": 1806.7634,
      "encoder_q-layer.9": 1325.3214,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1925,
      "inbatch_pos_score": 0.855,
      "learning_rate": 3.272222222222223e-05,
      "loss": 3.3638,
      "norm_diff": 0.0138,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2383.9609,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1919,
      "query_norm": 1.3933,
      "queue_k_norm": 1.4119,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0518,
      "sent_len_1": 66.6861,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.215,
      "stdk": 0.0483,
      "stdq": 0.046,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41100
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.361,
      "doc_norm": 1.4181,
      "encoder_q-embeddings": 651.9974,
      "encoder_q-layer.0": 454.0811,
      "encoder_q-layer.1": 481.657,
      "encoder_q-layer.10": 582.196,
      "encoder_q-layer.11": 1317.8243,
      "encoder_q-layer.2": 559.9383,
      "encoder_q-layer.3": 569.7593,
      "encoder_q-layer.4": 627.4794,
      "encoder_q-layer.5": 594.5319,
      "encoder_q-layer.6": 653.8876,
      "encoder_q-layer.7": 719.9487,
      "encoder_q-layer.8": 784.9109,
      "encoder_q-layer.9": 645.6687,
      "epoch": 0.4,
      "inbatch_neg_score": 0.2043,
      "inbatch_pos_score": 0.8623,
      "learning_rate": 3.266666666666667e-05,
      "loss": 3.361,
      "norm_diff": 0.035,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1054.0056,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2036,
      "query_norm": 1.3831,
      "queue_k_norm": 1.4117,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8879,
      "sent_len_1": 66.867,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.18,
      "stdk": 0.0487,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41200
    },
    {
      "accuracy": 51.1719,
      "active_queue_size": 16384.0,
      "cl_loss": 3.333,
      "doc_norm": 1.4172,
      "encoder_q-embeddings": 881.0007,
      "encoder_q-layer.0": 627.4488,
      "encoder_q-layer.1": 682.9316,
      "encoder_q-layer.10": 588.9636,
      "encoder_q-layer.11": 1305.8643,
      "encoder_q-layer.2": 828.9022,
      "encoder_q-layer.3": 906.1451,
      "encoder_q-layer.4": 1053.873,
      "encoder_q-layer.5": 1152.8063,
      "encoder_q-layer.6": 1304.0927,
      "encoder_q-layer.7": 1513.939,
      "encoder_q-layer.8": 1534.394,
      "encoder_q-layer.9": 975.4919,
      "epoch": 0.4,
      "inbatch_neg_score": 0.1964,
      "inbatch_pos_score": 0.8418,
      "learning_rate": 3.261111111111111e-05,
      "loss": 3.333,
      "norm_diff": 0.0317,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1588.3506,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1963,
      "query_norm": 1.3855,
      "queue_k_norm": 1.4127,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6193,
      "sent_len_1": 66.6505,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2413,
      "stdk": 0.0486,
      "stdq": 0.0456,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 41300
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3662,
      "doc_norm": 1.4129,
      "encoder_q-embeddings": 948.6764,
      "encoder_q-layer.0": 626.0134,
      "encoder_q-layer.1": 703.9127,
      "encoder_q-layer.10": 666.0769,
      "encoder_q-layer.11": 1320.6569,
      "encoder_q-layer.2": 763.251,
      "encoder_q-layer.3": 782.788,
      "encoder_q-layer.4": 818.9739,
      "encoder_q-layer.5": 901.8481,
      "encoder_q-layer.6": 969.7435,
      "encoder_q-layer.7": 1185.9785,
      "encoder_q-layer.8": 1242.5299,
      "encoder_q-layer.9": 736.8083,
      "epoch": 0.4,
      "inbatch_neg_score": 0.2035,
      "inbatch_pos_score": 0.8701,
      "learning_rate": 3.2555555555555555e-05,
      "loss": 3.3662,
      "norm_diff": 0.0214,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1376.6371,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2041,
      "query_norm": 1.3997,
      "queue_k_norm": 1.4138,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.043,
      "sent_len_1": 66.8378,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.81,
      "stdk": 0.0485,
      "stdq": 0.046,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 41400
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3347,
      "doc_norm": 1.41,
      "encoder_q-embeddings": 709.8692,
      "encoder_q-layer.0": 478.3379,
      "encoder_q-layer.1": 504.4007,
      "encoder_q-layer.10": 681.1744,
      "encoder_q-layer.11": 1298.7441,
      "encoder_q-layer.2": 564.3962,
      "encoder_q-layer.3": 623.1088,
      "encoder_q-layer.4": 657.1346,
      "encoder_q-layer.5": 671.4012,
      "encoder_q-layer.6": 742.2836,
      "encoder_q-layer.7": 855.5963,
      "encoder_q-layer.8": 1140.4753,
      "encoder_q-layer.9": 904.9538,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2103,
      "inbatch_pos_score": 0.8574,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 3.3347,
      "norm_diff": 0.0152,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1165.7722,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.209,
      "query_norm": 1.3976,
      "queue_k_norm": 1.4112,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9307,
      "sent_len_1": 66.7849,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1687,
      "stdk": 0.0484,
      "stdq": 0.0459,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41500
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3183,
      "doc_norm": 1.4108,
      "encoder_q-embeddings": 644.4532,
      "encoder_q-layer.0": 400.3166,
      "encoder_q-layer.1": 435.7461,
      "encoder_q-layer.10": 627.6603,
      "encoder_q-layer.11": 1395.3156,
      "encoder_q-layer.2": 489.9265,
      "encoder_q-layer.3": 528.2087,
      "encoder_q-layer.4": 545.0742,
      "encoder_q-layer.5": 624.3137,
      "encoder_q-layer.6": 724.6846,
      "encoder_q-layer.7": 912.7904,
      "encoder_q-layer.8": 1157.3857,
      "encoder_q-layer.9": 904.647,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2164,
      "inbatch_pos_score": 0.8774,
      "learning_rate": 3.2444444444444446e-05,
      "loss": 3.3183,
      "norm_diff": 0.009,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1176.9049,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2161,
      "query_norm": 1.4058,
      "queue_k_norm": 1.4118,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8785,
      "sent_len_1": 66.6637,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7562,
      "stdk": 0.0484,
      "stdq": 0.0462,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41600
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.358,
      "doc_norm": 1.4149,
      "encoder_q-embeddings": 651.5169,
      "encoder_q-layer.0": 416.6948,
      "encoder_q-layer.1": 490.3955,
      "encoder_q-layer.10": 628.271,
      "encoder_q-layer.11": 1381.8673,
      "encoder_q-layer.2": 597.9086,
      "encoder_q-layer.3": 608.101,
      "encoder_q-layer.4": 628.2769,
      "encoder_q-layer.5": 660.2829,
      "encoder_q-layer.6": 683.5198,
      "encoder_q-layer.7": 759.3759,
      "encoder_q-layer.8": 748.0917,
      "encoder_q-layer.9": 616.4691,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2182,
      "inbatch_pos_score": 0.8701,
      "learning_rate": 3.238888888888889e-05,
      "loss": 3.358,
      "norm_diff": 0.0433,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1072.2726,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2183,
      "query_norm": 1.3716,
      "queue_k_norm": 1.4129,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7131,
      "sent_len_1": 66.5699,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5863,
      "stdk": 0.0485,
      "stdq": 0.045,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41700
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3211,
      "doc_norm": 1.4112,
      "encoder_q-embeddings": 772.6971,
      "encoder_q-layer.0": 492.7506,
      "encoder_q-layer.1": 606.5897,
      "encoder_q-layer.10": 579.0443,
      "encoder_q-layer.11": 1296.6702,
      "encoder_q-layer.2": 726.4323,
      "encoder_q-layer.3": 789.9049,
      "encoder_q-layer.4": 784.8657,
      "encoder_q-layer.5": 777.8354,
      "encoder_q-layer.6": 835.2467,
      "encoder_q-layer.7": 824.7648,
      "encoder_q-layer.8": 786.7727,
      "encoder_q-layer.9": 625.9105,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2192,
      "inbatch_pos_score": 0.8911,
      "learning_rate": 3.233333333333333e-05,
      "loss": 3.3211,
      "norm_diff": 0.0298,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1166.9606,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.219,
      "query_norm": 1.3813,
      "queue_k_norm": 1.4133,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9235,
      "sent_len_1": 66.7277,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3938,
      "stdk": 0.0483,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41800
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3322,
      "doc_norm": 1.413,
      "encoder_q-embeddings": 991.7081,
      "encoder_q-layer.0": 655.4652,
      "encoder_q-layer.1": 780.5581,
      "encoder_q-layer.10": 590.3641,
      "encoder_q-layer.11": 1272.3623,
      "encoder_q-layer.2": 935.1348,
      "encoder_q-layer.3": 906.8386,
      "encoder_q-layer.4": 1054.3077,
      "encoder_q-layer.5": 1034.5135,
      "encoder_q-layer.6": 1089.2478,
      "encoder_q-layer.7": 1194.8716,
      "encoder_q-layer.8": 931.3466,
      "encoder_q-layer.9": 642.7172,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2181,
      "inbatch_pos_score": 0.8848,
      "learning_rate": 3.227777777777778e-05,
      "loss": 3.3322,
      "norm_diff": 0.038,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1381.1061,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2181,
      "query_norm": 1.375,
      "queue_k_norm": 1.4148,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2039,
      "sent_len_1": 67.0027,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.715,
      "stdk": 0.0484,
      "stdq": 0.0455,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 41900
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3387,
      "doc_norm": 1.4151,
      "encoder_q-embeddings": 685.9819,
      "encoder_q-layer.0": 434.4364,
      "encoder_q-layer.1": 463.9839,
      "encoder_q-layer.10": 568.3271,
      "encoder_q-layer.11": 1287.0354,
      "encoder_q-layer.2": 525.3226,
      "encoder_q-layer.3": 561.5266,
      "encoder_q-layer.4": 613.5241,
      "encoder_q-layer.5": 561.6017,
      "encoder_q-layer.6": 576.6347,
      "encoder_q-layer.7": 621.4035,
      "encoder_q-layer.8": 690.1999,
      "encoder_q-layer.9": 584.9044,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2143,
      "inbatch_pos_score": 0.8462,
      "learning_rate": 3.222222222222223e-05,
      "loss": 3.3387,
      "norm_diff": 0.0571,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 989.1467,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2142,
      "query_norm": 1.358,
      "queue_k_norm": 1.4155,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9631,
      "sent_len_1": 66.8029,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6962,
      "stdk": 0.0484,
      "stdq": 0.045,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 42000
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3566,
      "doc_norm": 1.4181,
      "encoder_q-embeddings": 871.4882,
      "encoder_q-layer.0": 606.5714,
      "encoder_q-layer.1": 643.5759,
      "encoder_q-layer.10": 621.8083,
      "encoder_q-layer.11": 1331.6365,
      "encoder_q-layer.2": 741.368,
      "encoder_q-layer.3": 781.0949,
      "encoder_q-layer.4": 826.465,
      "encoder_q-layer.5": 854.5236,
      "encoder_q-layer.6": 857.2499,
      "encoder_q-layer.7": 967.4739,
      "encoder_q-layer.8": 933.974,
      "encoder_q-layer.9": 681.6948,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2131,
      "inbatch_pos_score": 0.8438,
      "learning_rate": 3.2166666666666665e-05,
      "loss": 3.3566,
      "norm_diff": 0.061,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1264.5897,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2131,
      "query_norm": 1.3572,
      "queue_k_norm": 1.4154,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9644,
      "sent_len_1": 66.5739,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8725,
      "stdk": 0.0485,
      "stdq": 0.045,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 42100
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3255,
      "doc_norm": 1.4164,
      "encoder_q-embeddings": 1921.0294,
      "encoder_q-layer.0": 1611.7421,
      "encoder_q-layer.1": 1869.7224,
      "encoder_q-layer.10": 610.2836,
      "encoder_q-layer.11": 1311.0249,
      "encoder_q-layer.2": 1833.1721,
      "encoder_q-layer.3": 1735.321,
      "encoder_q-layer.4": 1861.1237,
      "encoder_q-layer.5": 1904.1305,
      "encoder_q-layer.6": 2038.0642,
      "encoder_q-layer.7": 1703.3829,
      "encoder_q-layer.8": 1303.5468,
      "encoder_q-layer.9": 678.4182,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2137,
      "inbatch_pos_score": 0.8804,
      "learning_rate": 3.2111111111111114e-05,
      "loss": 3.3255,
      "norm_diff": 0.0469,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2470.1605,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2125,
      "query_norm": 1.3695,
      "queue_k_norm": 1.4168,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0218,
      "sent_len_1": 66.6533,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.035,
      "stdk": 0.0485,
      "stdq": 0.0455,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42200
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3255,
      "doc_norm": 1.4199,
      "encoder_q-embeddings": 618.2369,
      "encoder_q-layer.0": 413.3218,
      "encoder_q-layer.1": 463.7318,
      "encoder_q-layer.10": 559.4675,
      "encoder_q-layer.11": 1218.4253,
      "encoder_q-layer.2": 592.8334,
      "encoder_q-layer.3": 627.7633,
      "encoder_q-layer.4": 688.7883,
      "encoder_q-layer.5": 698.4971,
      "encoder_q-layer.6": 744.4785,
      "encoder_q-layer.7": 824.3997,
      "encoder_q-layer.8": 816.7671,
      "encoder_q-layer.9": 610.4033,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2068,
      "inbatch_pos_score": 0.8882,
      "learning_rate": 3.2055555555555556e-05,
      "loss": 3.3255,
      "norm_diff": 0.0258,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1043.2435,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2074,
      "query_norm": 1.3941,
      "queue_k_norm": 1.417,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9682,
      "sent_len_1": 66.7896,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7663,
      "stdk": 0.0486,
      "stdq": 0.0465,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42300
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3321,
      "doc_norm": 1.419,
      "encoder_q-embeddings": 667.4397,
      "encoder_q-layer.0": 463.5435,
      "encoder_q-layer.1": 453.595,
      "encoder_q-layer.10": 602.5529,
      "encoder_q-layer.11": 1290.707,
      "encoder_q-layer.2": 513.7106,
      "encoder_q-layer.3": 541.5753,
      "encoder_q-layer.4": 557.9749,
      "encoder_q-layer.5": 586.0375,
      "encoder_q-layer.6": 642.5804,
      "encoder_q-layer.7": 727.7101,
      "encoder_q-layer.8": 757.5709,
      "encoder_q-layer.9": 596.0701,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2033,
      "inbatch_pos_score": 0.9004,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 3.3321,
      "norm_diff": 0.0354,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1020.3585,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2054,
      "query_norm": 1.3836,
      "queue_k_norm": 1.4177,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9438,
      "sent_len_1": 66.6694,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.025,
      "stdk": 0.0486,
      "stdq": 0.0463,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42400
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3608,
      "doc_norm": 1.4197,
      "encoder_q-embeddings": 640.9758,
      "encoder_q-layer.0": 457.6689,
      "encoder_q-layer.1": 526.881,
      "encoder_q-layer.10": 620.9788,
      "encoder_q-layer.11": 1265.7422,
      "encoder_q-layer.2": 648.8992,
      "encoder_q-layer.3": 612.3759,
      "encoder_q-layer.4": 619.426,
      "encoder_q-layer.5": 642.4967,
      "encoder_q-layer.6": 682.1539,
      "encoder_q-layer.7": 872.7705,
      "encoder_q-layer.8": 826.871,
      "encoder_q-layer.9": 670.5676,
      "epoch": 0.41,
      "inbatch_neg_score": 0.2048,
      "inbatch_pos_score": 0.8867,
      "learning_rate": 3.194444444444444e-05,
      "loss": 3.3608,
      "norm_diff": 0.0458,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1075.3256,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2036,
      "query_norm": 1.3738,
      "queue_k_norm": 1.416,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9105,
      "sent_len_1": 66.56,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9762,
      "stdk": 0.0486,
      "stdq": 0.0459,
      "stdqueue_k": 0.0485,
      "stdqueue_q": 0.0,
      "step": 42500
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3175,
      "doc_norm": 1.4142,
      "encoder_q-embeddings": 888.2637,
      "encoder_q-layer.0": 628.2617,
      "encoder_q-layer.1": 696.5986,
      "encoder_q-layer.10": 646.1299,
      "encoder_q-layer.11": 1317.4252,
      "encoder_q-layer.2": 766.0026,
      "encoder_q-layer.3": 793.8303,
      "encoder_q-layer.4": 791.4348,
      "encoder_q-layer.5": 831.7206,
      "encoder_q-layer.6": 866.4728,
      "encoder_q-layer.7": 888.8124,
      "encoder_q-layer.8": 854.0746,
      "encoder_q-layer.9": 677.5018,
      "epoch": 0.42,
      "inbatch_neg_score": 0.2001,
      "inbatch_pos_score": 0.8564,
      "learning_rate": 3.188888888888889e-05,
      "loss": 3.3175,
      "norm_diff": 0.0518,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1251.3704,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2007,
      "query_norm": 1.3625,
      "queue_k_norm": 1.417,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9007,
      "sent_len_1": 66.724,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7063,
      "stdk": 0.0484,
      "stdq": 0.0456,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42600
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3132,
      "doc_norm": 1.4262,
      "encoder_q-embeddings": 646.3419,
      "encoder_q-layer.0": 450.6562,
      "encoder_q-layer.1": 480.3102,
      "encoder_q-layer.10": 585.9885,
      "encoder_q-layer.11": 1276.0741,
      "encoder_q-layer.2": 533.89,
      "encoder_q-layer.3": 544.4371,
      "encoder_q-layer.4": 563.3469,
      "encoder_q-layer.5": 583.7206,
      "encoder_q-layer.6": 592.3884,
      "encoder_q-layer.7": 696.0845,
      "encoder_q-layer.8": 731.8017,
      "encoder_q-layer.9": 604.5693,
      "epoch": 0.42,
      "inbatch_neg_score": 0.2042,
      "inbatch_pos_score": 0.9004,
      "learning_rate": 3.183333333333334e-05,
      "loss": 3.3132,
      "norm_diff": 0.0546,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1000.3268,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2036,
      "query_norm": 1.3715,
      "queue_k_norm": 1.4165,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9338,
      "sent_len_1": 66.6621,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8638,
      "stdk": 0.0488,
      "stdq": 0.0459,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42700
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3339,
      "doc_norm": 1.4142,
      "encoder_q-embeddings": 565.9507,
      "encoder_q-layer.0": 356.8761,
      "encoder_q-layer.1": 382.2411,
      "encoder_q-layer.10": 622.2243,
      "encoder_q-layer.11": 1338.1033,
      "encoder_q-layer.2": 427.2748,
      "encoder_q-layer.3": 452.442,
      "encoder_q-layer.4": 468.8776,
      "encoder_q-layer.5": 483.1323,
      "encoder_q-layer.6": 559.0262,
      "encoder_q-layer.7": 670.1913,
      "encoder_q-layer.8": 742.8456,
      "encoder_q-layer.9": 647.2837,
      "epoch": 0.42,
      "inbatch_neg_score": 0.2058,
      "inbatch_pos_score": 0.8828,
      "learning_rate": 3.177777777777778e-05,
      "loss": 3.3339,
      "norm_diff": 0.0369,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 945.7569,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2052,
      "query_norm": 1.3773,
      "queue_k_norm": 1.418,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9515,
      "sent_len_1": 66.7725,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.22,
      "stdk": 0.0484,
      "stdq": 0.0461,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42800
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3196,
      "doc_norm": 1.422,
      "encoder_q-embeddings": 652.2833,
      "encoder_q-layer.0": 427.7899,
      "encoder_q-layer.1": 459.8958,
      "encoder_q-layer.10": 617.1335,
      "encoder_q-layer.11": 1294.7386,
      "encoder_q-layer.2": 526.9139,
      "encoder_q-layer.3": 587.7241,
      "encoder_q-layer.4": 629.0897,
      "encoder_q-layer.5": 649.8214,
      "encoder_q-layer.6": 735.3761,
      "encoder_q-layer.7": 901.1779,
      "encoder_q-layer.8": 939.2888,
      "encoder_q-layer.9": 766.4161,
      "epoch": 0.42,
      "inbatch_neg_score": 0.211,
      "inbatch_pos_score": 0.8887,
      "learning_rate": 3.1722222222222224e-05,
      "loss": 3.3196,
      "norm_diff": 0.0597,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1103.248,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2109,
      "query_norm": 1.3622,
      "queue_k_norm": 1.4168,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9112,
      "sent_len_1": 66.7907,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0288,
      "stdk": 0.0487,
      "stdq": 0.0454,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 42900
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3412,
      "doc_norm": 1.4182,
      "encoder_q-embeddings": 652.647,
      "encoder_q-layer.0": 415.5762,
      "encoder_q-layer.1": 442.8323,
      "encoder_q-layer.10": 650.6154,
      "encoder_q-layer.11": 1426.2939,
      "encoder_q-layer.2": 485.8511,
      "encoder_q-layer.3": 538.4459,
      "encoder_q-layer.4": 582.2295,
      "encoder_q-layer.5": 631.8797,
      "encoder_q-layer.6": 694.65,
      "encoder_q-layer.7": 749.7224,
      "encoder_q-layer.8": 793.1775,
      "encoder_q-layer.9": 681.6508,
      "epoch": 0.42,
      "inbatch_neg_score": 0.2114,
      "inbatch_pos_score": 0.8467,
      "learning_rate": 3.1666666666666666e-05,
      "loss": 3.3412,
      "norm_diff": 0.0498,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1074.2289,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2113,
      "query_norm": 1.3684,
      "queue_k_norm": 1.4196,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9238,
      "sent_len_1": 66.9152,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.59,
      "stdk": 0.0485,
      "stdq": 0.0455,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 43000
    },
    {
      "accuracy": 52.9297,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3403,
      "doc_norm": 1.4247,
      "encoder_q-embeddings": 324.053,
      "encoder_q-layer.0": 216.7407,
      "encoder_q-layer.1": 255.9641,
      "encoder_q-layer.10": 318.1435,
      "encoder_q-layer.11": 676.2004,
      "encoder_q-layer.2": 301.6189,
      "encoder_q-layer.3": 321.2693,
      "encoder_q-layer.4": 392.086,
      "encoder_q-layer.5": 382.9891,
      "encoder_q-layer.6": 449.8855,
      "encoder_q-layer.7": 437.3492,
      "encoder_q-layer.8": 397.9259,
      "encoder_q-layer.9": 316.1306,
      "epoch": 0.42,
      "inbatch_neg_score": 0.2081,
      "inbatch_pos_score": 0.8545,
      "learning_rate": 3.1611111111111115e-05,
      "loss": 3.3403,
      "norm_diff": 0.0688,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 568.2427,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2076,
      "query_norm": 1.3559,
      "queue_k_norm": 1.4182,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9151,
      "sent_len_1": 66.6933,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5637,
      "stdk": 0.0488,
      "stdq": 0.0451,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43100
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3147,
      "doc_norm": 1.4173,
      "encoder_q-embeddings": 642.0334,
      "encoder_q-layer.0": 447.5197,
      "encoder_q-layer.1": 441.7097,
      "encoder_q-layer.10": 308.4878,
      "encoder_q-layer.11": 668.8505,
      "encoder_q-layer.2": 490.8317,
      "encoder_q-layer.3": 509.7571,
      "encoder_q-layer.4": 579.9095,
      "encoder_q-layer.5": 565.4186,
      "encoder_q-layer.6": 627.3035,
      "encoder_q-layer.7": 767.9724,
      "encoder_q-layer.8": 640.5216,
      "encoder_q-layer.9": 335.2906,
      "epoch": 0.42,
      "inbatch_neg_score": 0.2048,
      "inbatch_pos_score": 0.8511,
      "learning_rate": 3.155555555555556e-05,
      "loss": 3.3147,
      "norm_diff": 0.0469,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 828.5536,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2039,
      "query_norm": 1.3704,
      "queue_k_norm": 1.4176,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9996,
      "sent_len_1": 66.5101,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.0163,
      "stdk": 0.0485,
      "stdq": 0.0458,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43200
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3457,
      "doc_norm": 1.4139,
      "encoder_q-embeddings": 643.7472,
      "encoder_q-layer.0": 433.6754,
      "encoder_q-layer.1": 489.727,
      "encoder_q-layer.10": 275.3564,
      "encoder_q-layer.11": 602.9526,
      "encoder_q-layer.2": 580.1151,
      "encoder_q-layer.3": 626.0225,
      "encoder_q-layer.4": 586.2537,
      "encoder_q-layer.5": 560.5406,
      "encoder_q-layer.6": 552.6998,
      "encoder_q-layer.7": 493.8007,
      "encoder_q-layer.8": 386.7263,
      "encoder_q-layer.9": 288.9821,
      "epoch": 0.42,
      "inbatch_neg_score": 0.1956,
      "inbatch_pos_score": 0.8784,
      "learning_rate": 3.15e-05,
      "loss": 3.3457,
      "norm_diff": 0.0456,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 786.5784,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1962,
      "query_norm": 1.3683,
      "queue_k_norm": 1.4175,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8636,
      "sent_len_1": 66.7243,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6825,
      "stdk": 0.0484,
      "stdq": 0.0458,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43300
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2983,
      "doc_norm": 1.419,
      "encoder_q-embeddings": 1714.1003,
      "encoder_q-layer.0": 1258.0769,
      "encoder_q-layer.1": 1255.7178,
      "encoder_q-layer.10": 352.252,
      "encoder_q-layer.11": 672.8743,
      "encoder_q-layer.2": 1478.099,
      "encoder_q-layer.3": 1699.9553,
      "encoder_q-layer.4": 1821.2969,
      "encoder_q-layer.5": 1572.5426,
      "encoder_q-layer.6": 1852.5077,
      "encoder_q-layer.7": 1883.7384,
      "encoder_q-layer.8": 1264.2515,
      "encoder_q-layer.9": 418.8791,
      "epoch": 0.42,
      "inbatch_neg_score": 0.1932,
      "inbatch_pos_score": 0.8486,
      "learning_rate": 3.144444444444445e-05,
      "loss": 3.2983,
      "norm_diff": 0.0493,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2184.948,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.194,
      "query_norm": 1.3697,
      "queue_k_norm": 1.4172,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0522,
      "sent_len_1": 66.7646,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9563,
      "stdk": 0.0486,
      "stdq": 0.046,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43400
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3228,
      "doc_norm": 1.4201,
      "encoder_q-embeddings": 547.2619,
      "encoder_q-layer.0": 370.1842,
      "encoder_q-layer.1": 368.4086,
      "encoder_q-layer.10": 319.984,
      "encoder_q-layer.11": 665.526,
      "encoder_q-layer.2": 483.1642,
      "encoder_q-layer.3": 460.4322,
      "encoder_q-layer.4": 514.5997,
      "encoder_q-layer.5": 560.1196,
      "encoder_q-layer.6": 602.4464,
      "encoder_q-layer.7": 595.7255,
      "encoder_q-layer.8": 458.3723,
      "encoder_q-layer.9": 316.4586,
      "epoch": 0.42,
      "inbatch_neg_score": 0.1964,
      "inbatch_pos_score": 0.8682,
      "learning_rate": 3.138888888888889e-05,
      "loss": 3.3228,
      "norm_diff": 0.0584,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 739.394,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1969,
      "query_norm": 1.3617,
      "queue_k_norm": 1.4157,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.884,
      "sent_len_1": 66.8179,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.165,
      "stdk": 0.0486,
      "stdq": 0.0455,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43500
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3038,
      "doc_norm": 1.4131,
      "encoder_q-embeddings": 291.5234,
      "encoder_q-layer.0": 186.2465,
      "encoder_q-layer.1": 205.7701,
      "encoder_q-layer.10": 281.7762,
      "encoder_q-layer.11": 626.8435,
      "encoder_q-layer.2": 235.476,
      "encoder_q-layer.3": 257.949,
      "encoder_q-layer.4": 273.9988,
      "encoder_q-layer.5": 280.9333,
      "encoder_q-layer.6": 314.0224,
      "encoder_q-layer.7": 362.0202,
      "encoder_q-layer.8": 368.3067,
      "encoder_q-layer.9": 315.1465,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1984,
      "inbatch_pos_score": 0.8838,
      "learning_rate": 3.1333333333333334e-05,
      "loss": 3.3038,
      "norm_diff": 0.0394,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 480.4719,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1982,
      "query_norm": 1.3737,
      "queue_k_norm": 1.4167,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0912,
      "sent_len_1": 66.9852,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0775,
      "stdk": 0.0484,
      "stdq": 0.0459,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43600
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2844,
      "doc_norm": 1.4126,
      "encoder_q-embeddings": 247.6566,
      "encoder_q-layer.0": 165.9463,
      "encoder_q-layer.1": 180.7304,
      "encoder_q-layer.10": 286.089,
      "encoder_q-layer.11": 639.0943,
      "encoder_q-layer.2": 201.9543,
      "encoder_q-layer.3": 205.0382,
      "encoder_q-layer.4": 220.9973,
      "encoder_q-layer.5": 230.0922,
      "encoder_q-layer.6": 260.8263,
      "encoder_q-layer.7": 298.4917,
      "encoder_q-layer.8": 338.9518,
      "encoder_q-layer.9": 296.118,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1943,
      "inbatch_pos_score": 0.8384,
      "learning_rate": 3.1277777777777776e-05,
      "loss": 3.2844,
      "norm_diff": 0.0613,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 440.7413,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1946,
      "query_norm": 1.3512,
      "queue_k_norm": 1.4165,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9848,
      "sent_len_1": 66.7873,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.31,
      "stdk": 0.0484,
      "stdq": 0.0452,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43700
    },
    {
      "accuracy": 51.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3191,
      "doc_norm": 1.4138,
      "encoder_q-embeddings": 462.9347,
      "encoder_q-layer.0": 319.8843,
      "encoder_q-layer.1": 325.8416,
      "encoder_q-layer.10": 338.8963,
      "encoder_q-layer.11": 670.9633,
      "encoder_q-layer.2": 357.0933,
      "encoder_q-layer.3": 385.9306,
      "encoder_q-layer.4": 406.1962,
      "encoder_q-layer.5": 443.9343,
      "encoder_q-layer.6": 473.923,
      "encoder_q-layer.7": 460.0476,
      "encoder_q-layer.8": 429.6974,
      "encoder_q-layer.9": 327.0101,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1965,
      "inbatch_pos_score": 0.8169,
      "learning_rate": 3.1222222222222225e-05,
      "loss": 3.3191,
      "norm_diff": 0.065,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 634.4755,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1967,
      "query_norm": 1.3488,
      "queue_k_norm": 1.4162,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9725,
      "sent_len_1": 66.7369,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8275,
      "stdk": 0.0484,
      "stdq": 0.0449,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43800
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3195,
      "doc_norm": 1.416,
      "encoder_q-embeddings": 262.5099,
      "encoder_q-layer.0": 177.4348,
      "encoder_q-layer.1": 194.7729,
      "encoder_q-layer.10": 320.0033,
      "encoder_q-layer.11": 624.7731,
      "encoder_q-layer.2": 213.2238,
      "encoder_q-layer.3": 225.2675,
      "encoder_q-layer.4": 232.2906,
      "encoder_q-layer.5": 251.0912,
      "encoder_q-layer.6": 282.781,
      "encoder_q-layer.7": 326.2191,
      "encoder_q-layer.8": 394.0247,
      "encoder_q-layer.9": 321.9742,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1973,
      "inbatch_pos_score": 0.8535,
      "learning_rate": 3.116666666666667e-05,
      "loss": 3.3195,
      "norm_diff": 0.0349,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 465.0568,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1965,
      "query_norm": 1.3811,
      "queue_k_norm": 1.4163,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0631,
      "sent_len_1": 66.7689,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.165,
      "stdk": 0.0485,
      "stdq": 0.0463,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 43900
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3244,
      "doc_norm": 1.4169,
      "encoder_q-embeddings": 289.4416,
      "encoder_q-layer.0": 179.4188,
      "encoder_q-layer.1": 186.8819,
      "encoder_q-layer.10": 290.0345,
      "encoder_q-layer.11": 638.5543,
      "encoder_q-layer.2": 209.779,
      "encoder_q-layer.3": 225.065,
      "encoder_q-layer.4": 247.7596,
      "encoder_q-layer.5": 259.0912,
      "encoder_q-layer.6": 291.816,
      "encoder_q-layer.7": 337.4434,
      "encoder_q-layer.8": 378.7804,
      "encoder_q-layer.9": 313.7754,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1965,
      "inbatch_pos_score": 0.8955,
      "learning_rate": 3.111111111111111e-05,
      "loss": 3.3244,
      "norm_diff": 0.0239,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 469.6961,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1959,
      "query_norm": 1.393,
      "queue_k_norm": 1.415,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.732,
      "sent_len_1": 66.8841,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4025,
      "stdk": 0.0485,
      "stdq": 0.0466,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 44000
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3419,
      "doc_norm": 1.4165,
      "encoder_q-embeddings": 488.8597,
      "encoder_q-layer.0": 325.5974,
      "encoder_q-layer.1": 335.3644,
      "encoder_q-layer.10": 285.6553,
      "encoder_q-layer.11": 633.9598,
      "encoder_q-layer.2": 401.3664,
      "encoder_q-layer.3": 406.3181,
      "encoder_q-layer.4": 494.4386,
      "encoder_q-layer.5": 538.5054,
      "encoder_q-layer.6": 671.9125,
      "encoder_q-layer.7": 717.4082,
      "encoder_q-layer.8": 578.7513,
      "encoder_q-layer.9": 399.9344,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1967,
      "inbatch_pos_score": 0.8296,
      "learning_rate": 3.105555555555555e-05,
      "loss": 3.3419,
      "norm_diff": 0.0591,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 739.6686,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1968,
      "query_norm": 1.3574,
      "queue_k_norm": 1.4158,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9649,
      "sent_len_1": 66.6701,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7912,
      "stdk": 0.0485,
      "stdq": 0.0453,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 44100
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3189,
      "doc_norm": 1.4208,
      "encoder_q-embeddings": 286.7896,
      "encoder_q-layer.0": 184.0093,
      "encoder_q-layer.1": 200.5699,
      "encoder_q-layer.10": 296.0781,
      "encoder_q-layer.11": 608.0043,
      "encoder_q-layer.2": 229.4677,
      "encoder_q-layer.3": 237.2131,
      "encoder_q-layer.4": 241.926,
      "encoder_q-layer.5": 261.6401,
      "encoder_q-layer.6": 301.3804,
      "encoder_q-layer.7": 358.3695,
      "encoder_q-layer.8": 339.4464,
      "encoder_q-layer.9": 288.5755,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1961,
      "inbatch_pos_score": 0.9106,
      "learning_rate": 3.1e-05,
      "loss": 3.3189,
      "norm_diff": 0.0292,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 461.4437,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1965,
      "query_norm": 1.3916,
      "queue_k_norm": 1.4182,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.112,
      "sent_len_1": 66.7092,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4162,
      "stdk": 0.0487,
      "stdq": 0.0466,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 44200
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.329,
      "doc_norm": 1.4123,
      "encoder_q-embeddings": 584.4637,
      "encoder_q-layer.0": 424.2446,
      "encoder_q-layer.1": 482.28,
      "encoder_q-layer.10": 288.3389,
      "encoder_q-layer.11": 637.2018,
      "encoder_q-layer.2": 558.9258,
      "encoder_q-layer.3": 596.7576,
      "encoder_q-layer.4": 612.652,
      "encoder_q-layer.5": 626.6475,
      "encoder_q-layer.6": 596.9974,
      "encoder_q-layer.7": 638.5096,
      "encoder_q-layer.8": 550.6765,
      "encoder_q-layer.9": 308.0601,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1981,
      "inbatch_pos_score": 0.8301,
      "learning_rate": 3.094444444444445e-05,
      "loss": 3.329,
      "norm_diff": 0.0521,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 823.6682,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1975,
      "query_norm": 1.3602,
      "queue_k_norm": 1.4159,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.964,
      "sent_len_1": 66.7947,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1138,
      "stdk": 0.0484,
      "stdq": 0.0453,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 44300
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2984,
      "doc_norm": 1.4138,
      "encoder_q-embeddings": 600.5946,
      "encoder_q-layer.0": 428.7651,
      "encoder_q-layer.1": 458.8206,
      "encoder_q-layer.10": 322.7917,
      "encoder_q-layer.11": 664.9639,
      "encoder_q-layer.2": 543.8619,
      "encoder_q-layer.3": 569.5438,
      "encoder_q-layer.4": 605.7654,
      "encoder_q-layer.5": 571.5055,
      "encoder_q-layer.6": 495.3671,
      "encoder_q-layer.7": 501.3279,
      "encoder_q-layer.8": 456.2144,
      "encoder_q-layer.9": 321.4072,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1969,
      "inbatch_pos_score": 0.8354,
      "learning_rate": 3.088888888888889e-05,
      "loss": 3.2984,
      "norm_diff": 0.039,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 772.0803,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.198,
      "query_norm": 1.3748,
      "queue_k_norm": 1.416,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9307,
      "sent_len_1": 66.8017,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9013,
      "stdk": 0.0485,
      "stdq": 0.0457,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 44400
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3144,
      "doc_norm": 1.4173,
      "encoder_q-embeddings": 336.8156,
      "encoder_q-layer.0": 218.2655,
      "encoder_q-layer.1": 232.7871,
      "encoder_q-layer.10": 293.2474,
      "encoder_q-layer.11": 658.017,
      "encoder_q-layer.2": 268.3653,
      "encoder_q-layer.3": 274.4176,
      "encoder_q-layer.4": 297.9587,
      "encoder_q-layer.5": 310.8348,
      "encoder_q-layer.6": 332.9094,
      "encoder_q-layer.7": 367.5707,
      "encoder_q-layer.8": 388.3803,
      "encoder_q-layer.9": 317.8849,
      "epoch": 0.43,
      "inbatch_neg_score": 0.1983,
      "inbatch_pos_score": 0.8296,
      "learning_rate": 3.0833333333333335e-05,
      "loss": 3.3144,
      "norm_diff": 0.0472,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 519.2402,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1989,
      "query_norm": 1.3701,
      "queue_k_norm": 1.4143,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8951,
      "sent_len_1": 66.6784,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6875,
      "stdk": 0.0486,
      "stdq": 0.0456,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 44500
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3146,
      "doc_norm": 1.419,
      "encoder_q-embeddings": 297.1875,
      "encoder_q-layer.0": 192.2595,
      "encoder_q-layer.1": 206.4339,
      "encoder_q-layer.10": 316.444,
      "encoder_q-layer.11": 659.6421,
      "encoder_q-layer.2": 227.7933,
      "encoder_q-layer.3": 241.623,
      "encoder_q-layer.4": 266.461,
      "encoder_q-layer.5": 277.622,
      "encoder_q-layer.6": 303.7605,
      "encoder_q-layer.7": 337.3526,
      "encoder_q-layer.8": 363.2707,
      "encoder_q-layer.9": 311.3642,
      "epoch": 0.44,
      "inbatch_neg_score": 0.2025,
      "inbatch_pos_score": 0.8647,
      "learning_rate": 3.077777777777778e-05,
      "loss": 3.3146,
      "norm_diff": 0.0509,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 487.4019,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2023,
      "query_norm": 1.3681,
      "queue_k_norm": 1.4168,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8687,
      "sent_len_1": 66.6894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.555,
      "stdk": 0.0486,
      "stdq": 0.0453,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 44600
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3036,
      "doc_norm": 1.4166,
      "encoder_q-embeddings": 374.2867,
      "encoder_q-layer.0": 253.1553,
      "encoder_q-layer.1": 272.7285,
      "encoder_q-layer.10": 285.3405,
      "encoder_q-layer.11": 637.5209,
      "encoder_q-layer.2": 325.6043,
      "encoder_q-layer.3": 357.7886,
      "encoder_q-layer.4": 351.4905,
      "encoder_q-layer.5": 338.3356,
      "encoder_q-layer.6": 387.7497,
      "encoder_q-layer.7": 430.6668,
      "encoder_q-layer.8": 378.2032,
      "encoder_q-layer.9": 285.5622,
      "epoch": 0.44,
      "inbatch_neg_score": 0.1996,
      "inbatch_pos_score": 0.8594,
      "learning_rate": 3.0722222222222227e-05,
      "loss": 3.3036,
      "norm_diff": 0.0399,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 555.4492,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1997,
      "query_norm": 1.3767,
      "queue_k_norm": 1.4171,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6456,
      "sent_len_1": 66.8564,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 193.5762,
      "stdk": 0.0485,
      "stdq": 0.0457,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 44700
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2875,
      "doc_norm": 1.4188,
      "encoder_q-embeddings": 438.8046,
      "encoder_q-layer.0": 339.4998,
      "encoder_q-layer.1": 393.7716,
      "encoder_q-layer.10": 330.1195,
      "encoder_q-layer.11": 685.0272,
      "encoder_q-layer.2": 448.0167,
      "encoder_q-layer.3": 438.2965,
      "encoder_q-layer.4": 422.6928,
      "encoder_q-layer.5": 379.284,
      "encoder_q-layer.6": 363.7701,
      "encoder_q-layer.7": 384.4494,
      "encoder_q-layer.8": 422.1562,
      "encoder_q-layer.9": 330.9393,
      "epoch": 0.44,
      "inbatch_neg_score": 0.1997,
      "inbatch_pos_score": 0.8525,
      "learning_rate": 3.066666666666667e-05,
      "loss": 3.2875,
      "norm_diff": 0.0507,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 630.8868,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.1993,
      "query_norm": 1.3681,
      "queue_k_norm": 1.4181,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9654,
      "sent_len_1": 66.8566,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1037,
      "stdk": 0.0486,
      "stdq": 0.045,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 44800
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3101,
      "doc_norm": 1.4072,
      "encoder_q-embeddings": 1008.6836,
      "encoder_q-layer.0": 698.0264,
      "encoder_q-layer.1": 773.3721,
      "encoder_q-layer.10": 298.5735,
      "encoder_q-layer.11": 637.9417,
      "encoder_q-layer.2": 996.0127,
      "encoder_q-layer.3": 1067.2345,
      "encoder_q-layer.4": 1207.7457,
      "encoder_q-layer.5": 1315.9204,
      "encoder_q-layer.6": 1355.9625,
      "encoder_q-layer.7": 1271.5652,
      "encoder_q-layer.8": 865.9797,
      "encoder_q-layer.9": 358.8655,
      "epoch": 0.44,
      "inbatch_neg_score": 0.206,
      "inbatch_pos_score": 0.8638,
      "learning_rate": 3.061111111111111e-05,
      "loss": 3.3101,
      "norm_diff": 0.017,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1433.2065,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2057,
      "query_norm": 1.3917,
      "queue_k_norm": 1.4186,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9545,
      "sent_len_1": 66.8065,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.575,
      "stdk": 0.0482,
      "stdq": 0.0457,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 44900
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3055,
      "doc_norm": 1.4168,
      "encoder_q-embeddings": 443.9218,
      "encoder_q-layer.0": 319.0468,
      "encoder_q-layer.1": 314.9095,
      "encoder_q-layer.10": 282.2415,
      "encoder_q-layer.11": 593.1105,
      "encoder_q-layer.2": 361.2851,
      "encoder_q-layer.3": 363.5935,
      "encoder_q-layer.4": 364.8318,
      "encoder_q-layer.5": 395.7696,
      "encoder_q-layer.6": 402.6557,
      "encoder_q-layer.7": 408.2527,
      "encoder_q-layer.8": 371.869,
      "encoder_q-layer.9": 285.5411,
      "epoch": 0.44,
      "inbatch_neg_score": 0.2135,
      "inbatch_pos_score": 0.8984,
      "learning_rate": 3.055555555555556e-05,
      "loss": 3.3055,
      "norm_diff": 0.0141,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 576.6874,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2128,
      "query_norm": 1.4027,
      "queue_k_norm": 1.4179,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0397,
      "sent_len_1": 66.8432,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8825,
      "stdk": 0.0485,
      "stdq": 0.0456,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 45000
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3073,
      "doc_norm": 1.4158,
      "encoder_q-embeddings": 2151.3162,
      "encoder_q-layer.0": 1393.9089,
      "encoder_q-layer.1": 1658.1257,
      "encoder_q-layer.10": 574.8478,
      "encoder_q-layer.11": 1242.0784,
      "encoder_q-layer.2": 1899.3822,
      "encoder_q-layer.3": 1791.661,
      "encoder_q-layer.4": 1771.0856,
      "encoder_q-layer.5": 1488.2998,
      "encoder_q-layer.6": 1520.9319,
      "encoder_q-layer.7": 1455.3865,
      "encoder_q-layer.8": 1021.2946,
      "encoder_q-layer.9": 631.4149,
      "epoch": 0.44,
      "inbatch_neg_score": 0.2182,
      "inbatch_pos_score": 0.854,
      "learning_rate": 3.05e-05,
      "loss": 3.3073,
      "norm_diff": 0.0199,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2294.5212,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2181,
      "query_norm": 1.3959,
      "queue_k_norm": 1.4168,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0006,
      "sent_len_1": 67.0456,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0387,
      "stdk": 0.0485,
      "stdq": 0.0452,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45100
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3143,
      "doc_norm": 1.4155,
      "encoder_q-embeddings": 528.5461,
      "encoder_q-layer.0": 350.3493,
      "encoder_q-layer.1": 364.1764,
      "encoder_q-layer.10": 604.6742,
      "encoder_q-layer.11": 1253.2003,
      "encoder_q-layer.2": 395.5608,
      "encoder_q-layer.3": 401.8536,
      "encoder_q-layer.4": 432.2409,
      "encoder_q-layer.5": 459.5488,
      "encoder_q-layer.6": 541.5715,
      "encoder_q-layer.7": 592.0429,
      "encoder_q-layer.8": 714.8865,
      "encoder_q-layer.9": 612.745,
      "epoch": 0.44,
      "inbatch_neg_score": 0.2254,
      "inbatch_pos_score": 0.8999,
      "learning_rate": 3.044444444444445e-05,
      "loss": 3.3143,
      "norm_diff": 0.0089,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 901.3771,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2244,
      "query_norm": 1.4206,
      "queue_k_norm": 1.4169,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9405,
      "sent_len_1": 66.8898,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8625,
      "stdk": 0.0484,
      "stdq": 0.0461,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45200
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2908,
      "doc_norm": 1.4138,
      "encoder_q-embeddings": 707.9913,
      "encoder_q-layer.0": 499.0506,
      "encoder_q-layer.1": 552.5721,
      "encoder_q-layer.10": 602.9816,
      "encoder_q-layer.11": 1329.1221,
      "encoder_q-layer.2": 611.2488,
      "encoder_q-layer.3": 613.9766,
      "encoder_q-layer.4": 639.7579,
      "encoder_q-layer.5": 608.3501,
      "encoder_q-layer.6": 630.3896,
      "encoder_q-layer.7": 665.3107,
      "encoder_q-layer.8": 723.4641,
      "encoder_q-layer.9": 598.5833,
      "epoch": 0.44,
      "inbatch_neg_score": 0.2315,
      "inbatch_pos_score": 0.9136,
      "learning_rate": 3.0388888888888887e-05,
      "loss": 3.2908,
      "norm_diff": 0.0171,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1051.7396,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2308,
      "query_norm": 1.4299,
      "queue_k_norm": 1.4187,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8462,
      "sent_len_1": 66.7798,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5175,
      "stdk": 0.0484,
      "stdq": 0.0464,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45300
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3302,
      "doc_norm": 1.4223,
      "encoder_q-embeddings": 572.1978,
      "encoder_q-layer.0": 388.9594,
      "encoder_q-layer.1": 427.3064,
      "encoder_q-layer.10": 616.3926,
      "encoder_q-layer.11": 1361.8383,
      "encoder_q-layer.2": 476.388,
      "encoder_q-layer.3": 501.2525,
      "encoder_q-layer.4": 529.0745,
      "encoder_q-layer.5": 549.6539,
      "encoder_q-layer.6": 562.7726,
      "encoder_q-layer.7": 627.0963,
      "encoder_q-layer.8": 712.4395,
      "encoder_q-layer.9": 608.4832,
      "epoch": 0.44,
      "inbatch_neg_score": 0.236,
      "inbatch_pos_score": 0.9326,
      "learning_rate": 3.0333333333333337e-05,
      "loss": 3.3302,
      "norm_diff": 0.0073,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 993.3553,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.236,
      "query_norm": 1.4274,
      "queue_k_norm": 1.4195,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0496,
      "sent_len_1": 66.9502,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5012,
      "stdk": 0.0486,
      "stdq": 0.0463,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45400
    },
    {
      "accuracy": 58.0078,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3118,
      "doc_norm": 1.4234,
      "encoder_q-embeddings": 768.6887,
      "encoder_q-layer.0": 507.2254,
      "encoder_q-layer.1": 546.4923,
      "encoder_q-layer.10": 567.2425,
      "encoder_q-layer.11": 1220.38,
      "encoder_q-layer.2": 616.4326,
      "encoder_q-layer.3": 684.917,
      "encoder_q-layer.4": 667.4772,
      "encoder_q-layer.5": 690.4581,
      "encoder_q-layer.6": 761.5431,
      "encoder_q-layer.7": 821.6723,
      "encoder_q-layer.8": 835.9122,
      "encoder_q-layer.9": 613.6291,
      "epoch": 0.44,
      "inbatch_neg_score": 0.2373,
      "inbatch_pos_score": 0.9258,
      "learning_rate": 3.0277777777777776e-05,
      "loss": 3.3118,
      "norm_diff": 0.0076,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1103.7136,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2378,
      "query_norm": 1.4186,
      "queue_k_norm": 1.4203,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0134,
      "sent_len_1": 66.756,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2587,
      "stdk": 0.0487,
      "stdq": 0.0462,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45500
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3033,
      "doc_norm": 1.4213,
      "encoder_q-embeddings": 704.0043,
      "encoder_q-layer.0": 485.1963,
      "encoder_q-layer.1": 500.5159,
      "encoder_q-layer.10": 615.786,
      "encoder_q-layer.11": 1365.4988,
      "encoder_q-layer.2": 553.1611,
      "encoder_q-layer.3": 564.9895,
      "encoder_q-layer.4": 589.6924,
      "encoder_q-layer.5": 598.3147,
      "encoder_q-layer.6": 618.9741,
      "encoder_q-layer.7": 719.8652,
      "encoder_q-layer.8": 725.1595,
      "encoder_q-layer.9": 615.0979,
      "epoch": 0.45,
      "inbatch_neg_score": 0.234,
      "inbatch_pos_score": 0.8779,
      "learning_rate": 3.0222222222222225e-05,
      "loss": 3.3033,
      "norm_diff": 0.0464,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1046.2606,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2334,
      "query_norm": 1.3749,
      "queue_k_norm": 1.4214,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8107,
      "sent_len_1": 66.8491,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.72,
      "stdk": 0.0485,
      "stdq": 0.045,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 45600
    },
    {
      "accuracy": 50.1953,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3141,
      "doc_norm": 1.4181,
      "encoder_q-embeddings": 823.1373,
      "encoder_q-layer.0": 601.6436,
      "encoder_q-layer.1": 659.7247,
      "encoder_q-layer.10": 609.1884,
      "encoder_q-layer.11": 1463.3352,
      "encoder_q-layer.2": 758.6068,
      "encoder_q-layer.3": 821.4514,
      "encoder_q-layer.4": 866.547,
      "encoder_q-layer.5": 928.1512,
      "encoder_q-layer.6": 921.5577,
      "encoder_q-layer.7": 1017.5889,
      "encoder_q-layer.8": 987.1755,
      "encoder_q-layer.9": 698.1871,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2351,
      "inbatch_pos_score": 0.856,
      "learning_rate": 3.016666666666667e-05,
      "loss": 3.3141,
      "norm_diff": 0.0419,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1329.5789,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2346,
      "query_norm": 1.3762,
      "queue_k_norm": 1.4208,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6507,
      "sent_len_1": 66.7432,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8425,
      "stdk": 0.0484,
      "stdq": 0.0452,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45700
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3072,
      "doc_norm": 1.4251,
      "encoder_q-embeddings": 496.9846,
      "encoder_q-layer.0": 329.057,
      "encoder_q-layer.1": 344.3408,
      "encoder_q-layer.10": 576.1889,
      "encoder_q-layer.11": 1314.952,
      "encoder_q-layer.2": 376.309,
      "encoder_q-layer.3": 403.5558,
      "encoder_q-layer.4": 427.9603,
      "encoder_q-layer.5": 451.2594,
      "encoder_q-layer.6": 531.1075,
      "encoder_q-layer.7": 644.2263,
      "encoder_q-layer.8": 691.0531,
      "encoder_q-layer.9": 608.9666,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2378,
      "inbatch_pos_score": 0.9053,
      "learning_rate": 3.0111111111111113e-05,
      "loss": 3.3072,
      "norm_diff": 0.0429,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 907.2869,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2373,
      "query_norm": 1.3823,
      "queue_k_norm": 1.4227,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9742,
      "sent_len_1": 66.8795,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9725,
      "stdk": 0.0486,
      "stdq": 0.0454,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45800
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2884,
      "doc_norm": 1.4243,
      "encoder_q-embeddings": 722.6323,
      "encoder_q-layer.0": 496.2588,
      "encoder_q-layer.1": 551.7868,
      "encoder_q-layer.10": 634.8545,
      "encoder_q-layer.11": 1325.9576,
      "encoder_q-layer.2": 691.0233,
      "encoder_q-layer.3": 764.1688,
      "encoder_q-layer.4": 883.7959,
      "encoder_q-layer.5": 923.074,
      "encoder_q-layer.6": 971.5112,
      "encoder_q-layer.7": 841.146,
      "encoder_q-layer.8": 850.5674,
      "encoder_q-layer.9": 645.7567,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2358,
      "inbatch_pos_score": 0.8779,
      "learning_rate": 3.005555555555556e-05,
      "loss": 3.2884,
      "norm_diff": 0.0347,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1214.4133,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.236,
      "query_norm": 1.3896,
      "queue_k_norm": 1.422,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.882,
      "sent_len_1": 66.7835,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7038,
      "stdk": 0.0486,
      "stdq": 0.0458,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 45900
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3149,
      "doc_norm": 1.4218,
      "encoder_q-embeddings": 2466.2339,
      "encoder_q-layer.0": 1746.1521,
      "encoder_q-layer.1": 1474.5081,
      "encoder_q-layer.10": 600.1024,
      "encoder_q-layer.11": 1295.2843,
      "encoder_q-layer.2": 1651.5361,
      "encoder_q-layer.3": 1723.9873,
      "encoder_q-layer.4": 1759.2365,
      "encoder_q-layer.5": 1663.9301,
      "encoder_q-layer.6": 1569.5308,
      "encoder_q-layer.7": 1660.8876,
      "encoder_q-layer.8": 1353.2133,
      "encoder_q-layer.9": 771.9584,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2331,
      "inbatch_pos_score": 0.8926,
      "learning_rate": 3e-05,
      "loss": 3.3149,
      "norm_diff": 0.0265,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2519.3038,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2336,
      "query_norm": 1.3952,
      "queue_k_norm": 1.4261,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7013,
      "sent_len_1": 66.5543,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1738,
      "stdk": 0.0485,
      "stdq": 0.0462,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46000
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3141,
      "doc_norm": 1.4199,
      "encoder_q-embeddings": 538.4189,
      "encoder_q-layer.0": 358.2054,
      "encoder_q-layer.1": 376.3449,
      "encoder_q-layer.10": 622.5236,
      "encoder_q-layer.11": 1262.4795,
      "encoder_q-layer.2": 408.5656,
      "encoder_q-layer.3": 420.966,
      "encoder_q-layer.4": 470.7821,
      "encoder_q-layer.5": 485.1684,
      "encoder_q-layer.6": 525.8845,
      "encoder_q-layer.7": 608.8171,
      "encoder_q-layer.8": 715.424,
      "encoder_q-layer.9": 633.2883,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2252,
      "inbatch_pos_score": 0.8809,
      "learning_rate": 2.9944444444444446e-05,
      "loss": 3.3141,
      "norm_diff": 0.0464,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 918.3071,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2263,
      "query_norm": 1.3735,
      "queue_k_norm": 1.4249,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.833,
      "sent_len_1": 66.5855,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4863,
      "stdk": 0.0484,
      "stdq": 0.0455,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46100
    },
    {
      "accuracy": 52.9297,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2864,
      "doc_norm": 1.4326,
      "encoder_q-embeddings": 662.6136,
      "encoder_q-layer.0": 410.8615,
      "encoder_q-layer.1": 421.296,
      "encoder_q-layer.10": 643.6664,
      "encoder_q-layer.11": 1389.5165,
      "encoder_q-layer.2": 472.1631,
      "encoder_q-layer.3": 502.4929,
      "encoder_q-layer.4": 527.8853,
      "encoder_q-layer.5": 534.4405,
      "encoder_q-layer.6": 588.6668,
      "encoder_q-layer.7": 681.8404,
      "encoder_q-layer.8": 741.5972,
      "encoder_q-layer.9": 629.8555,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2299,
      "inbatch_pos_score": 0.8989,
      "learning_rate": 2.988888888888889e-05,
      "loss": 3.2864,
      "norm_diff": 0.0502,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1026.6205,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2303,
      "query_norm": 1.3824,
      "queue_k_norm": 1.4242,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8486,
      "sent_len_1": 66.4723,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.5737,
      "stdk": 0.0489,
      "stdq": 0.0457,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 46200
    },
    {
      "accuracy": 51.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2911,
      "doc_norm": 1.4249,
      "encoder_q-embeddings": 1285.2937,
      "encoder_q-layer.0": 940.409,
      "encoder_q-layer.1": 1091.1738,
      "encoder_q-layer.10": 619.2592,
      "encoder_q-layer.11": 1437.351,
      "encoder_q-layer.2": 1288.751,
      "encoder_q-layer.3": 1338.6329,
      "encoder_q-layer.4": 1411.8809,
      "encoder_q-layer.5": 1377.1304,
      "encoder_q-layer.6": 1281.236,
      "encoder_q-layer.7": 1096.6338,
      "encoder_q-layer.8": 997.6033,
      "encoder_q-layer.9": 666.907,
      "epoch": 0.45,
      "inbatch_neg_score": 0.229,
      "inbatch_pos_score": 0.8535,
      "learning_rate": 2.9833333333333335e-05,
      "loss": 3.2911,
      "norm_diff": 0.0721,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1744.4094,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2273,
      "query_norm": 1.3528,
      "queue_k_norm": 1.4249,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9992,
      "sent_len_1": 66.848,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9975,
      "stdk": 0.0485,
      "stdq": 0.0446,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46300
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2975,
      "doc_norm": 1.4247,
      "encoder_q-embeddings": 532.2657,
      "encoder_q-layer.0": 349.8211,
      "encoder_q-layer.1": 386.0331,
      "encoder_q-layer.10": 610.9373,
      "encoder_q-layer.11": 1254.9396,
      "encoder_q-layer.2": 431.7523,
      "encoder_q-layer.3": 447.0812,
      "encoder_q-layer.4": 467.7936,
      "encoder_q-layer.5": 493.9984,
      "encoder_q-layer.6": 543.5929,
      "encoder_q-layer.7": 630.6003,
      "encoder_q-layer.8": 675.6058,
      "encoder_q-layer.9": 593.9529,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2213,
      "inbatch_pos_score": 0.9121,
      "learning_rate": 2.9777777777777777e-05,
      "loss": 3.2975,
      "norm_diff": 0.0332,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 912.4508,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2208,
      "query_norm": 1.3915,
      "queue_k_norm": 1.4283,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9949,
      "sent_len_1": 66.8478,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3988,
      "stdk": 0.0485,
      "stdq": 0.0462,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 46400
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2937,
      "doc_norm": 1.4248,
      "encoder_q-embeddings": 516.7308,
      "encoder_q-layer.0": 336.9787,
      "encoder_q-layer.1": 341.7676,
      "encoder_q-layer.10": 543.8725,
      "encoder_q-layer.11": 1254.1589,
      "encoder_q-layer.2": 389.2842,
      "encoder_q-layer.3": 428.2521,
      "encoder_q-layer.4": 457.6279,
      "encoder_q-layer.5": 480.2539,
      "encoder_q-layer.6": 553.2525,
      "encoder_q-layer.7": 615.6038,
      "encoder_q-layer.8": 700.2859,
      "encoder_q-layer.9": 590.9703,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2192,
      "inbatch_pos_score": 0.8804,
      "learning_rate": 2.9722222222222223e-05,
      "loss": 3.2937,
      "norm_diff": 0.0605,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 901.372,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.22,
      "query_norm": 1.3644,
      "queue_k_norm": 1.4265,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6605,
      "sent_len_1": 66.5996,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2788,
      "stdk": 0.0486,
      "stdq": 0.0452,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46500
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3097,
      "doc_norm": 1.4273,
      "encoder_q-embeddings": 720.0128,
      "encoder_q-layer.0": 474.8589,
      "encoder_q-layer.1": 543.2939,
      "encoder_q-layer.10": 641.2438,
      "encoder_q-layer.11": 1376.1637,
      "encoder_q-layer.2": 631.0128,
      "encoder_q-layer.3": 660.0961,
      "encoder_q-layer.4": 753.6788,
      "encoder_q-layer.5": 821.9664,
      "encoder_q-layer.6": 889.2244,
      "encoder_q-layer.7": 887.1025,
      "encoder_q-layer.8": 993.5014,
      "encoder_q-layer.9": 768.4984,
      "epoch": 0.45,
      "inbatch_neg_score": 0.2199,
      "inbatch_pos_score": 0.8745,
      "learning_rate": 2.9666666666666672e-05,
      "loss": 3.3097,
      "norm_diff": 0.0529,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1207.3944,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.22,
      "query_norm": 1.3745,
      "queue_k_norm": 1.4263,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0609,
      "sent_len_1": 67.004,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.3325,
      "stdk": 0.0486,
      "stdq": 0.0455,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46600
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3014,
      "doc_norm": 1.4292,
      "encoder_q-embeddings": 1675.5349,
      "encoder_q-layer.0": 1184.3787,
      "encoder_q-layer.1": 1356.4994,
      "encoder_q-layer.10": 607.3243,
      "encoder_q-layer.11": 1314.6866,
      "encoder_q-layer.2": 1538.9652,
      "encoder_q-layer.3": 1724.3932,
      "encoder_q-layer.4": 1843.2448,
      "encoder_q-layer.5": 1823.2882,
      "encoder_q-layer.6": 1805.8909,
      "encoder_q-layer.7": 1722.225,
      "encoder_q-layer.8": 1196.9406,
      "encoder_q-layer.9": 691.4374,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2213,
      "inbatch_pos_score": 0.8687,
      "learning_rate": 2.961111111111111e-05,
      "loss": 3.3014,
      "norm_diff": 0.055,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2200.3962,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.22,
      "query_norm": 1.3742,
      "queue_k_norm": 1.4258,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7671,
      "sent_len_1": 66.9782,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1987,
      "stdk": 0.0487,
      "stdq": 0.0456,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46700
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.317,
      "doc_norm": 1.4279,
      "encoder_q-embeddings": 534.8832,
      "encoder_q-layer.0": 347.0648,
      "encoder_q-layer.1": 364.2167,
      "encoder_q-layer.10": 708.2047,
      "encoder_q-layer.11": 1416.8915,
      "encoder_q-layer.2": 424.5532,
      "encoder_q-layer.3": 449.3445,
      "encoder_q-layer.4": 483.2906,
      "encoder_q-layer.5": 486.7383,
      "encoder_q-layer.6": 551.3012,
      "encoder_q-layer.7": 635.5269,
      "encoder_q-layer.8": 718.485,
      "encoder_q-layer.9": 698.8461,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2153,
      "inbatch_pos_score": 0.8643,
      "learning_rate": 2.955555555555556e-05,
      "loss": 3.317,
      "norm_diff": 0.0388,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 964.8262,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2155,
      "query_norm": 1.3891,
      "queue_k_norm": 1.4253,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9584,
      "sent_len_1": 66.6899,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6463,
      "stdk": 0.0487,
      "stdq": 0.0462,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46800
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.27,
      "doc_norm": 1.4294,
      "encoder_q-embeddings": 514.0528,
      "encoder_q-layer.0": 340.7193,
      "encoder_q-layer.1": 344.8975,
      "encoder_q-layer.10": 598.6968,
      "encoder_q-layer.11": 1296.3027,
      "encoder_q-layer.2": 376.4624,
      "encoder_q-layer.3": 401.5954,
      "encoder_q-layer.4": 431.6165,
      "encoder_q-layer.5": 463.8053,
      "encoder_q-layer.6": 500.7421,
      "encoder_q-layer.7": 570.1464,
      "encoder_q-layer.8": 674.3207,
      "encoder_q-layer.9": 605.6843,
      "epoch": 0.46,
      "inbatch_neg_score": 0.221,
      "inbatch_pos_score": 0.9072,
      "learning_rate": 2.95e-05,
      "loss": 3.27,
      "norm_diff": 0.0265,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 893.4081,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2202,
      "query_norm": 1.4035,
      "queue_k_norm": 1.4254,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9031,
      "sent_len_1": 67.0401,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5662,
      "stdk": 0.0488,
      "stdq": 0.0465,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 46900
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3077,
      "doc_norm": 1.4238,
      "encoder_q-embeddings": 565.1782,
      "encoder_q-layer.0": 365.8199,
      "encoder_q-layer.1": 385.2644,
      "encoder_q-layer.10": 569.3797,
      "encoder_q-layer.11": 1321.4886,
      "encoder_q-layer.2": 424.0208,
      "encoder_q-layer.3": 454.5379,
      "encoder_q-layer.4": 494.619,
      "encoder_q-layer.5": 504.0933,
      "encoder_q-layer.6": 553.5225,
      "encoder_q-layer.7": 639.169,
      "encoder_q-layer.8": 696.531,
      "encoder_q-layer.9": 600.1984,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2213,
      "inbatch_pos_score": 0.9009,
      "learning_rate": 2.9444444444444448e-05,
      "loss": 3.3077,
      "norm_diff": 0.0425,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 952.8839,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2225,
      "query_norm": 1.3814,
      "queue_k_norm": 1.4251,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1263,
      "sent_len_1": 66.9611,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.475,
      "stdk": 0.0486,
      "stdq": 0.0456,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47000
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2852,
      "doc_norm": 1.426,
      "encoder_q-embeddings": 1006.1053,
      "encoder_q-layer.0": 654.6647,
      "encoder_q-layer.1": 677.7766,
      "encoder_q-layer.10": 1211.595,
      "encoder_q-layer.11": 2727.8237,
      "encoder_q-layer.2": 756.7513,
      "encoder_q-layer.3": 759.2128,
      "encoder_q-layer.4": 825.3746,
      "encoder_q-layer.5": 863.7272,
      "encoder_q-layer.6": 983.4937,
      "encoder_q-layer.7": 1243.5803,
      "encoder_q-layer.8": 1417.1707,
      "encoder_q-layer.9": 1266.5186,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2237,
      "inbatch_pos_score": 0.8916,
      "learning_rate": 2.9388888888888887e-05,
      "loss": 3.2852,
      "norm_diff": 0.039,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1828.7839,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.224,
      "query_norm": 1.387,
      "queue_k_norm": 1.4259,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8653,
      "sent_len_1": 66.6842,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1763,
      "stdk": 0.0487,
      "stdq": 0.0457,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47100
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3046,
      "doc_norm": 1.4251,
      "encoder_q-embeddings": 1916.7208,
      "encoder_q-layer.0": 1243.9668,
      "encoder_q-layer.1": 1467.0419,
      "encoder_q-layer.10": 1309.8733,
      "encoder_q-layer.11": 2745.0518,
      "encoder_q-layer.2": 1875.5293,
      "encoder_q-layer.3": 2231.8564,
      "encoder_q-layer.4": 2199.4424,
      "encoder_q-layer.5": 2329.6704,
      "encoder_q-layer.6": 2657.1677,
      "encoder_q-layer.7": 3054.7551,
      "encoder_q-layer.8": 2947.2256,
      "encoder_q-layer.9": 1735.4291,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2208,
      "inbatch_pos_score": 0.8535,
      "learning_rate": 2.9333333333333336e-05,
      "loss": 3.3046,
      "norm_diff": 0.0395,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3298.7609,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2209,
      "query_norm": 1.3856,
      "queue_k_norm": 1.4246,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0405,
      "sent_len_1": 66.8335,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.17,
      "stdk": 0.0486,
      "stdq": 0.0458,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47200
    },
    {
      "accuracy": 47.0703,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2956,
      "doc_norm": 1.421,
      "encoder_q-embeddings": 1280.3264,
      "encoder_q-layer.0": 900.5726,
      "encoder_q-layer.1": 916.1006,
      "encoder_q-layer.10": 1240.1953,
      "encoder_q-layer.11": 2969.8765,
      "encoder_q-layer.2": 1020.7275,
      "encoder_q-layer.3": 1076.554,
      "encoder_q-layer.4": 1158.0641,
      "encoder_q-layer.5": 1221.2217,
      "encoder_q-layer.6": 1314.1509,
      "encoder_q-layer.7": 1473.8191,
      "encoder_q-layer.8": 1543.2629,
      "encoder_q-layer.9": 1273.686,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2282,
      "inbatch_pos_score": 0.8076,
      "learning_rate": 2.927777777777778e-05,
      "loss": 3.2956,
      "norm_diff": 0.0735,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2199.7201,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2264,
      "query_norm": 1.3475,
      "queue_k_norm": 1.4253,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0511,
      "sent_len_1": 66.8818,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.37,
      "stdk": 0.0484,
      "stdq": 0.044,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47300
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2896,
      "doc_norm": 1.4256,
      "encoder_q-embeddings": 1336.8907,
      "encoder_q-layer.0": 893.0792,
      "encoder_q-layer.1": 971.411,
      "encoder_q-layer.10": 1185.0334,
      "encoder_q-layer.11": 2604.0833,
      "encoder_q-layer.2": 1104.0466,
      "encoder_q-layer.3": 1098.4115,
      "encoder_q-layer.4": 1078.5972,
      "encoder_q-layer.5": 1143.3488,
      "encoder_q-layer.6": 1240.9667,
      "encoder_q-layer.7": 1376.4022,
      "encoder_q-layer.8": 1451.6913,
      "encoder_q-layer.9": 1168.3917,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2244,
      "inbatch_pos_score": 0.8823,
      "learning_rate": 2.9222222222222224e-05,
      "loss": 3.2896,
      "norm_diff": 0.039,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2028.3423,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2249,
      "query_norm": 1.3866,
      "queue_k_norm": 1.4259,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8916,
      "sent_len_1": 66.5893,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8137,
      "stdk": 0.0486,
      "stdq": 0.0456,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47400
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3209,
      "doc_norm": 1.4286,
      "encoder_q-embeddings": 1032.4471,
      "encoder_q-layer.0": 674.157,
      "encoder_q-layer.1": 696.6463,
      "encoder_q-layer.10": 1135.1278,
      "encoder_q-layer.11": 2607.3789,
      "encoder_q-layer.2": 742.3719,
      "encoder_q-layer.3": 792.0936,
      "encoder_q-layer.4": 826.8855,
      "encoder_q-layer.5": 825.0183,
      "encoder_q-layer.6": 965.8436,
      "encoder_q-layer.7": 1138.9368,
      "encoder_q-layer.8": 1372.0851,
      "encoder_q-layer.9": 1207.9176,
      "epoch": 0.46,
      "inbatch_neg_score": 0.2306,
      "inbatch_pos_score": 0.9189,
      "learning_rate": 2.916666666666667e-05,
      "loss": 3.3209,
      "norm_diff": 0.024,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1800.0288,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2302,
      "query_norm": 1.4046,
      "queue_k_norm": 1.4271,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8666,
      "sent_len_1": 66.782,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2713,
      "stdk": 0.0487,
      "stdq": 0.0462,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47500
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2891,
      "doc_norm": 1.4248,
      "encoder_q-embeddings": 1026.5413,
      "encoder_q-layer.0": 668.6854,
      "encoder_q-layer.1": 691.8762,
      "encoder_q-layer.10": 1150.1545,
      "encoder_q-layer.11": 2657.7563,
      "encoder_q-layer.2": 750.4769,
      "encoder_q-layer.3": 786.4589,
      "encoder_q-layer.4": 821.9103,
      "encoder_q-layer.5": 917.7315,
      "encoder_q-layer.6": 1045.2136,
      "encoder_q-layer.7": 1177.535,
      "encoder_q-layer.8": 1365.605,
      "encoder_q-layer.9": 1187.5286,
      "epoch": 0.46,
      "inbatch_neg_score": 0.225,
      "inbatch_pos_score": 0.8789,
      "learning_rate": 2.9111111111111112e-05,
      "loss": 3.2891,
      "norm_diff": 0.0509,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1834.4356,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2247,
      "query_norm": 1.3738,
      "queue_k_norm": 1.4243,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.609,
      "sent_len_1": 66.4559,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7025,
      "stdk": 0.0486,
      "stdq": 0.0451,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 47600
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3104,
      "doc_norm": 1.4303,
      "encoder_q-embeddings": 1037.6136,
      "encoder_q-layer.0": 671.5725,
      "encoder_q-layer.1": 713.489,
      "encoder_q-layer.10": 1201.4991,
      "encoder_q-layer.11": 2634.6887,
      "encoder_q-layer.2": 804.9861,
      "encoder_q-layer.3": 839.2104,
      "encoder_q-layer.4": 885.7121,
      "encoder_q-layer.5": 909.1273,
      "encoder_q-layer.6": 995.6784,
      "encoder_q-layer.7": 1189.5145,
      "encoder_q-layer.8": 1399.7854,
      "encoder_q-layer.9": 1272.5129,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2293,
      "inbatch_pos_score": 0.8799,
      "learning_rate": 2.9055555555555558e-05,
      "loss": 3.3104,
      "norm_diff": 0.049,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1827.5978,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2302,
      "query_norm": 1.3813,
      "queue_k_norm": 1.4249,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8479,
      "sent_len_1": 66.7309,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4812,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47700
    },
    {
      "accuracy": 49.6094,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3088,
      "doc_norm": 1.4273,
      "encoder_q-embeddings": 1308.8805,
      "encoder_q-layer.0": 831.4935,
      "encoder_q-layer.1": 808.0121,
      "encoder_q-layer.10": 1265.2429,
      "encoder_q-layer.11": 2889.7161,
      "encoder_q-layer.2": 862.2923,
      "encoder_q-layer.3": 920.5638,
      "encoder_q-layer.4": 977.1323,
      "encoder_q-layer.5": 1002.98,
      "encoder_q-layer.6": 1114.0908,
      "encoder_q-layer.7": 1260.7939,
      "encoder_q-layer.8": 1482.0543,
      "encoder_q-layer.9": 1328.9124,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2276,
      "inbatch_pos_score": 0.8579,
      "learning_rate": 2.9e-05,
      "loss": 3.3088,
      "norm_diff": 0.0424,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2045.8082,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2273,
      "query_norm": 1.3849,
      "queue_k_norm": 1.4257,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8838,
      "sent_len_1": 66.7109,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4888,
      "stdk": 0.0487,
      "stdq": 0.0453,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47800
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2887,
      "doc_norm": 1.4283,
      "encoder_q-embeddings": 1398.6035,
      "encoder_q-layer.0": 964.3931,
      "encoder_q-layer.1": 1135.7676,
      "encoder_q-layer.10": 590.2095,
      "encoder_q-layer.11": 1305.5054,
      "encoder_q-layer.2": 1240.3499,
      "encoder_q-layer.3": 1293.2319,
      "encoder_q-layer.4": 1400.6847,
      "encoder_q-layer.5": 1555.4417,
      "encoder_q-layer.6": 1570.9894,
      "encoder_q-layer.7": 1513.9745,
      "encoder_q-layer.8": 1018.3575,
      "encoder_q-layer.9": 645.3621,
      "epoch": 0.47,
      "inbatch_neg_score": 0.235,
      "inbatch_pos_score": 0.9062,
      "learning_rate": 2.8944444444444446e-05,
      "loss": 3.2887,
      "norm_diff": 0.022,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1848.1127,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.234,
      "query_norm": 1.4063,
      "queue_k_norm": 1.4259,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0827,
      "sent_len_1": 66.8307,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.3938,
      "stdk": 0.0487,
      "stdq": 0.0461,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 47900
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2759,
      "doc_norm": 1.4328,
      "encoder_q-embeddings": 892.9286,
      "encoder_q-layer.0": 587.9846,
      "encoder_q-layer.1": 670.3327,
      "encoder_q-layer.10": 591.7316,
      "encoder_q-layer.11": 1246.304,
      "encoder_q-layer.2": 788.8891,
      "encoder_q-layer.3": 830.2289,
      "encoder_q-layer.4": 916.0156,
      "encoder_q-layer.5": 1014.6574,
      "encoder_q-layer.6": 1033.6362,
      "encoder_q-layer.7": 1132.2083,
      "encoder_q-layer.8": 1025.2064,
      "encoder_q-layer.9": 697.8023,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2378,
      "inbatch_pos_score": 0.9072,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 3.2759,
      "norm_diff": 0.0481,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1349.5552,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2382,
      "query_norm": 1.3846,
      "queue_k_norm": 1.4257,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.952,
      "sent_len_1": 66.7182,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7625,
      "stdk": 0.0488,
      "stdq": 0.0451,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 48000
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2723,
      "doc_norm": 1.422,
      "encoder_q-embeddings": 1356.514,
      "encoder_q-layer.0": 911.4519,
      "encoder_q-layer.1": 992.5809,
      "encoder_q-layer.10": 607.5475,
      "encoder_q-layer.11": 1265.5281,
      "encoder_q-layer.2": 1088.2566,
      "encoder_q-layer.3": 1115.301,
      "encoder_q-layer.4": 1084.0438,
      "encoder_q-layer.5": 1178.6055,
      "encoder_q-layer.6": 1029.9767,
      "encoder_q-layer.7": 1099.9539,
      "encoder_q-layer.8": 924.0941,
      "encoder_q-layer.9": 620.5746,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2427,
      "inbatch_pos_score": 0.9058,
      "learning_rate": 2.8833333333333334e-05,
      "loss": 3.2723,
      "norm_diff": 0.0231,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1596.2087,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2428,
      "query_norm": 1.3989,
      "queue_k_norm": 1.4258,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0153,
      "sent_len_1": 66.6307,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6987,
      "stdk": 0.0484,
      "stdq": 0.0456,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 48100
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2747,
      "doc_norm": 1.4288,
      "encoder_q-embeddings": 546.8296,
      "encoder_q-layer.0": 348.5963,
      "encoder_q-layer.1": 371.2564,
      "encoder_q-layer.10": 648.1476,
      "encoder_q-layer.11": 1380.618,
      "encoder_q-layer.2": 409.53,
      "encoder_q-layer.3": 418.1068,
      "encoder_q-layer.4": 457.9181,
      "encoder_q-layer.5": 462.685,
      "encoder_q-layer.6": 575.0729,
      "encoder_q-layer.7": 648.1972,
      "encoder_q-layer.8": 757.8553,
      "encoder_q-layer.9": 666.8765,
      "epoch": 0.47,
      "inbatch_neg_score": 0.246,
      "inbatch_pos_score": 0.9092,
      "learning_rate": 2.877777777777778e-05,
      "loss": 3.2747,
      "norm_diff": 0.0126,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 961.0177,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2468,
      "query_norm": 1.4272,
      "queue_k_norm": 1.4248,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0243,
      "sent_len_1": 66.7676,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.44,
      "stdk": 0.0487,
      "stdq": 0.0467,
      "stdqueue_k": 0.0486,
      "stdqueue_q": 0.0,
      "step": 48200
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3041,
      "doc_norm": 1.4258,
      "encoder_q-embeddings": 569.5185,
      "encoder_q-layer.0": 367.6384,
      "encoder_q-layer.1": 399.3294,
      "encoder_q-layer.10": 602.2183,
      "encoder_q-layer.11": 1380.363,
      "encoder_q-layer.2": 464.5112,
      "encoder_q-layer.3": 481.264,
      "encoder_q-layer.4": 539.3295,
      "encoder_q-layer.5": 545.8685,
      "encoder_q-layer.6": 590.405,
      "encoder_q-layer.7": 642.0701,
      "encoder_q-layer.8": 706.6587,
      "encoder_q-layer.9": 606.5648,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2503,
      "inbatch_pos_score": 0.8936,
      "learning_rate": 2.8722222222222222e-05,
      "loss": 3.3041,
      "norm_diff": 0.0388,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 992.7243,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2498,
      "query_norm": 1.387,
      "queue_k_norm": 1.4285,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9645,
      "sent_len_1": 66.9948,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9425,
      "stdk": 0.0485,
      "stdq": 0.0453,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 48300
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2632,
      "doc_norm": 1.4301,
      "encoder_q-embeddings": 553.0736,
      "encoder_q-layer.0": 362.5482,
      "encoder_q-layer.1": 390.148,
      "encoder_q-layer.10": 591.7563,
      "encoder_q-layer.11": 1281.4136,
      "encoder_q-layer.2": 434.1026,
      "encoder_q-layer.3": 452.7967,
      "encoder_q-layer.4": 509.3332,
      "encoder_q-layer.5": 526.9597,
      "encoder_q-layer.6": 550.8844,
      "encoder_q-layer.7": 618.018,
      "encoder_q-layer.8": 748.4677,
      "encoder_q-layer.9": 618.9684,
      "epoch": 0.47,
      "inbatch_neg_score": 0.253,
      "inbatch_pos_score": 0.9487,
      "learning_rate": 2.8666666666666668e-05,
      "loss": 3.2632,
      "norm_diff": 0.0103,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 938.1393,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2524,
      "query_norm": 1.425,
      "queue_k_norm": 1.4307,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8994,
      "sent_len_1": 66.9176,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2388,
      "stdk": 0.0487,
      "stdq": 0.0468,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 48400
    },
    {
      "accuracy": 58.8867,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2852,
      "doc_norm": 1.4306,
      "encoder_q-embeddings": 1498.9424,
      "encoder_q-layer.0": 1012.7009,
      "encoder_q-layer.1": 1190.5234,
      "encoder_q-layer.10": 580.6085,
      "encoder_q-layer.11": 1252.8499,
      "encoder_q-layer.2": 1180.6609,
      "encoder_q-layer.3": 1350.1953,
      "encoder_q-layer.4": 1283.1575,
      "encoder_q-layer.5": 1223.6936,
      "encoder_q-layer.6": 1275.7114,
      "encoder_q-layer.7": 1182.236,
      "encoder_q-layer.8": 858.3002,
      "encoder_q-layer.9": 649.9968,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2503,
      "inbatch_pos_score": 0.9561,
      "learning_rate": 2.861111111111111e-05,
      "loss": 3.2852,
      "norm_diff": 0.0312,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1744.4753,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2524,
      "query_norm": 1.3994,
      "queue_k_norm": 1.4308,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8606,
      "sent_len_1": 66.6842,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9125,
      "stdk": 0.0486,
      "stdq": 0.046,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 48500
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3052,
      "doc_norm": 1.4311,
      "encoder_q-embeddings": 549.3686,
      "encoder_q-layer.0": 350.7133,
      "encoder_q-layer.1": 354.2456,
      "encoder_q-layer.10": 570.1596,
      "encoder_q-layer.11": 1347.975,
      "encoder_q-layer.2": 396.1655,
      "encoder_q-layer.3": 417.18,
      "encoder_q-layer.4": 448.7313,
      "encoder_q-layer.5": 493.0651,
      "encoder_q-layer.6": 529.5985,
      "encoder_q-layer.7": 627.5674,
      "encoder_q-layer.8": 692.7375,
      "encoder_q-layer.9": 601.0941,
      "epoch": 0.47,
      "inbatch_neg_score": 0.2562,
      "inbatch_pos_score": 0.9067,
      "learning_rate": 2.855555555555556e-05,
      "loss": 3.3052,
      "norm_diff": 0.0568,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 936.9608,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2571,
      "query_norm": 1.3744,
      "queue_k_norm": 1.4305,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9522,
      "sent_len_1": 66.7514,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.025,
      "stdk": 0.0487,
      "stdq": 0.045,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 48600
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2944,
      "doc_norm": 1.4319,
      "encoder_q-embeddings": 494.7448,
      "encoder_q-layer.0": 325.32,
      "encoder_q-layer.1": 343.7641,
      "encoder_q-layer.10": 554.3524,
      "encoder_q-layer.11": 1255.1483,
      "encoder_q-layer.2": 382.2431,
      "encoder_q-layer.3": 395.2921,
      "encoder_q-layer.4": 430.5898,
      "encoder_q-layer.5": 447.0279,
      "encoder_q-layer.6": 493.0802,
      "encoder_q-layer.7": 569.2576,
      "encoder_q-layer.8": 682.7518,
      "encoder_q-layer.9": 588.1035,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2516,
      "inbatch_pos_score": 0.9634,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 3.2944,
      "norm_diff": 0.0282,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 884.4212,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2522,
      "query_norm": 1.4063,
      "queue_k_norm": 1.4314,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8827,
      "sent_len_1": 66.6237,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5225,
      "stdk": 0.0487,
      "stdq": 0.0464,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 48700
    },
    {
      "accuracy": 58.0078,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2453,
      "doc_norm": 1.4333,
      "encoder_q-embeddings": 712.2699,
      "encoder_q-layer.0": 491.7051,
      "encoder_q-layer.1": 536.4409,
      "encoder_q-layer.10": 558.1549,
      "encoder_q-layer.11": 1236.0571,
      "encoder_q-layer.2": 592.5216,
      "encoder_q-layer.3": 636.8582,
      "encoder_q-layer.4": 646.5758,
      "encoder_q-layer.5": 689.6365,
      "encoder_q-layer.6": 668.5109,
      "encoder_q-layer.7": 793.3224,
      "encoder_q-layer.8": 782.2405,
      "encoder_q-layer.9": 609.9998,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2516,
      "inbatch_pos_score": 0.9639,
      "learning_rate": 2.8444444444444447e-05,
      "loss": 3.2453,
      "norm_diff": 0.0228,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1071.5103,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2527,
      "query_norm": 1.4104,
      "queue_k_norm": 1.4314,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9028,
      "sent_len_1": 66.7626,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9863,
      "stdk": 0.0487,
      "stdq": 0.0465,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 48800
    },
    {
      "accuracy": 50.6836,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2805,
      "doc_norm": 1.4355,
      "encoder_q-embeddings": 565.7724,
      "encoder_q-layer.0": 368.4667,
      "encoder_q-layer.1": 383.0261,
      "encoder_q-layer.10": 591.6637,
      "encoder_q-layer.11": 1419.0225,
      "encoder_q-layer.2": 436.8559,
      "encoder_q-layer.3": 458.7169,
      "encoder_q-layer.4": 478.0298,
      "encoder_q-layer.5": 507.2666,
      "encoder_q-layer.6": 547.9071,
      "encoder_q-layer.7": 638.4215,
      "encoder_q-layer.8": 719.4729,
      "encoder_q-layer.9": 616.3262,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2566,
      "inbatch_pos_score": 0.8892,
      "learning_rate": 2.8388888888888893e-05,
      "loss": 3.2805,
      "norm_diff": 0.0637,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 976.9166,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2559,
      "query_norm": 1.3718,
      "queue_k_norm": 1.4336,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0095,
      "sent_len_1": 66.8235,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5938,
      "stdk": 0.0487,
      "stdq": 0.0449,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 48900
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3087,
      "doc_norm": 1.4296,
      "encoder_q-embeddings": 573.9216,
      "encoder_q-layer.0": 380.9545,
      "encoder_q-layer.1": 411.9406,
      "encoder_q-layer.10": 645.8431,
      "encoder_q-layer.11": 1409.3793,
      "encoder_q-layer.2": 462.9188,
      "encoder_q-layer.3": 469.1431,
      "encoder_q-layer.4": 524.4514,
      "encoder_q-layer.5": 516.006,
      "encoder_q-layer.6": 583.3431,
      "encoder_q-layer.7": 673.5248,
      "encoder_q-layer.8": 744.4202,
      "encoder_q-layer.9": 648.5178,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2572,
      "inbatch_pos_score": 0.8984,
      "learning_rate": 2.8333333333333335e-05,
      "loss": 3.3087,
      "norm_diff": 0.0378,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1018.6196,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2578,
      "query_norm": 1.3918,
      "queue_k_norm": 1.4336,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8964,
      "sent_len_1": 66.7379,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2988,
      "stdk": 0.0485,
      "stdq": 0.0456,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 49000
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.254,
      "doc_norm": 1.4307,
      "encoder_q-embeddings": 480.1978,
      "encoder_q-layer.0": 312.885,
      "encoder_q-layer.1": 328.4211,
      "encoder_q-layer.10": 595.6025,
      "encoder_q-layer.11": 1179.4508,
      "encoder_q-layer.2": 366.0674,
      "encoder_q-layer.3": 403.5346,
      "encoder_q-layer.4": 415.4842,
      "encoder_q-layer.5": 442.2807,
      "encoder_q-layer.6": 505.0235,
      "encoder_q-layer.7": 548.2421,
      "encoder_q-layer.8": 637.4025,
      "encoder_q-layer.9": 561.2772,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2529,
      "inbatch_pos_score": 0.9331,
      "learning_rate": 2.827777777777778e-05,
      "loss": 3.254,
      "norm_diff": 0.036,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 827.0843,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2522,
      "query_norm": 1.3947,
      "queue_k_norm": 1.4348,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7881,
      "sent_len_1": 66.9515,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9712,
      "stdk": 0.0485,
      "stdq": 0.0461,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 49100
    },
    {
      "accuracy": 60.1562,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2674,
      "doc_norm": 1.4362,
      "encoder_q-embeddings": 669.8634,
      "encoder_q-layer.0": 451.548,
      "encoder_q-layer.1": 461.2002,
      "encoder_q-layer.10": 556.2665,
      "encoder_q-layer.11": 1195.0251,
      "encoder_q-layer.2": 538.9232,
      "encoder_q-layer.3": 537.0554,
      "encoder_q-layer.4": 602.7024,
      "encoder_q-layer.5": 586.8381,
      "encoder_q-layer.6": 648.6224,
      "encoder_q-layer.7": 693.9544,
      "encoder_q-layer.8": 744.7681,
      "encoder_q-layer.9": 602.6758,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2496,
      "inbatch_pos_score": 0.9878,
      "learning_rate": 2.8222222222222223e-05,
      "loss": 3.2674,
      "norm_diff": 0.0189,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 996.8091,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2498,
      "query_norm": 1.4181,
      "queue_k_norm": 1.4342,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8133,
      "sent_len_1": 66.7709,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6025,
      "stdk": 0.0487,
      "stdq": 0.0469,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 49200
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2776,
      "doc_norm": 1.438,
      "encoder_q-embeddings": 743.2507,
      "encoder_q-layer.0": 499.1469,
      "encoder_q-layer.1": 528.4717,
      "encoder_q-layer.10": 575.9436,
      "encoder_q-layer.11": 1237.1012,
      "encoder_q-layer.2": 549.3284,
      "encoder_q-layer.3": 541.8404,
      "encoder_q-layer.4": 558.2244,
      "encoder_q-layer.5": 556.8466,
      "encoder_q-layer.6": 609.2246,
      "encoder_q-layer.7": 688.6274,
      "encoder_q-layer.8": 798.4581,
      "encoder_q-layer.9": 639.2142,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2491,
      "inbatch_pos_score": 0.9307,
      "learning_rate": 2.816666666666667e-05,
      "loss": 3.2776,
      "norm_diff": 0.0384,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1044.7238,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2487,
      "query_norm": 1.3996,
      "queue_k_norm": 1.4361,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.146,
      "sent_len_1": 66.8046,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1413,
      "stdk": 0.0488,
      "stdq": 0.0463,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 49300
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2815,
      "doc_norm": 1.43,
      "encoder_q-embeddings": 741.7136,
      "encoder_q-layer.0": 479.366,
      "encoder_q-layer.1": 568.0845,
      "encoder_q-layer.10": 599.6125,
      "encoder_q-layer.11": 1266.66,
      "encoder_q-layer.2": 649.4241,
      "encoder_q-layer.3": 808.4289,
      "encoder_q-layer.4": 824.9398,
      "encoder_q-layer.5": 843.7504,
      "encoder_q-layer.6": 936.3667,
      "encoder_q-layer.7": 995.6174,
      "encoder_q-layer.8": 887.8491,
      "encoder_q-layer.9": 609.2858,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2515,
      "inbatch_pos_score": 0.8994,
      "learning_rate": 2.811111111111111e-05,
      "loss": 3.2815,
      "norm_diff": 0.0632,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1209.0235,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2522,
      "query_norm": 1.3668,
      "queue_k_norm": 1.4353,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8605,
      "sent_len_1": 66.7924,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.065,
      "stdk": 0.0485,
      "stdq": 0.045,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 49400
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.261,
      "doc_norm": 1.4377,
      "encoder_q-embeddings": 670.2951,
      "encoder_q-layer.0": 464.3238,
      "encoder_q-layer.1": 529.0363,
      "encoder_q-layer.10": 559.6861,
      "encoder_q-layer.11": 1190.3253,
      "encoder_q-layer.2": 588.9677,
      "encoder_q-layer.3": 637.9328,
      "encoder_q-layer.4": 697.6728,
      "encoder_q-layer.5": 711.5396,
      "encoder_q-layer.6": 717.5536,
      "encoder_q-layer.7": 812.1486,
      "encoder_q-layer.8": 814.0258,
      "encoder_q-layer.9": 585.0984,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2544,
      "inbatch_pos_score": 0.9131,
      "learning_rate": 2.8055555555555557e-05,
      "loss": 3.261,
      "norm_diff": 0.0582,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1055.6176,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2542,
      "query_norm": 1.3795,
      "queue_k_norm": 1.4351,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8175,
      "sent_len_1": 66.6153,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5163,
      "stdk": 0.0487,
      "stdq": 0.0454,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 49500
    },
    {
      "accuracy": 51.2695,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2698,
      "doc_norm": 1.4374,
      "encoder_q-embeddings": 639.9524,
      "encoder_q-layer.0": 433.8493,
      "encoder_q-layer.1": 477.6505,
      "encoder_q-layer.10": 596.1375,
      "encoder_q-layer.11": 1267.7473,
      "encoder_q-layer.2": 529.642,
      "encoder_q-layer.3": 549.7073,
      "encoder_q-layer.4": 576.8627,
      "encoder_q-layer.5": 609.8173,
      "encoder_q-layer.6": 636.2068,
      "encoder_q-layer.7": 693.9382,
      "encoder_q-layer.8": 758.7448,
      "encoder_q-layer.9": 601.1351,
      "epoch": 0.48,
      "inbatch_neg_score": 0.2551,
      "inbatch_pos_score": 0.894,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 3.2698,
      "norm_diff": 0.0724,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1014.3912,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2542,
      "query_norm": 1.365,
      "queue_k_norm": 1.4362,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9943,
      "sent_len_1": 66.8838,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.58,
      "stdk": 0.0487,
      "stdq": 0.0449,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 49600
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2781,
      "doc_norm": 1.4354,
      "encoder_q-embeddings": 527.7044,
      "encoder_q-layer.0": 339.0107,
      "encoder_q-layer.1": 361.2082,
      "encoder_q-layer.10": 635.0565,
      "encoder_q-layer.11": 1340.6644,
      "encoder_q-layer.2": 404.3798,
      "encoder_q-layer.3": 410.4211,
      "encoder_q-layer.4": 450.1935,
      "encoder_q-layer.5": 469.7354,
      "encoder_q-layer.6": 555.3041,
      "encoder_q-layer.7": 629.2714,
      "encoder_q-layer.8": 734.3181,
      "encoder_q-layer.9": 647.3668,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2541,
      "inbatch_pos_score": 0.9014,
      "learning_rate": 2.7944444444444445e-05,
      "loss": 3.2781,
      "norm_diff": 0.0473,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 941.5673,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2532,
      "query_norm": 1.3882,
      "queue_k_norm": 1.4366,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9019,
      "sent_len_1": 66.6103,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2325,
      "stdk": 0.0486,
      "stdq": 0.0457,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 49700
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2694,
      "doc_norm": 1.4388,
      "encoder_q-embeddings": 1114.4673,
      "encoder_q-layer.0": 790.269,
      "encoder_q-layer.1": 983.6457,
      "encoder_q-layer.10": 587.7758,
      "encoder_q-layer.11": 1348.8569,
      "encoder_q-layer.2": 1148.7114,
      "encoder_q-layer.3": 1195.8994,
      "encoder_q-layer.4": 1274.5671,
      "encoder_q-layer.5": 1309.0521,
      "encoder_q-layer.6": 1134.8138,
      "encoder_q-layer.7": 1142.5459,
      "encoder_q-layer.8": 928.8678,
      "encoder_q-layer.9": 643.5111,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2569,
      "inbatch_pos_score": 0.8794,
      "learning_rate": 2.788888888888889e-05,
      "loss": 3.2694,
      "norm_diff": 0.0729,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1597.4312,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2559,
      "query_norm": 1.366,
      "queue_k_norm": 1.4353,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8295,
      "sent_len_1": 66.595,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.665,
      "stdk": 0.0487,
      "stdq": 0.0449,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 49800
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2659,
      "doc_norm": 1.434,
      "encoder_q-embeddings": 1040.2062,
      "encoder_q-layer.0": 676.837,
      "encoder_q-layer.1": 714.893,
      "encoder_q-layer.10": 1234.5951,
      "encoder_q-layer.11": 2755.4761,
      "encoder_q-layer.2": 794.3683,
      "encoder_q-layer.3": 821.9664,
      "encoder_q-layer.4": 874.4783,
      "encoder_q-layer.5": 948.3627,
      "encoder_q-layer.6": 1071.0786,
      "encoder_q-layer.7": 1297.5349,
      "encoder_q-layer.8": 1513.7731,
      "encoder_q-layer.9": 1282.4891,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2594,
      "inbatch_pos_score": 0.9253,
      "learning_rate": 2.7833333333333333e-05,
      "loss": 3.2659,
      "norm_diff": 0.041,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1911.9599,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2583,
      "query_norm": 1.393,
      "queue_k_norm": 1.436,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.974,
      "sent_len_1": 66.7733,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9162,
      "stdk": 0.0485,
      "stdq": 0.0458,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 49900
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2853,
      "doc_norm": 1.4345,
      "encoder_q-embeddings": 1055.1362,
      "encoder_q-layer.0": 684.3519,
      "encoder_q-layer.1": 746.9404,
      "encoder_q-layer.10": 1129.9023,
      "encoder_q-layer.11": 2476.2202,
      "encoder_q-layer.2": 855.759,
      "encoder_q-layer.3": 920.9887,
      "encoder_q-layer.4": 1028.2083,
      "encoder_q-layer.5": 993.6367,
      "encoder_q-layer.6": 1116.8638,
      "encoder_q-layer.7": 1245.4443,
      "encoder_q-layer.8": 1435.1512,
      "encoder_q-layer.9": 1256.2905,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2621,
      "inbatch_pos_score": 0.9404,
      "learning_rate": 2.777777777777778e-05,
      "loss": 3.2853,
      "norm_diff": 0.0289,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1847.4731,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2607,
      "query_norm": 1.4056,
      "queue_k_norm": 1.4362,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7732,
      "sent_len_1": 66.7012,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3587,
      "stdk": 0.0486,
      "stdq": 0.0462,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 50000
    },
    {
      "dev_runtime": 37.2579,
      "dev_samples_per_second": 1.718,
      "dev_steps_per_second": 0.027,
      "epoch": 0.49,
      "step": 50000,
      "test_accuracy": 94.2626953125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.32980793714523315,
      "test_doc_norm": 1.4416618347167969,
      "test_inbatch_neg_score": 0.6341235637664795,
      "test_inbatch_pos_score": 1.6145732402801514,
      "test_loss": 0.32980793714523315,
      "test_loss_align": 1.106376051902771,
      "test_loss_unif": 3.8641538619995117,
      "test_loss_unif_q@queue": 3.8641536235809326,
      "test_norm_diff": 0.05056659132242203,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.243794783949852,
      "test_query_norm": 1.4922285079956055,
      "test_queue_k_norm": 1.4359393119812012,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04328026995062828,
      "test_stdq": 0.04397352784872055,
      "test_stdqueue_k": 0.048719923943281174,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 37.2579,
      "dev_samples_per_second": 1.718,
      "dev_steps_per_second": 0.027,
      "epoch": 0.49,
      "eval_beir-arguana_ndcg@10": 0.35385,
      "eval_beir-arguana_recall@10": 0.60953,
      "eval_beir-arguana_recall@100": 0.9111,
      "eval_beir-arguana_recall@20": 0.74395,
      "eval_beir-avg_ndcg@10": 0.3668061666666667,
      "eval_beir-avg_recall@10": 0.433861,
      "eval_beir-avg_recall@100": 0.606643,
      "eval_beir-avg_recall@20": 0.4888535,
      "eval_beir-cqadupstack_ndcg@10": 0.25952166666666665,
      "eval_beir-cqadupstack_recall@10": 0.34896,
      "eval_beir-cqadupstack_recall@100": 0.5782299999999999,
      "eval_beir-cqadupstack_recall@20": 0.411715,
      "eval_beir-fiqa_ndcg@10": 0.21034,
      "eval_beir-fiqa_recall@10": 0.26955,
      "eval_beir-fiqa_recall@100": 0.50806,
      "eval_beir-fiqa_recall@20": 0.33142,
      "eval_beir-nfcorpus_ndcg@10": 0.27642,
      "eval_beir-nfcorpus_recall@10": 0.13751,
      "eval_beir-nfcorpus_recall@100": 0.25811,
      "eval_beir-nfcorpus_recall@20": 0.16814,
      "eval_beir-nq_ndcg@10": 0.26792,
      "eval_beir-nq_recall@10": 0.43839,
      "eval_beir-nq_recall@100": 0.77083,
      "eval_beir-nq_recall@20": 0.55123,
      "eval_beir-quora_ndcg@10": 0.81686,
      "eval_beir-quora_recall@10": 0.91333,
      "eval_beir-quora_recall@100": 0.98438,
      "eval_beir-quora_recall@20": 0.94872,
      "eval_beir-scidocs_ndcg@10": 0.13662,
      "eval_beir-scidocs_recall@10": 0.14343,
      "eval_beir-scidocs_recall@100": 0.33678,
      "eval_beir-scidocs_recall@20": 0.19703,
      "eval_beir-scifact_ndcg@10": 0.59899,
      "eval_beir-scifact_recall@10": 0.7465,
      "eval_beir-scifact_recall@100": 0.91522,
      "eval_beir-scifact_recall@20": 0.80067,
      "eval_beir-trec-covid_ndcg@10": 0.56107,
      "eval_beir-trec-covid_recall@10": 0.592,
      "eval_beir-trec-covid_recall@100": 0.394,
      "eval_beir-trec-covid_recall@20": 0.525,
      "eval_beir-webis-touche2020_ndcg@10": 0.18647,
      "eval_beir-webis-touche2020_recall@10": 0.13941,
      "eval_beir-webis-touche2020_recall@100": 0.40972,
      "eval_beir-webis-touche2020_recall@20": 0.21066,
      "eval_senteval-avg_sts": 0.7340157330939798,
      "eval_senteval-sickr_spearman": 0.7026975116568629,
      "eval_senteval-stsb_spearman": 0.7653339545310969,
      "step": 50000,
      "test_accuracy": 94.2626953125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.32980793714523315,
      "test_doc_norm": 1.4416618347167969,
      "test_inbatch_neg_score": 0.6341235637664795,
      "test_inbatch_pos_score": 1.6145732402801514,
      "test_loss": 0.32980793714523315,
      "test_loss_align": 1.106376051902771,
      "test_loss_unif": 3.8641538619995117,
      "test_loss_unif_q@queue": 3.8641536235809326,
      "test_norm_diff": 0.05056659132242203,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.243794783949852,
      "test_query_norm": 1.4922285079956055,
      "test_queue_k_norm": 1.4359393119812012,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04328026995062828,
      "test_stdq": 0.04397352784872055,
      "test_stdqueue_k": 0.048719923943281174,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2947,
      "doc_norm": 1.4341,
      "encoder_q-embeddings": 1211.6646,
      "encoder_q-layer.0": 792.7898,
      "encoder_q-layer.1": 899.413,
      "encoder_q-layer.10": 1147.7826,
      "encoder_q-layer.11": 2531.7803,
      "encoder_q-layer.2": 1033.4694,
      "encoder_q-layer.3": 1052.1979,
      "encoder_q-layer.4": 1126.5465,
      "encoder_q-layer.5": 1142.4198,
      "encoder_q-layer.6": 1304.8063,
      "encoder_q-layer.7": 1445.0179,
      "encoder_q-layer.8": 1658.5203,
      "encoder_q-layer.9": 1365.713,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2606,
      "inbatch_pos_score": 0.9136,
      "learning_rate": 2.772222222222222e-05,
      "loss": 3.2947,
      "norm_diff": 0.0481,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2024.693,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2612,
      "query_norm": 1.3861,
      "queue_k_norm": 1.4358,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8835,
      "sent_len_1": 66.579,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3088,
      "stdk": 0.0485,
      "stdq": 0.0454,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 50100
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2655,
      "doc_norm": 1.4414,
      "encoder_q-embeddings": 1207.978,
      "encoder_q-layer.0": 744.3564,
      "encoder_q-layer.1": 795.8477,
      "encoder_q-layer.10": 1262.3108,
      "encoder_q-layer.11": 2605.5647,
      "encoder_q-layer.2": 899.9904,
      "encoder_q-layer.3": 969.5113,
      "encoder_q-layer.4": 1031.207,
      "encoder_q-layer.5": 1077.5168,
      "encoder_q-layer.6": 1199.4794,
      "encoder_q-layer.7": 1361.3185,
      "encoder_q-layer.8": 1563.3794,
      "encoder_q-layer.9": 1360.3708,
      "epoch": 0.49,
      "inbatch_neg_score": 0.258,
      "inbatch_pos_score": 0.9175,
      "learning_rate": 2.7666666666666667e-05,
      "loss": 3.2655,
      "norm_diff": 0.0542,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1957.6102,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2573,
      "query_norm": 1.3873,
      "queue_k_norm": 1.4368,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8582,
      "sent_len_1": 66.7203,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1087,
      "stdk": 0.0488,
      "stdq": 0.0455,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 50200
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2784,
      "doc_norm": 1.4374,
      "encoder_q-embeddings": 1765.6559,
      "encoder_q-layer.0": 1168.9316,
      "encoder_q-layer.1": 1422.0564,
      "encoder_q-layer.10": 1167.4447,
      "encoder_q-layer.11": 2598.1196,
      "encoder_q-layer.2": 1700.7476,
      "encoder_q-layer.3": 1859.4659,
      "encoder_q-layer.4": 1934.5746,
      "encoder_q-layer.5": 2215.3433,
      "encoder_q-layer.6": 2326.8479,
      "encoder_q-layer.7": 2360.5522,
      "encoder_q-layer.8": 2127.3074,
      "encoder_q-layer.9": 1484.8362,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2588,
      "inbatch_pos_score": 0.936,
      "learning_rate": 2.761111111111111e-05,
      "loss": 3.2784,
      "norm_diff": 0.0455,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2848.7958,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2598,
      "query_norm": 1.3919,
      "queue_k_norm": 1.4381,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9632,
      "sent_len_1": 66.783,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1813,
      "stdk": 0.0486,
      "stdq": 0.0456,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 50300
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2976,
      "doc_norm": 1.4375,
      "encoder_q-embeddings": 954.5184,
      "encoder_q-layer.0": 635.1726,
      "encoder_q-layer.1": 675.2886,
      "encoder_q-layer.10": 1310.5186,
      "encoder_q-layer.11": 2435.7488,
      "encoder_q-layer.2": 767.007,
      "encoder_q-layer.3": 800.6089,
      "encoder_q-layer.4": 831.1558,
      "encoder_q-layer.5": 877.3548,
      "encoder_q-layer.6": 998.9734,
      "encoder_q-layer.7": 1153.4221,
      "encoder_q-layer.8": 1359.4242,
      "encoder_q-layer.9": 1198.1189,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2615,
      "inbatch_pos_score": 0.9043,
      "learning_rate": 2.7555555555555555e-05,
      "loss": 3.2976,
      "norm_diff": 0.0406,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1725.3252,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2617,
      "query_norm": 1.3969,
      "queue_k_norm": 1.4382,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9739,
      "sent_len_1": 66.8045,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2537,
      "stdk": 0.0487,
      "stdq": 0.0456,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 50400
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2802,
      "doc_norm": 1.4408,
      "encoder_q-embeddings": 1428.1772,
      "encoder_q-layer.0": 1018.8244,
      "encoder_q-layer.1": 1098.9192,
      "encoder_q-layer.10": 1123.7081,
      "encoder_q-layer.11": 2463.4946,
      "encoder_q-layer.2": 1337.5638,
      "encoder_q-layer.3": 1381.8877,
      "encoder_q-layer.4": 1373.7654,
      "encoder_q-layer.5": 1310.3585,
      "encoder_q-layer.6": 1230.3354,
      "encoder_q-layer.7": 1378.9727,
      "encoder_q-layer.8": 1487.8324,
      "encoder_q-layer.9": 1234.2593,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2674,
      "inbatch_pos_score": 0.9287,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 3.2802,
      "norm_diff": 0.0535,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2114.4005,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2664,
      "query_norm": 1.3873,
      "queue_k_norm": 1.439,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0151,
      "sent_len_1": 66.7572,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1413,
      "stdk": 0.0487,
      "stdq": 0.045,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 50500
    },
    {
      "accuracy": 57.2266,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3005,
      "doc_norm": 1.4325,
      "encoder_q-embeddings": 1153.769,
      "encoder_q-layer.0": 759.8857,
      "encoder_q-layer.1": 803.3187,
      "encoder_q-layer.10": 1167.4299,
      "encoder_q-layer.11": 2452.1709,
      "encoder_q-layer.2": 883.792,
      "encoder_q-layer.3": 926.6732,
      "encoder_q-layer.4": 986.6658,
      "encoder_q-layer.5": 1102.5579,
      "encoder_q-layer.6": 1163.6055,
      "encoder_q-layer.7": 1263.8777,
      "encoder_q-layer.8": 1357.5933,
      "encoder_q-layer.9": 1128.3422,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2708,
      "inbatch_pos_score": 0.9409,
      "learning_rate": 2.7444444444444443e-05,
      "loss": 3.3005,
      "norm_diff": 0.0338,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1838.3568,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2705,
      "query_norm": 1.3987,
      "queue_k_norm": 1.4403,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7369,
      "sent_len_1": 66.7605,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0462,
      "stdk": 0.0484,
      "stdq": 0.0453,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 50600
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2622,
      "doc_norm": 1.4413,
      "encoder_q-embeddings": 1097.0942,
      "encoder_q-layer.0": 690.6096,
      "encoder_q-layer.1": 723.7868,
      "encoder_q-layer.10": 1250.9301,
      "encoder_q-layer.11": 2711.6902,
      "encoder_q-layer.2": 802.0071,
      "encoder_q-layer.3": 865.0005,
      "encoder_q-layer.4": 944.0429,
      "encoder_q-layer.5": 1006.4625,
      "encoder_q-layer.6": 1092.9701,
      "encoder_q-layer.7": 1243.7129,
      "encoder_q-layer.8": 1426.7479,
      "encoder_q-layer.9": 1237.1437,
      "epoch": 0.49,
      "inbatch_neg_score": 0.2763,
      "inbatch_pos_score": 0.9658,
      "learning_rate": 2.7388888888888892e-05,
      "loss": 3.2622,
      "norm_diff": 0.0268,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1908.1977,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2756,
      "query_norm": 1.4145,
      "queue_k_norm": 1.4406,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9272,
      "sent_len_1": 66.982,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7325,
      "stdk": 0.0487,
      "stdq": 0.0458,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 50700
    },
    {
      "accuracy": 56.9336,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2665,
      "doc_norm": 1.4479,
      "encoder_q-embeddings": 1019.4509,
      "encoder_q-layer.0": 686.3734,
      "encoder_q-layer.1": 727.0787,
      "encoder_q-layer.10": 1128.7607,
      "encoder_q-layer.11": 2583.7151,
      "encoder_q-layer.2": 813.1434,
      "encoder_q-layer.3": 845.6356,
      "encoder_q-layer.4": 919.1758,
      "encoder_q-layer.5": 912.0933,
      "encoder_q-layer.6": 1089.1222,
      "encoder_q-layer.7": 1226.6792,
      "encoder_q-layer.8": 1466.3184,
      "encoder_q-layer.9": 1266.4097,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2804,
      "inbatch_pos_score": 0.978,
      "learning_rate": 2.733333333333333e-05,
      "loss": 3.2665,
      "norm_diff": 0.0183,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1842.2471,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.281,
      "query_norm": 1.4296,
      "queue_k_norm": 1.4423,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9366,
      "sent_len_1": 66.8677,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8613,
      "stdk": 0.0489,
      "stdq": 0.0462,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 50800
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2491,
      "doc_norm": 1.4432,
      "encoder_q-embeddings": 1123.4386,
      "encoder_q-layer.0": 790.5886,
      "encoder_q-layer.1": 844.0469,
      "encoder_q-layer.10": 1426.6189,
      "encoder_q-layer.11": 2770.3386,
      "encoder_q-layer.2": 925.0001,
      "encoder_q-layer.3": 945.0209,
      "encoder_q-layer.4": 1034.1024,
      "encoder_q-layer.5": 1089.2164,
      "encoder_q-layer.6": 1322.8473,
      "encoder_q-layer.7": 1510.824,
      "encoder_q-layer.8": 1740.3009,
      "encoder_q-layer.9": 1489.2295,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2869,
      "inbatch_pos_score": 0.9414,
      "learning_rate": 2.727777777777778e-05,
      "loss": 3.2491,
      "norm_diff": 0.0496,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2064.3833,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2869,
      "query_norm": 1.3936,
      "queue_k_norm": 1.4403,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.822,
      "sent_len_1": 66.5197,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4538,
      "stdk": 0.0487,
      "stdq": 0.0448,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 50900
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2815,
      "doc_norm": 1.438,
      "encoder_q-embeddings": 2220.0398,
      "encoder_q-layer.0": 1568.1479,
      "encoder_q-layer.1": 1792.9387,
      "encoder_q-layer.10": 1147.9126,
      "encoder_q-layer.11": 2522.082,
      "encoder_q-layer.2": 2104.2612,
      "encoder_q-layer.3": 2024.3868,
      "encoder_q-layer.4": 2317.907,
      "encoder_q-layer.5": 2161.7134,
      "encoder_q-layer.6": 2181.7244,
      "encoder_q-layer.7": 1956.7344,
      "encoder_q-layer.8": 1953.9933,
      "encoder_q-layer.9": 1487.8584,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2899,
      "inbatch_pos_score": 0.9741,
      "learning_rate": 2.7222222222222223e-05,
      "loss": 3.2815,
      "norm_diff": 0.0069,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2952.3577,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2893,
      "query_norm": 1.4377,
      "queue_k_norm": 1.4426,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7962,
      "sent_len_1": 66.6016,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1037,
      "stdk": 0.0485,
      "stdq": 0.0465,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51000
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2602,
      "doc_norm": 1.4486,
      "encoder_q-embeddings": 1009.5827,
      "encoder_q-layer.0": 636.9775,
      "encoder_q-layer.1": 688.6617,
      "encoder_q-layer.10": 1254.7466,
      "encoder_q-layer.11": 2659.0186,
      "encoder_q-layer.2": 753.02,
      "encoder_q-layer.3": 779.0068,
      "encoder_q-layer.4": 807.4285,
      "encoder_q-layer.5": 903.6761,
      "encoder_q-layer.6": 995.0179,
      "encoder_q-layer.7": 1139.2833,
      "encoder_q-layer.8": 1404.1119,
      "encoder_q-layer.9": 1196.4259,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2901,
      "inbatch_pos_score": 0.9688,
      "learning_rate": 2.716666666666667e-05,
      "loss": 3.2602,
      "norm_diff": 0.0159,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1802.0676,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2898,
      "query_norm": 1.438,
      "queue_k_norm": 1.4454,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0127,
      "sent_len_1": 66.8526,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.365,
      "stdk": 0.0489,
      "stdq": 0.0464,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 51100
    },
    {
      "accuracy": 56.9336,
      "active_queue_size": 16384.0,
      "cl_loss": 3.276,
      "doc_norm": 1.4508,
      "encoder_q-embeddings": 1723.6058,
      "encoder_q-layer.0": 1122.0077,
      "encoder_q-layer.1": 1274.3641,
      "encoder_q-layer.10": 1170.4948,
      "encoder_q-layer.11": 2614.2214,
      "encoder_q-layer.2": 1557.7628,
      "encoder_q-layer.3": 1737.8528,
      "encoder_q-layer.4": 1749.2527,
      "encoder_q-layer.5": 1836.8586,
      "encoder_q-layer.6": 1768.502,
      "encoder_q-layer.7": 1814.5582,
      "encoder_q-layer.8": 1624.8115,
      "encoder_q-layer.9": 1199.8207,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2922,
      "inbatch_pos_score": 0.9756,
      "learning_rate": 2.7111111111111114e-05,
      "loss": 3.276,
      "norm_diff": 0.0315,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2528.7995,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2935,
      "query_norm": 1.4192,
      "queue_k_norm": 1.4435,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8029,
      "sent_len_1": 66.8065,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.38,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51200
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.3002,
      "doc_norm": 1.4441,
      "encoder_q-embeddings": 1338.3572,
      "encoder_q-layer.0": 900.2236,
      "encoder_q-layer.1": 967.2972,
      "encoder_q-layer.10": 1194.7307,
      "encoder_q-layer.11": 2587.0698,
      "encoder_q-layer.2": 1060.7356,
      "encoder_q-layer.3": 1107.3979,
      "encoder_q-layer.4": 1201.5321,
      "encoder_q-layer.5": 1228.3876,
      "encoder_q-layer.6": 1302.9744,
      "encoder_q-layer.7": 1426.8362,
      "encoder_q-layer.8": 1510.8982,
      "encoder_q-layer.9": 1284.0896,
      "epoch": 0.5,
      "inbatch_neg_score": 0.291,
      "inbatch_pos_score": 0.9414,
      "learning_rate": 2.7055555555555557e-05,
      "loss": 3.3002,
      "norm_diff": 0.0403,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2075.3779,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2903,
      "query_norm": 1.4039,
      "queue_k_norm": 1.4441,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8996,
      "sent_len_1": 66.4845,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0975,
      "stdk": 0.0486,
      "stdq": 0.0454,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51300
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2871,
      "doc_norm": 1.4496,
      "encoder_q-embeddings": 2659.5273,
      "encoder_q-layer.0": 1715.2466,
      "encoder_q-layer.1": 1758.1676,
      "encoder_q-layer.10": 1199.6017,
      "encoder_q-layer.11": 2804.0176,
      "encoder_q-layer.2": 2002.5017,
      "encoder_q-layer.3": 2171.3596,
      "encoder_q-layer.4": 2447.6096,
      "encoder_q-layer.5": 2344.2251,
      "encoder_q-layer.6": 2013.2656,
      "encoder_q-layer.7": 1987.2859,
      "encoder_q-layer.8": 1710.4736,
      "encoder_q-layer.9": 1271.9417,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2979,
      "inbatch_pos_score": 0.9336,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 3.2871,
      "norm_diff": 0.0484,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3144.9979,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2971,
      "query_norm": 1.4011,
      "queue_k_norm": 1.4444,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7978,
      "sent_len_1": 66.6128,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8375,
      "stdk": 0.0488,
      "stdq": 0.0454,
      "stdqueue_k": 0.0487,
      "stdqueue_q": 0.0,
      "step": 51400
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2952,
      "doc_norm": 1.4472,
      "encoder_q-embeddings": 1237.239,
      "encoder_q-layer.0": 791.0626,
      "encoder_q-layer.1": 828.5809,
      "encoder_q-layer.10": 1214.4211,
      "encoder_q-layer.11": 2795.3115,
      "encoder_q-layer.2": 928.469,
      "encoder_q-layer.3": 998.6166,
      "encoder_q-layer.4": 1062.6223,
      "encoder_q-layer.5": 1077.6219,
      "encoder_q-layer.6": 1169.0847,
      "encoder_q-layer.7": 1389.251,
      "encoder_q-layer.8": 1468.6864,
      "encoder_q-layer.9": 1223.3564,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2958,
      "inbatch_pos_score": 0.9712,
      "learning_rate": 2.6944444444444445e-05,
      "loss": 3.2952,
      "norm_diff": 0.0304,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2013.6185,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.2961,
      "query_norm": 1.4167,
      "queue_k_norm": 1.4474,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9549,
      "sent_len_1": 66.7634,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.81,
      "stdk": 0.0487,
      "stdq": 0.0461,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51500
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2544,
      "doc_norm": 1.4486,
      "encoder_q-embeddings": 2165.8784,
      "encoder_q-layer.0": 1487.7036,
      "encoder_q-layer.1": 1703.0875,
      "encoder_q-layer.10": 1157.6433,
      "encoder_q-layer.11": 2769.354,
      "encoder_q-layer.2": 1987.9492,
      "encoder_q-layer.3": 2278.6904,
      "encoder_q-layer.4": 2472.2561,
      "encoder_q-layer.5": 2458.7173,
      "encoder_q-layer.6": 2414.5176,
      "encoder_q-layer.7": 2220.2673,
      "encoder_q-layer.8": 2168.0903,
      "encoder_q-layer.9": 1457.9832,
      "epoch": 0.5,
      "inbatch_neg_score": 0.2993,
      "inbatch_pos_score": 0.9707,
      "learning_rate": 2.688888888888889e-05,
      "loss": 3.2544,
      "norm_diff": 0.0435,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3209.2698,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3,
      "query_norm": 1.4051,
      "queue_k_norm": 1.4478,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1709,
      "sent_len_1": 66.7817,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.88,
      "stdk": 0.0488,
      "stdq": 0.0456,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51600
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2536,
      "doc_norm": 1.4475,
      "encoder_q-embeddings": 969.1141,
      "encoder_q-layer.0": 642.3048,
      "encoder_q-layer.1": 664.6626,
      "encoder_q-layer.10": 1174.2742,
      "encoder_q-layer.11": 2858.5254,
      "encoder_q-layer.2": 753.7573,
      "encoder_q-layer.3": 768.1957,
      "encoder_q-layer.4": 818.3869,
      "encoder_q-layer.5": 855.6999,
      "encoder_q-layer.6": 988.6628,
      "encoder_q-layer.7": 1162.6769,
      "encoder_q-layer.8": 1373.1758,
      "encoder_q-layer.9": 1195.8572,
      "epoch": 0.5,
      "inbatch_neg_score": 0.3022,
      "inbatch_pos_score": 0.9409,
      "learning_rate": 2.6833333333333333e-05,
      "loss": 3.2536,
      "norm_diff": 0.058,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1877.6536,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.302,
      "query_norm": 1.3895,
      "queue_k_norm": 1.4486,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2201,
      "sent_len_1": 67.0927,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.0437,
      "stdk": 0.0487,
      "stdq": 0.045,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51700
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2774,
      "doc_norm": 1.4489,
      "encoder_q-embeddings": 1256.1582,
      "encoder_q-layer.0": 833.2198,
      "encoder_q-layer.1": 924.1592,
      "encoder_q-layer.10": 1233.8258,
      "encoder_q-layer.11": 2581.1216,
      "encoder_q-layer.2": 1032.7875,
      "encoder_q-layer.3": 1147.6764,
      "encoder_q-layer.4": 1222.186,
      "encoder_q-layer.5": 1325.7236,
      "encoder_q-layer.6": 1368.3606,
      "encoder_q-layer.7": 1427.3041,
      "encoder_q-layer.8": 1489.9845,
      "encoder_q-layer.9": 1160.4004,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3013,
      "inbatch_pos_score": 0.9595,
      "learning_rate": 2.677777777777778e-05,
      "loss": 3.2774,
      "norm_diff": 0.0626,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2066.7077,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3013,
      "query_norm": 1.3864,
      "queue_k_norm": 1.4496,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9184,
      "sent_len_1": 67.0613,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6962,
      "stdk": 0.0487,
      "stdq": 0.0449,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51800
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2684,
      "doc_norm": 1.4526,
      "encoder_q-embeddings": 2529.9556,
      "encoder_q-layer.0": 1686.6755,
      "encoder_q-layer.1": 1853.4084,
      "encoder_q-layer.10": 2231.1294,
      "encoder_q-layer.11": 5462.7642,
      "encoder_q-layer.2": 2087.5757,
      "encoder_q-layer.3": 2233.9949,
      "encoder_q-layer.4": 2348.4836,
      "encoder_q-layer.5": 2384.6833,
      "encoder_q-layer.6": 2538.8872,
      "encoder_q-layer.7": 2657.7949,
      "encoder_q-layer.8": 2804.5466,
      "encoder_q-layer.9": 2362.1243,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3038,
      "inbatch_pos_score": 0.9629,
      "learning_rate": 2.6722222222222228e-05,
      "loss": 3.2684,
      "norm_diff": 0.0539,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4013.4874,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3049,
      "query_norm": 1.3986,
      "queue_k_norm": 1.4503,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0697,
      "sent_len_1": 66.7456,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5012,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 51900
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.259,
      "doc_norm": 1.4467,
      "encoder_q-embeddings": 3033.0173,
      "encoder_q-layer.0": 2047.2709,
      "encoder_q-layer.1": 2433.8833,
      "encoder_q-layer.10": 2427.7825,
      "encoder_q-layer.11": 5564.5396,
      "encoder_q-layer.2": 2812.0322,
      "encoder_q-layer.3": 2979.5,
      "encoder_q-layer.4": 3297.4167,
      "encoder_q-layer.5": 3086.604,
      "encoder_q-layer.6": 3170.0203,
      "encoder_q-layer.7": 3106.5581,
      "encoder_q-layer.8": 2979.2588,
      "encoder_q-layer.9": 2399.1279,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3015,
      "inbatch_pos_score": 0.9614,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 3.259,
      "norm_diff": 0.0498,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4684.1284,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3027,
      "query_norm": 1.3969,
      "queue_k_norm": 1.4504,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9123,
      "sent_len_1": 66.7628,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5425,
      "stdk": 0.0486,
      "stdq": 0.0453,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52000
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2551,
      "doc_norm": 1.4589,
      "encoder_q-embeddings": 2361.54,
      "encoder_q-layer.0": 1590.246,
      "encoder_q-layer.1": 1831.9202,
      "encoder_q-layer.10": 2434.6265,
      "encoder_q-layer.11": 5340.8027,
      "encoder_q-layer.2": 2049.5586,
      "encoder_q-layer.3": 2122.2317,
      "encoder_q-layer.4": 2228.5354,
      "encoder_q-layer.5": 2277.3682,
      "encoder_q-layer.6": 2430.8271,
      "encoder_q-layer.7": 2695.4519,
      "encoder_q-layer.8": 2876.448,
      "encoder_q-layer.9": 2487.7043,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3062,
      "inbatch_pos_score": 0.9907,
      "learning_rate": 2.6611111111111116e-05,
      "loss": 3.2551,
      "norm_diff": 0.0558,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4035.6363,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3062,
      "query_norm": 1.4031,
      "queue_k_norm": 1.4502,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2445,
      "sent_len_1": 66.8362,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0825,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52100
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2885,
      "doc_norm": 1.4576,
      "encoder_q-embeddings": 2441.5417,
      "encoder_q-layer.0": 1555.229,
      "encoder_q-layer.1": 1641.8291,
      "encoder_q-layer.10": 2439.7886,
      "encoder_q-layer.11": 5521.5518,
      "encoder_q-layer.2": 1849.7288,
      "encoder_q-layer.3": 1942.5228,
      "encoder_q-layer.4": 2137.7949,
      "encoder_q-layer.5": 2126.2378,
      "encoder_q-layer.6": 2435.2314,
      "encoder_q-layer.7": 2958.1292,
      "encoder_q-layer.8": 3134.2014,
      "encoder_q-layer.9": 2578.3206,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3043,
      "inbatch_pos_score": 0.9595,
      "learning_rate": 2.6555555555555555e-05,
      "loss": 3.2885,
      "norm_diff": 0.058,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4103.4679,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.304,
      "query_norm": 1.3995,
      "queue_k_norm": 1.452,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9268,
      "sent_len_1": 66.894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 193.245,
      "stdk": 0.0489,
      "stdq": 0.0454,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52200
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2659,
      "doc_norm": 1.4536,
      "encoder_q-embeddings": 2475.5537,
      "encoder_q-layer.0": 1583.9683,
      "encoder_q-layer.1": 1677.0433,
      "encoder_q-layer.10": 2380.0686,
      "encoder_q-layer.11": 5387.8457,
      "encoder_q-layer.2": 1963.1552,
      "encoder_q-layer.3": 2071.1589,
      "encoder_q-layer.4": 2141.7131,
      "encoder_q-layer.5": 2152.915,
      "encoder_q-layer.6": 2380.1091,
      "encoder_q-layer.7": 2629.6094,
      "encoder_q-layer.8": 2887.5278,
      "encoder_q-layer.9": 2459.5403,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3111,
      "inbatch_pos_score": 0.9497,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 3.2659,
      "norm_diff": 0.0515,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4031.8182,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3105,
      "query_norm": 1.402,
      "queue_k_norm": 1.452,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8795,
      "sent_len_1": 66.6744,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4625,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52300
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2809,
      "doc_norm": 1.4559,
      "encoder_q-embeddings": 1941.0386,
      "encoder_q-layer.0": 1242.5665,
      "encoder_q-layer.1": 1299.5121,
      "encoder_q-layer.10": 2304.7559,
      "encoder_q-layer.11": 5088.1494,
      "encoder_q-layer.2": 1467.5684,
      "encoder_q-layer.3": 1581.2965,
      "encoder_q-layer.4": 1678.145,
      "encoder_q-layer.5": 1670.7913,
      "encoder_q-layer.6": 1954.3693,
      "encoder_q-layer.7": 2251.0762,
      "encoder_q-layer.8": 2541.8018,
      "encoder_q-layer.9": 2363.2793,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3086,
      "inbatch_pos_score": 0.9956,
      "learning_rate": 2.6444444444444443e-05,
      "loss": 3.2809,
      "norm_diff": 0.0404,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3520.0242,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3091,
      "query_norm": 1.4155,
      "queue_k_norm": 1.4527,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9978,
      "sent_len_1": 66.6688,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8162,
      "stdk": 0.0489,
      "stdq": 0.0459,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52400
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2383,
      "doc_norm": 1.45,
      "encoder_q-embeddings": 3780.7537,
      "encoder_q-layer.0": 2482.3518,
      "encoder_q-layer.1": 2575.7002,
      "encoder_q-layer.10": 2306.2466,
      "encoder_q-layer.11": 5196.2178,
      "encoder_q-layer.2": 3109.2112,
      "encoder_q-layer.3": 3506.55,
      "encoder_q-layer.4": 3633.5725,
      "encoder_q-layer.5": 3547.5344,
      "encoder_q-layer.6": 3566.6033,
      "encoder_q-layer.7": 3479.3955,
      "encoder_q-layer.8": 3368.0918,
      "encoder_q-layer.9": 2499.9163,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3083,
      "inbatch_pos_score": 0.9741,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 3.2383,
      "norm_diff": 0.0386,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5142.7177,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3066,
      "query_norm": 1.4114,
      "queue_k_norm": 1.4536,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9209,
      "sent_len_1": 66.9175,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4038,
      "stdk": 0.0486,
      "stdq": 0.0458,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52500
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2823,
      "doc_norm": 1.4598,
      "encoder_q-embeddings": 2638.0964,
      "encoder_q-layer.0": 1720.6609,
      "encoder_q-layer.1": 1858.2151,
      "encoder_q-layer.10": 2358.6245,
      "encoder_q-layer.11": 5248.0176,
      "encoder_q-layer.2": 2119.7834,
      "encoder_q-layer.3": 2252.0027,
      "encoder_q-layer.4": 2408.9929,
      "encoder_q-layer.5": 2546.8865,
      "encoder_q-layer.6": 2687.0647,
      "encoder_q-layer.7": 3328.6409,
      "encoder_q-layer.8": 3383.615,
      "encoder_q-layer.9": 2583.5813,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3054,
      "inbatch_pos_score": 0.9468,
      "learning_rate": 2.633333333333333e-05,
      "loss": 3.2823,
      "norm_diff": 0.059,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4286.1376,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3047,
      "query_norm": 1.4008,
      "queue_k_norm": 1.4528,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7756,
      "sent_len_1": 66.692,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1525,
      "stdk": 0.049,
      "stdq": 0.0455,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52600
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2822,
      "doc_norm": 1.4599,
      "encoder_q-embeddings": 4765.2139,
      "encoder_q-layer.0": 3521.3997,
      "encoder_q-layer.1": 3877.8953,
      "encoder_q-layer.10": 2457.4885,
      "encoder_q-layer.11": 5721.958,
      "encoder_q-layer.2": 4464.8438,
      "encoder_q-layer.3": 4443.0586,
      "encoder_q-layer.4": 4301.6904,
      "encoder_q-layer.5": 4207.5479,
      "encoder_q-layer.6": 4598.3896,
      "encoder_q-layer.7": 4283.707,
      "encoder_q-layer.8": 3743.491,
      "encoder_q-layer.9": 2870.7637,
      "epoch": 0.51,
      "inbatch_neg_score": 0.3104,
      "inbatch_pos_score": 0.96,
      "learning_rate": 2.627777777777778e-05,
      "loss": 3.2822,
      "norm_diff": 0.0385,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6331.3822,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3098,
      "query_norm": 1.4214,
      "queue_k_norm": 1.4527,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8837,
      "sent_len_1": 66.8305,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.955,
      "stdk": 0.049,
      "stdq": 0.0462,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52700
    },
    {
      "accuracy": 51.5625,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2553,
      "doc_norm": 1.4525,
      "encoder_q-embeddings": 2021.922,
      "encoder_q-layer.0": 1325.9564,
      "encoder_q-layer.1": 1414.6069,
      "encoder_q-layer.10": 2595.0339,
      "encoder_q-layer.11": 5686.7417,
      "encoder_q-layer.2": 1664.1935,
      "encoder_q-layer.3": 1721.0531,
      "encoder_q-layer.4": 1866.9073,
      "encoder_q-layer.5": 1957.0818,
      "encoder_q-layer.6": 2246.4702,
      "encoder_q-layer.7": 2743.0239,
      "encoder_q-layer.8": 2966.917,
      "encoder_q-layer.9": 2690.3845,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3113,
      "inbatch_pos_score": 0.9668,
      "learning_rate": 2.6222222222222226e-05,
      "loss": 3.2553,
      "norm_diff": 0.0432,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3861.1265,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3105,
      "query_norm": 1.4094,
      "queue_k_norm": 1.4545,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7171,
      "sent_len_1": 66.5659,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.77,
      "stdk": 0.0486,
      "stdq": 0.0457,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52800
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2451,
      "doc_norm": 1.4565,
      "encoder_q-embeddings": 3015.7744,
      "encoder_q-layer.0": 1996.3484,
      "encoder_q-layer.1": 2244.9617,
      "encoder_q-layer.10": 2590.7407,
      "encoder_q-layer.11": 5759.564,
      "encoder_q-layer.2": 2624.3474,
      "encoder_q-layer.3": 2862.9958,
      "encoder_q-layer.4": 3108.7222,
      "encoder_q-layer.5": 3161.0122,
      "encoder_q-layer.6": 3328.1479,
      "encoder_q-layer.7": 3686.8052,
      "encoder_q-layer.8": 3903.0813,
      "encoder_q-layer.9": 3088.0574,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3055,
      "inbatch_pos_score": 0.9917,
      "learning_rate": 2.6166666666666668e-05,
      "loss": 3.2451,
      "norm_diff": 0.0202,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4926.5298,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3066,
      "query_norm": 1.4389,
      "queue_k_norm": 1.4539,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.832,
      "sent_len_1": 66.8759,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.91,
      "stdk": 0.0488,
      "stdq": 0.0469,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 52900
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2678,
      "doc_norm": 1.4541,
      "encoder_q-embeddings": 2934.2634,
      "encoder_q-layer.0": 1905.2008,
      "encoder_q-layer.1": 2104.4646,
      "encoder_q-layer.10": 2431.6289,
      "encoder_q-layer.11": 5374.3926,
      "encoder_q-layer.2": 2326.0884,
      "encoder_q-layer.3": 2422.9021,
      "encoder_q-layer.4": 2541.1909,
      "encoder_q-layer.5": 2700.042,
      "encoder_q-layer.6": 2792.2139,
      "encoder_q-layer.7": 2876.5571,
      "encoder_q-layer.8": 3180.6663,
      "encoder_q-layer.9": 2570.0298,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3093,
      "inbatch_pos_score": 0.978,
      "learning_rate": 2.6111111111111114e-05,
      "loss": 3.2678,
      "norm_diff": 0.0397,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4361.9361,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3086,
      "query_norm": 1.4144,
      "queue_k_norm": 1.4554,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0636,
      "sent_len_1": 66.8272,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9375,
      "stdk": 0.0487,
      "stdq": 0.046,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53000
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2631,
      "doc_norm": 1.4556,
      "encoder_q-embeddings": 1864.6393,
      "encoder_q-layer.0": 1180.8322,
      "encoder_q-layer.1": 1250.1144,
      "encoder_q-layer.10": 2364.3118,
      "encoder_q-layer.11": 5327.8789,
      "encoder_q-layer.2": 1402.3622,
      "encoder_q-layer.3": 1481.8773,
      "encoder_q-layer.4": 1607.0627,
      "encoder_q-layer.5": 1653.6544,
      "encoder_q-layer.6": 1921.6986,
      "encoder_q-layer.7": 2236.167,
      "encoder_q-layer.8": 2647.5989,
      "encoder_q-layer.9": 2396.5813,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3038,
      "inbatch_pos_score": 0.9771,
      "learning_rate": 2.6055555555555556e-05,
      "loss": 3.2631,
      "norm_diff": 0.0549,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3563.4463,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3044,
      "query_norm": 1.4007,
      "queue_k_norm": 1.4575,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0608,
      "sent_len_1": 66.7487,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3625,
      "stdk": 0.0487,
      "stdq": 0.0455,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 53100
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2623,
      "doc_norm": 1.4541,
      "encoder_q-embeddings": 1289.6663,
      "encoder_q-layer.0": 855.8948,
      "encoder_q-layer.1": 903.3082,
      "encoder_q-layer.10": 1281.5227,
      "encoder_q-layer.11": 2682.9382,
      "encoder_q-layer.2": 1033.2267,
      "encoder_q-layer.3": 1123.7324,
      "encoder_q-layer.4": 1255.8892,
      "encoder_q-layer.5": 1259.7306,
      "encoder_q-layer.6": 1341.1338,
      "encoder_q-layer.7": 1394.8898,
      "encoder_q-layer.8": 1537.7391,
      "encoder_q-layer.9": 1323.8893,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3047,
      "inbatch_pos_score": 0.9707,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 3.2623,
      "norm_diff": 0.0575,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2088.1066,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3047,
      "query_norm": 1.3966,
      "queue_k_norm": 1.4556,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7924,
      "sent_len_1": 66.6957,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2738,
      "stdk": 0.0487,
      "stdq": 0.0453,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53200
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.257,
      "doc_norm": 1.4485,
      "encoder_q-embeddings": 1919.9004,
      "encoder_q-layer.0": 1310.1681,
      "encoder_q-layer.1": 1275.0701,
      "encoder_q-layer.10": 1266.5015,
      "encoder_q-layer.11": 2817.6836,
      "encoder_q-layer.2": 1515.9335,
      "encoder_q-layer.3": 1701.0081,
      "encoder_q-layer.4": 1870.8693,
      "encoder_q-layer.5": 1711.2993,
      "encoder_q-layer.6": 1760.2579,
      "encoder_q-layer.7": 1712.668,
      "encoder_q-layer.8": 1718.6309,
      "encoder_q-layer.9": 1293.0289,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3053,
      "inbatch_pos_score": 0.9717,
      "learning_rate": 2.5944444444444444e-05,
      "loss": 3.257,
      "norm_diff": 0.0396,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2652.2284,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3054,
      "query_norm": 1.4089,
      "queue_k_norm": 1.455,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8865,
      "sent_len_1": 66.7807,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6875,
      "stdk": 0.0484,
      "stdq": 0.0458,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53300
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2503,
      "doc_norm": 1.4535,
      "encoder_q-embeddings": 2664.2913,
      "encoder_q-layer.0": 1682.5398,
      "encoder_q-layer.1": 1990.8265,
      "encoder_q-layer.10": 1276.0438,
      "encoder_q-layer.11": 2669.5229,
      "encoder_q-layer.2": 2200.8228,
      "encoder_q-layer.3": 2317.3169,
      "encoder_q-layer.4": 2563.3738,
      "encoder_q-layer.5": 2814.3625,
      "encoder_q-layer.6": 2937.0488,
      "encoder_q-layer.7": 2968.968,
      "encoder_q-layer.8": 2961.187,
      "encoder_q-layer.9": 1879.399,
      "epoch": 0.52,
      "inbatch_neg_score": 0.305,
      "inbatch_pos_score": 0.9751,
      "learning_rate": 2.588888888888889e-05,
      "loss": 3.2503,
      "norm_diff": 0.0395,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3696.9173,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3032,
      "query_norm": 1.414,
      "queue_k_norm": 1.4559,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0034,
      "sent_len_1": 66.8238,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2562,
      "stdk": 0.0486,
      "stdq": 0.046,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53400
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2516,
      "doc_norm": 1.4573,
      "encoder_q-embeddings": 679.1169,
      "encoder_q-layer.0": 465.4492,
      "encoder_q-layer.1": 519.6705,
      "encoder_q-layer.10": 609.7736,
      "encoder_q-layer.11": 1306.3805,
      "encoder_q-layer.2": 573.0762,
      "encoder_q-layer.3": 608.7589,
      "encoder_q-layer.4": 620.0797,
      "encoder_q-layer.5": 622.5662,
      "encoder_q-layer.6": 600.4145,
      "encoder_q-layer.7": 678.2234,
      "encoder_q-layer.8": 772.5187,
      "encoder_q-layer.9": 642.8572,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3083,
      "inbatch_pos_score": 1.0186,
      "learning_rate": 2.5833333333333336e-05,
      "loss": 3.2516,
      "norm_diff": 0.0174,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1046.9606,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3062,
      "query_norm": 1.4455,
      "queue_k_norm": 1.4563,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1461,
      "sent_len_1": 66.995,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.255,
      "stdk": 0.0487,
      "stdq": 0.0471,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53500
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.264,
      "doc_norm": 1.4515,
      "encoder_q-embeddings": 532.1332,
      "encoder_q-layer.0": 339.4643,
      "encoder_q-layer.1": 364.5318,
      "encoder_q-layer.10": 579.0833,
      "encoder_q-layer.11": 1255.015,
      "encoder_q-layer.2": 403.6624,
      "encoder_q-layer.3": 410.9844,
      "encoder_q-layer.4": 443.556,
      "encoder_q-layer.5": 467.5862,
      "encoder_q-layer.6": 542.5486,
      "encoder_q-layer.7": 614.436,
      "encoder_q-layer.8": 692.5894,
      "encoder_q-layer.9": 597.0106,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3095,
      "inbatch_pos_score": 0.9956,
      "learning_rate": 2.5777777777777778e-05,
      "loss": 3.264,
      "norm_diff": 0.0295,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 907.4439,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3105,
      "query_norm": 1.422,
      "queue_k_norm": 1.4564,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9817,
      "sent_len_1": 66.6722,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9313,
      "stdk": 0.0485,
      "stdq": 0.0461,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53600
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2855,
      "doc_norm": 1.4492,
      "encoder_q-embeddings": 978.6429,
      "encoder_q-layer.0": 664.0924,
      "encoder_q-layer.1": 663.9759,
      "encoder_q-layer.10": 596.2518,
      "encoder_q-layer.11": 1356.7678,
      "encoder_q-layer.2": 813.6729,
      "encoder_q-layer.3": 820.6057,
      "encoder_q-layer.4": 987.0893,
      "encoder_q-layer.5": 988.9365,
      "encoder_q-layer.6": 1206.7094,
      "encoder_q-layer.7": 1346.1925,
      "encoder_q-layer.8": 1388.4393,
      "encoder_q-layer.9": 1030.9988,
      "epoch": 0.52,
      "inbatch_neg_score": 0.3138,
      "inbatch_pos_score": 0.9497,
      "learning_rate": 2.5722222222222224e-05,
      "loss": 3.2855,
      "norm_diff": 0.0405,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1550.6922,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3142,
      "query_norm": 1.4088,
      "queue_k_norm": 1.4566,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8617,
      "sent_len_1": 66.857,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1062,
      "stdk": 0.0484,
      "stdq": 0.0455,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53700
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2475,
      "doc_norm": 1.4602,
      "encoder_q-embeddings": 518.2703,
      "encoder_q-layer.0": 344.2548,
      "encoder_q-layer.1": 355.8439,
      "encoder_q-layer.10": 613.1191,
      "encoder_q-layer.11": 1371.0914,
      "encoder_q-layer.2": 391.3366,
      "encoder_q-layer.3": 425.5739,
      "encoder_q-layer.4": 459.1058,
      "encoder_q-layer.5": 483.7762,
      "encoder_q-layer.6": 533.8994,
      "encoder_q-layer.7": 659.2462,
      "encoder_q-layer.8": 702.2338,
      "encoder_q-layer.9": 618.1033,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3086,
      "inbatch_pos_score": 0.9546,
      "learning_rate": 2.5666666666666666e-05,
      "loss": 3.2475,
      "norm_diff": 0.0689,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 960.9503,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3083,
      "query_norm": 1.3914,
      "queue_k_norm": 1.4588,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8782,
      "sent_len_1": 66.8841,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0163,
      "stdk": 0.0488,
      "stdq": 0.0449,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 53800
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2615,
      "doc_norm": 1.4474,
      "encoder_q-embeddings": 875.852,
      "encoder_q-layer.0": 589.3999,
      "encoder_q-layer.1": 707.1741,
      "encoder_q-layer.10": 627.925,
      "encoder_q-layer.11": 1272.8533,
      "encoder_q-layer.2": 802.228,
      "encoder_q-layer.3": 842.9493,
      "encoder_q-layer.4": 877.1678,
      "encoder_q-layer.5": 892.139,
      "encoder_q-layer.6": 957.5679,
      "encoder_q-layer.7": 1001.8243,
      "encoder_q-layer.8": 855.2566,
      "encoder_q-layer.9": 640.3227,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3069,
      "inbatch_pos_score": 0.9873,
      "learning_rate": 2.5611111111111115e-05,
      "loss": 3.2615,
      "norm_diff": 0.0274,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1268.4015,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3062,
      "query_norm": 1.42,
      "queue_k_norm": 1.4567,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9722,
      "sent_len_1": 66.6646,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6863,
      "stdk": 0.0483,
      "stdq": 0.0459,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 53900
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2351,
      "doc_norm": 1.4623,
      "encoder_q-embeddings": 650.297,
      "encoder_q-layer.0": 442.5269,
      "encoder_q-layer.1": 485.9955,
      "encoder_q-layer.10": 581.5555,
      "encoder_q-layer.11": 1290.7361,
      "encoder_q-layer.2": 523.8799,
      "encoder_q-layer.3": 533.4045,
      "encoder_q-layer.4": 596.726,
      "encoder_q-layer.5": 618.5184,
      "encoder_q-layer.6": 660.2542,
      "encoder_q-layer.7": 738.7957,
      "encoder_q-layer.8": 715.33,
      "encoder_q-layer.9": 586.7059,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3061,
      "inbatch_pos_score": 0.9785,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 3.2351,
      "norm_diff": 0.0534,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1028.0005,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3057,
      "query_norm": 1.4089,
      "queue_k_norm": 1.4573,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1747,
      "sent_len_1": 67.0577,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.32,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54000
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2615,
      "doc_norm": 1.4531,
      "encoder_q-embeddings": 549.6766,
      "encoder_q-layer.0": 349.2867,
      "encoder_q-layer.1": 379.9253,
      "encoder_q-layer.10": 597.7563,
      "encoder_q-layer.11": 1404.0586,
      "encoder_q-layer.2": 439.146,
      "encoder_q-layer.3": 484.714,
      "encoder_q-layer.4": 497.5535,
      "encoder_q-layer.5": 496.144,
      "encoder_q-layer.6": 595.456,
      "encoder_q-layer.7": 647.8751,
      "encoder_q-layer.8": 707.4572,
      "encoder_q-layer.9": 617.5822,
      "epoch": 0.53,
      "inbatch_neg_score": 0.312,
      "inbatch_pos_score": 0.9761,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 3.2615,
      "norm_diff": 0.0402,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 979.6412,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3123,
      "query_norm": 1.413,
      "queue_k_norm": 1.4563,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0144,
      "sent_len_1": 66.9094,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.88,
      "stdk": 0.0486,
      "stdq": 0.0457,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54100
    },
    {
      "accuracy": 59.8633,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2704,
      "doc_norm": 1.4629,
      "encoder_q-embeddings": 1223.2585,
      "encoder_q-layer.0": 812.2845,
      "encoder_q-layer.1": 958.4854,
      "encoder_q-layer.10": 617.2843,
      "encoder_q-layer.11": 1323.0123,
      "encoder_q-layer.2": 1127.2628,
      "encoder_q-layer.3": 1329.5723,
      "encoder_q-layer.4": 1505.9457,
      "encoder_q-layer.5": 1491.264,
      "encoder_q-layer.6": 1679.8549,
      "encoder_q-layer.7": 1439.2255,
      "encoder_q-layer.8": 977.5849,
      "encoder_q-layer.9": 631.3735,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3136,
      "inbatch_pos_score": 1.0449,
      "learning_rate": 2.5444444444444442e-05,
      "loss": 3.2704,
      "norm_diff": 0.0187,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1775.3638,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3135,
      "query_norm": 1.4455,
      "queue_k_norm": 1.4572,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9794,
      "sent_len_1": 66.7084,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4625,
      "stdk": 0.049,
      "stdq": 0.0469,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54200
    },
    {
      "accuracy": 56.9336,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2603,
      "doc_norm": 1.4638,
      "encoder_q-embeddings": 527.3731,
      "encoder_q-layer.0": 348.3551,
      "encoder_q-layer.1": 373.1879,
      "encoder_q-layer.10": 566.4696,
      "encoder_q-layer.11": 1251.1072,
      "encoder_q-layer.2": 440.6123,
      "encoder_q-layer.3": 447.9819,
      "encoder_q-layer.4": 445.7251,
      "encoder_q-layer.5": 440.5097,
      "encoder_q-layer.6": 488.3738,
      "encoder_q-layer.7": 553.5131,
      "encoder_q-layer.8": 651.3451,
      "encoder_q-layer.9": 576.3845,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3145,
      "inbatch_pos_score": 1.0117,
      "learning_rate": 2.538888888888889e-05,
      "loss": 3.2603,
      "norm_diff": 0.0472,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 902.265,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3149,
      "query_norm": 1.4166,
      "queue_k_norm": 1.4567,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0504,
      "sent_len_1": 66.9104,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8025,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54300
    },
    {
      "accuracy": 59.8633,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2337,
      "doc_norm": 1.4575,
      "encoder_q-embeddings": 561.8256,
      "encoder_q-layer.0": 380.8225,
      "encoder_q-layer.1": 404.3314,
      "encoder_q-layer.10": 566.2388,
      "encoder_q-layer.11": 1287.0874,
      "encoder_q-layer.2": 482.0276,
      "encoder_q-layer.3": 498.191,
      "encoder_q-layer.4": 540.1487,
      "encoder_q-layer.5": 563.5172,
      "encoder_q-layer.6": 565.3632,
      "encoder_q-layer.7": 621.2231,
      "encoder_q-layer.8": 704.9969,
      "encoder_q-layer.9": 602.4878,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3181,
      "inbatch_pos_score": 1.0527,
      "learning_rate": 2.5333333333333337e-05,
      "loss": 3.2337,
      "norm_diff": 0.0197,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 952.2864,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3181,
      "query_norm": 1.4378,
      "queue_k_norm": 1.4581,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9668,
      "sent_len_1": 66.777,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6337,
      "stdk": 0.0487,
      "stdq": 0.0467,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54400
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2583,
      "doc_norm": 1.4588,
      "encoder_q-embeddings": 515.7585,
      "encoder_q-layer.0": 332.3135,
      "encoder_q-layer.1": 344.2042,
      "encoder_q-layer.10": 610.6855,
      "encoder_q-layer.11": 1274.7162,
      "encoder_q-layer.2": 403.2712,
      "encoder_q-layer.3": 417.4149,
      "encoder_q-layer.4": 453.1838,
      "encoder_q-layer.5": 456.5516,
      "encoder_q-layer.6": 519.8853,
      "encoder_q-layer.7": 634.5446,
      "encoder_q-layer.8": 672.9125,
      "encoder_q-layer.9": 581.3373,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3202,
      "inbatch_pos_score": 1.0088,
      "learning_rate": 2.527777777777778e-05,
      "loss": 3.2583,
      "norm_diff": 0.0461,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 911.3234,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3203,
      "query_norm": 1.4127,
      "queue_k_norm": 1.4586,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0535,
      "sent_len_1": 67.0125,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7862,
      "stdk": 0.0488,
      "stdq": 0.0457,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54500
    },
    {
      "accuracy": 58.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 3.243,
      "doc_norm": 1.4581,
      "encoder_q-embeddings": 815.7733,
      "encoder_q-layer.0": 529.7228,
      "encoder_q-layer.1": 610.9584,
      "encoder_q-layer.10": 563.5249,
      "encoder_q-layer.11": 1247.4277,
      "encoder_q-layer.2": 718.024,
      "encoder_q-layer.3": 757.3364,
      "encoder_q-layer.4": 791.0568,
      "encoder_q-layer.5": 823.9933,
      "encoder_q-layer.6": 730.6739,
      "encoder_q-layer.7": 698.1954,
      "encoder_q-layer.8": 706.26,
      "encoder_q-layer.9": 584.9048,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3247,
      "inbatch_pos_score": 1.0283,
      "learning_rate": 2.5222222222222225e-05,
      "loss": 3.243,
      "norm_diff": 0.0357,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1143.8077,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3252,
      "query_norm": 1.4224,
      "queue_k_norm": 1.458,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9519,
      "sent_len_1": 67.0095,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.065,
      "stdk": 0.0487,
      "stdq": 0.0462,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54600
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2385,
      "doc_norm": 1.4587,
      "encoder_q-embeddings": 691.7759,
      "encoder_q-layer.0": 468.3908,
      "encoder_q-layer.1": 538.395,
      "encoder_q-layer.10": 585.0952,
      "encoder_q-layer.11": 1337.7781,
      "encoder_q-layer.2": 592.1096,
      "encoder_q-layer.3": 610.0435,
      "encoder_q-layer.4": 673.6176,
      "encoder_q-layer.5": 665.4385,
      "encoder_q-layer.6": 695.9367,
      "encoder_q-layer.7": 738.8861,
      "encoder_q-layer.8": 737.2371,
      "encoder_q-layer.9": 624.432,
      "epoch": 0.53,
      "inbatch_neg_score": 0.3288,
      "inbatch_pos_score": 0.9961,
      "learning_rate": 2.5166666666666667e-05,
      "loss": 3.2385,
      "norm_diff": 0.0438,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1069.2969,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3281,
      "query_norm": 1.4149,
      "queue_k_norm": 1.4594,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0694,
      "sent_len_1": 66.8877,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1113,
      "stdk": 0.0487,
      "stdq": 0.0459,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54700
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.234,
      "doc_norm": 1.4514,
      "encoder_q-embeddings": 657.139,
      "encoder_q-layer.0": 428.9868,
      "encoder_q-layer.1": 474.3978,
      "encoder_q-layer.10": 613.6863,
      "encoder_q-layer.11": 1414.6897,
      "encoder_q-layer.2": 541.6478,
      "encoder_q-layer.3": 559.39,
      "encoder_q-layer.4": 596.8987,
      "encoder_q-layer.5": 634.2571,
      "encoder_q-layer.6": 648.9056,
      "encoder_q-layer.7": 675.4987,
      "encoder_q-layer.8": 752.1203,
      "encoder_q-layer.9": 656.2859,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3333,
      "inbatch_pos_score": 1.0078,
      "learning_rate": 2.5111111111111113e-05,
      "loss": 3.234,
      "norm_diff": 0.0217,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1051.717,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3337,
      "query_norm": 1.4296,
      "queue_k_norm": 1.4603,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9689,
      "sent_len_1": 66.8485,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7287,
      "stdk": 0.0484,
      "stdq": 0.0463,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 54800
    },
    {
      "accuracy": 49.707,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2684,
      "doc_norm": 1.4626,
      "encoder_q-embeddings": 1019.385,
      "encoder_q-layer.0": 689.5947,
      "encoder_q-layer.1": 746.9528,
      "encoder_q-layer.10": 610.3276,
      "encoder_q-layer.11": 1420.1444,
      "encoder_q-layer.2": 849.4861,
      "encoder_q-layer.3": 882.6152,
      "encoder_q-layer.4": 957.4644,
      "encoder_q-layer.5": 1002.7396,
      "encoder_q-layer.6": 1097.0497,
      "encoder_q-layer.7": 1033.2451,
      "encoder_q-layer.8": 1057.6633,
      "encoder_q-layer.9": 709.7014,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3299,
      "inbatch_pos_score": 0.9673,
      "learning_rate": 2.5055555555555555e-05,
      "loss": 3.2684,
      "norm_diff": 0.0582,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1451.6316,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3306,
      "query_norm": 1.4044,
      "queue_k_norm": 1.4617,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7743,
      "sent_len_1": 66.7653,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4888,
      "stdk": 0.0488,
      "stdq": 0.0454,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 54900
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2528,
      "doc_norm": 1.4647,
      "encoder_q-embeddings": 706.4089,
      "encoder_q-layer.0": 472.3781,
      "encoder_q-layer.1": 528.5988,
      "encoder_q-layer.10": 571.3318,
      "encoder_q-layer.11": 1289.877,
      "encoder_q-layer.2": 622.8236,
      "encoder_q-layer.3": 635.8906,
      "encoder_q-layer.4": 628.5106,
      "encoder_q-layer.5": 626.7506,
      "encoder_q-layer.6": 642.0452,
      "encoder_q-layer.7": 664.2176,
      "encoder_q-layer.8": 708.6227,
      "encoder_q-layer.9": 578.4404,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3312,
      "inbatch_pos_score": 1.0361,
      "learning_rate": 2.5e-05,
      "loss": 3.2528,
      "norm_diff": 0.0415,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1049.1153,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.332,
      "query_norm": 1.4232,
      "queue_k_norm": 1.4609,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9241,
      "sent_len_1": 66.6066,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6087,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 55000
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2409,
      "doc_norm": 1.4663,
      "encoder_q-embeddings": 552.5273,
      "encoder_q-layer.0": 354.8134,
      "encoder_q-layer.1": 370.6664,
      "encoder_q-layer.10": 652.8088,
      "encoder_q-layer.11": 1386.8669,
      "encoder_q-layer.2": 431.3096,
      "encoder_q-layer.3": 457.5678,
      "encoder_q-layer.4": 485.9171,
      "encoder_q-layer.5": 507.1512,
      "encoder_q-layer.6": 559.1821,
      "encoder_q-layer.7": 650.7742,
      "encoder_q-layer.8": 731.1226,
      "encoder_q-layer.9": 659.0688,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3332,
      "inbatch_pos_score": 1.0146,
      "learning_rate": 2.4944444444444447e-05,
      "loss": 3.2409,
      "norm_diff": 0.0579,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 977.4664,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.334,
      "query_norm": 1.4084,
      "queue_k_norm": 1.46,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8752,
      "sent_len_1": 66.6037,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5037,
      "stdk": 0.0489,
      "stdq": 0.0454,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 55100
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2534,
      "doc_norm": 1.4579,
      "encoder_q-embeddings": 605.8903,
      "encoder_q-layer.0": 399.2472,
      "encoder_q-layer.1": 433.6851,
      "encoder_q-layer.10": 625.2964,
      "encoder_q-layer.11": 1312.9022,
      "encoder_q-layer.2": 489.0856,
      "encoder_q-layer.3": 540.7819,
      "encoder_q-layer.4": 572.4464,
      "encoder_q-layer.5": 604.6945,
      "encoder_q-layer.6": 659.9494,
      "encoder_q-layer.7": 702.3077,
      "encoder_q-layer.8": 710.2654,
      "encoder_q-layer.9": 612.6099,
      "epoch": 0.54,
      "inbatch_neg_score": 0.338,
      "inbatch_pos_score": 1.0166,
      "learning_rate": 2.488888888888889e-05,
      "loss": 3.2534,
      "norm_diff": 0.0443,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1001.4008,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3372,
      "query_norm": 1.4137,
      "queue_k_norm": 1.4628,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0638,
      "sent_len_1": 66.8353,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.095,
      "stdk": 0.0486,
      "stdq": 0.0455,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 55200
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2419,
      "doc_norm": 1.4686,
      "encoder_q-embeddings": 604.7387,
      "encoder_q-layer.0": 396.3202,
      "encoder_q-layer.1": 445.6085,
      "encoder_q-layer.10": 623.955,
      "encoder_q-layer.11": 1393.0629,
      "encoder_q-layer.2": 534.3174,
      "encoder_q-layer.3": 533.3201,
      "encoder_q-layer.4": 534.2905,
      "encoder_q-layer.5": 562.9941,
      "encoder_q-layer.6": 579.3804,
      "encoder_q-layer.7": 619.6907,
      "encoder_q-layer.8": 695.8157,
      "encoder_q-layer.9": 610.1759,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3383,
      "inbatch_pos_score": 0.9932,
      "learning_rate": 2.4833333333333335e-05,
      "loss": 3.2419,
      "norm_diff": 0.0414,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1009.7636,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3379,
      "query_norm": 1.4272,
      "queue_k_norm": 1.4653,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2167,
      "sent_len_1": 66.997,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4162,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 55300
    },
    {
      "accuracy": 57.3242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2253,
      "doc_norm": 1.4706,
      "encoder_q-embeddings": 761.5033,
      "encoder_q-layer.0": 514.7718,
      "encoder_q-layer.1": 568.1295,
      "encoder_q-layer.10": 611.9531,
      "encoder_q-layer.11": 1348.5198,
      "encoder_q-layer.2": 648.4449,
      "encoder_q-layer.3": 665.0541,
      "encoder_q-layer.4": 748.3412,
      "encoder_q-layer.5": 818.2358,
      "encoder_q-layer.6": 804.3041,
      "encoder_q-layer.7": 824.9991,
      "encoder_q-layer.8": 818.051,
      "encoder_q-layer.9": 662.1336,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3374,
      "inbatch_pos_score": 1.0439,
      "learning_rate": 2.477777777777778e-05,
      "loss": 3.2253,
      "norm_diff": 0.0352,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1168.4945,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3374,
      "query_norm": 1.4355,
      "queue_k_norm": 1.4662,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1988,
      "sent_len_1": 66.9594,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5087,
      "stdk": 0.049,
      "stdq": 0.0464,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 55400
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2539,
      "doc_norm": 1.4596,
      "encoder_q-embeddings": 1113.8564,
      "encoder_q-layer.0": 691.0915,
      "encoder_q-layer.1": 765.1619,
      "encoder_q-layer.10": 1328.6445,
      "encoder_q-layer.11": 2629.5701,
      "encoder_q-layer.2": 868.0745,
      "encoder_q-layer.3": 904.5007,
      "encoder_q-layer.4": 1055.4968,
      "encoder_q-layer.5": 1085.4646,
      "encoder_q-layer.6": 1191.2274,
      "encoder_q-layer.7": 1353.3383,
      "encoder_q-layer.8": 1532.7472,
      "encoder_q-layer.9": 1260.0905,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3382,
      "inbatch_pos_score": 1.0234,
      "learning_rate": 2.4722222222222223e-05,
      "loss": 3.2539,
      "norm_diff": 0.0341,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1919.0622,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3386,
      "query_norm": 1.4254,
      "queue_k_norm": 1.4655,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9384,
      "sent_len_1": 66.7859,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.065,
      "stdk": 0.0486,
      "stdq": 0.046,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 55500
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2218,
      "doc_norm": 1.4635,
      "encoder_q-embeddings": 1076.5718,
      "encoder_q-layer.0": 724.0466,
      "encoder_q-layer.1": 809.8141,
      "encoder_q-layer.10": 1147.8977,
      "encoder_q-layer.11": 2659.3982,
      "encoder_q-layer.2": 907.1464,
      "encoder_q-layer.3": 973.7999,
      "encoder_q-layer.4": 997.3129,
      "encoder_q-layer.5": 1053.1294,
      "encoder_q-layer.6": 1183.2086,
      "encoder_q-layer.7": 1301.6,
      "encoder_q-layer.8": 1415.193,
      "encoder_q-layer.9": 1259.3058,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3392,
      "inbatch_pos_score": 1.0264,
      "learning_rate": 2.466666666666667e-05,
      "loss": 3.2218,
      "norm_diff": 0.0335,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1951.9241,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3396,
      "query_norm": 1.43,
      "queue_k_norm": 1.4663,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8371,
      "sent_len_1": 66.8118,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1387,
      "stdk": 0.0487,
      "stdq": 0.0461,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 55600
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2619,
      "doc_norm": 1.4663,
      "encoder_q-embeddings": 1158.2615,
      "encoder_q-layer.0": 772.6184,
      "encoder_q-layer.1": 860.2629,
      "encoder_q-layer.10": 1223.9451,
      "encoder_q-layer.11": 2740.335,
      "encoder_q-layer.2": 959.3987,
      "encoder_q-layer.3": 1035.8611,
      "encoder_q-layer.4": 1092.4092,
      "encoder_q-layer.5": 1155.2024,
      "encoder_q-layer.6": 1288.3085,
      "encoder_q-layer.7": 1454.7865,
      "encoder_q-layer.8": 1492.7397,
      "encoder_q-layer.9": 1297.1665,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3454,
      "inbatch_pos_score": 1.0059,
      "learning_rate": 2.461111111111111e-05,
      "loss": 3.2619,
      "norm_diff": 0.0361,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2028.941,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3447,
      "query_norm": 1.4302,
      "queue_k_norm": 1.4652,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7971,
      "sent_len_1": 66.385,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2912,
      "stdk": 0.0488,
      "stdq": 0.046,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 55700
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2371,
      "doc_norm": 1.4674,
      "encoder_q-embeddings": 1214.3668,
      "encoder_q-layer.0": 774.3855,
      "encoder_q-layer.1": 867.9693,
      "encoder_q-layer.10": 1221.4121,
      "encoder_q-layer.11": 2695.4668,
      "encoder_q-layer.2": 1015.8002,
      "encoder_q-layer.3": 1072.7611,
      "encoder_q-layer.4": 1164.3916,
      "encoder_q-layer.5": 1224.7582,
      "encoder_q-layer.6": 1244.7915,
      "encoder_q-layer.7": 1541.5612,
      "encoder_q-layer.8": 1601.4204,
      "encoder_q-layer.9": 1288.6572,
      "epoch": 0.54,
      "inbatch_neg_score": 0.3461,
      "inbatch_pos_score": 1.0234,
      "learning_rate": 2.4555555555555557e-05,
      "loss": 3.2371,
      "norm_diff": 0.0339,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2069.9696,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3462,
      "query_norm": 1.4336,
      "queue_k_norm": 1.4677,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9736,
      "sent_len_1": 66.8878,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.2562,
      "stdk": 0.0488,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 55800
    },
    {
      "accuracy": 55.1758,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2729,
      "doc_norm": 1.4699,
      "encoder_q-embeddings": 1553.7365,
      "encoder_q-layer.0": 1056.7343,
      "encoder_q-layer.1": 1262.2657,
      "encoder_q-layer.10": 1257.2861,
      "encoder_q-layer.11": 2634.0977,
      "encoder_q-layer.2": 1579.9639,
      "encoder_q-layer.3": 1606.0625,
      "encoder_q-layer.4": 1892.3091,
      "encoder_q-layer.5": 2129.9036,
      "encoder_q-layer.6": 2344.4836,
      "encoder_q-layer.7": 2306.573,
      "encoder_q-layer.8": 2512.6797,
      "encoder_q-layer.9": 1673.5944,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3431,
      "inbatch_pos_score": 1.0,
      "learning_rate": 2.45e-05,
      "loss": 3.2729,
      "norm_diff": 0.0653,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2785.8197,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3428,
      "query_norm": 1.4046,
      "queue_k_norm": 1.4664,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0716,
      "sent_len_1": 66.7217,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5762,
      "stdk": 0.0488,
      "stdq": 0.0452,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 55900
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.245,
      "doc_norm": 1.4685,
      "encoder_q-embeddings": 1941.3864,
      "encoder_q-layer.0": 1392.0302,
      "encoder_q-layer.1": 1502.5179,
      "encoder_q-layer.10": 1241.9531,
      "encoder_q-layer.11": 2620.4133,
      "encoder_q-layer.2": 1934.3741,
      "encoder_q-layer.3": 1985.3712,
      "encoder_q-layer.4": 2089.7507,
      "encoder_q-layer.5": 1979.4423,
      "encoder_q-layer.6": 1786.5112,
      "encoder_q-layer.7": 1956.6042,
      "encoder_q-layer.8": 1948.205,
      "encoder_q-layer.9": 1381.2026,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3496,
      "inbatch_pos_score": 1.0166,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 3.245,
      "norm_diff": 0.0638,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2816.2145,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3489,
      "query_norm": 1.4047,
      "queue_k_norm": 1.4688,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1076,
      "sent_len_1": 66.8001,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5125,
      "stdk": 0.0488,
      "stdq": 0.045,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56000
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2302,
      "doc_norm": 1.4699,
      "encoder_q-embeddings": 1932.437,
      "encoder_q-layer.0": 1331.4807,
      "encoder_q-layer.1": 1426.1594,
      "encoder_q-layer.10": 1218.0751,
      "encoder_q-layer.11": 2637.4973,
      "encoder_q-layer.2": 1699.9277,
      "encoder_q-layer.3": 1497.6577,
      "encoder_q-layer.4": 1475.2361,
      "encoder_q-layer.5": 1453.0848,
      "encoder_q-layer.6": 1614.3583,
      "encoder_q-layer.7": 1646.2424,
      "encoder_q-layer.8": 1500.7961,
      "encoder_q-layer.9": 1292.0261,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3515,
      "inbatch_pos_score": 1.0176,
      "learning_rate": 2.4388888888888887e-05,
      "loss": 3.2302,
      "norm_diff": 0.0476,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2485.5265,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3516,
      "query_norm": 1.4224,
      "queue_k_norm": 1.4687,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0403,
      "sent_len_1": 66.9177,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.845,
      "stdk": 0.0488,
      "stdq": 0.0456,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56100
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.285,
      "doc_norm": 1.4631,
      "encoder_q-embeddings": 2873.1699,
      "encoder_q-layer.0": 1884.551,
      "encoder_q-layer.1": 2305.9473,
      "encoder_q-layer.10": 1245.3198,
      "encoder_q-layer.11": 2764.3979,
      "encoder_q-layer.2": 2905.4241,
      "encoder_q-layer.3": 2849.0935,
      "encoder_q-layer.4": 2837.6611,
      "encoder_q-layer.5": 3001.2612,
      "encoder_q-layer.6": 3488.2278,
      "encoder_q-layer.7": 2959.9648,
      "encoder_q-layer.8": 2173.813,
      "encoder_q-layer.9": 1216.2323,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3532,
      "inbatch_pos_score": 1.0244,
      "learning_rate": 2.4333333333333336e-05,
      "loss": 3.285,
      "norm_diff": 0.0354,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3892.8125,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.353,
      "query_norm": 1.4277,
      "queue_k_norm": 1.4706,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7505,
      "sent_len_1": 66.6796,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.39,
      "stdk": 0.0485,
      "stdq": 0.0459,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56200
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2536,
      "doc_norm": 1.4674,
      "encoder_q-embeddings": 1937.3875,
      "encoder_q-layer.0": 1347.7516,
      "encoder_q-layer.1": 1551.9657,
      "encoder_q-layer.10": 1190.6985,
      "encoder_q-layer.11": 2704.7681,
      "encoder_q-layer.2": 1754.1162,
      "encoder_q-layer.3": 2020.4833,
      "encoder_q-layer.4": 2139.5757,
      "encoder_q-layer.5": 2382.2102,
      "encoder_q-layer.6": 2714.1995,
      "encoder_q-layer.7": 3217.2952,
      "encoder_q-layer.8": 2676.5774,
      "encoder_q-layer.9": 1323.2303,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3522,
      "inbatch_pos_score": 1.0293,
      "learning_rate": 2.427777777777778e-05,
      "loss": 3.2536,
      "norm_diff": 0.0346,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3183.0784,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3528,
      "query_norm": 1.4328,
      "queue_k_norm": 1.4692,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9175,
      "sent_len_1": 66.7791,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0263,
      "stdk": 0.0487,
      "stdq": 0.0461,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 56300
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2445,
      "doc_norm": 1.4707,
      "encoder_q-embeddings": 1114.9711,
      "encoder_q-layer.0": 696.4008,
      "encoder_q-layer.1": 770.7186,
      "encoder_q-layer.10": 1201.3453,
      "encoder_q-layer.11": 2769.646,
      "encoder_q-layer.2": 920.1508,
      "encoder_q-layer.3": 948.6283,
      "encoder_q-layer.4": 1023.6074,
      "encoder_q-layer.5": 1050.1355,
      "encoder_q-layer.6": 1200.1552,
      "encoder_q-layer.7": 1213.83,
      "encoder_q-layer.8": 1397.7942,
      "encoder_q-layer.9": 1224.0399,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3521,
      "inbatch_pos_score": 1.0059,
      "learning_rate": 2.4222222222222224e-05,
      "loss": 3.2445,
      "norm_diff": 0.0527,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1981.2985,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.353,
      "query_norm": 1.418,
      "queue_k_norm": 1.4706,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8415,
      "sent_len_1": 66.5935,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4512,
      "stdk": 0.0488,
      "stdq": 0.0455,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56400
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2412,
      "doc_norm": 1.4785,
      "encoder_q-embeddings": 1165.6338,
      "encoder_q-layer.0": 705.7196,
      "encoder_q-layer.1": 746.6288,
      "encoder_q-layer.10": 1233.3812,
      "encoder_q-layer.11": 2798.0444,
      "encoder_q-layer.2": 816.2272,
      "encoder_q-layer.3": 904.0092,
      "encoder_q-layer.4": 928.2448,
      "encoder_q-layer.5": 990.3612,
      "encoder_q-layer.6": 1103.5435,
      "encoder_q-layer.7": 1252.2617,
      "encoder_q-layer.8": 1427.126,
      "encoder_q-layer.9": 1239.7184,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3555,
      "inbatch_pos_score": 1.0537,
      "learning_rate": 2.4166666666666667e-05,
      "loss": 3.2412,
      "norm_diff": 0.0333,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1967.5881,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.355,
      "query_norm": 1.4451,
      "queue_k_norm": 1.4692,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9538,
      "sent_len_1": 66.5102,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7825,
      "stdk": 0.049,
      "stdq": 0.0466,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 56500
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.226,
      "doc_norm": 1.4694,
      "encoder_q-embeddings": 1471.0878,
      "encoder_q-layer.0": 943.5052,
      "encoder_q-layer.1": 1096.6223,
      "encoder_q-layer.10": 1203.8684,
      "encoder_q-layer.11": 2615.9224,
      "encoder_q-layer.2": 1292.9255,
      "encoder_q-layer.3": 1445.2753,
      "encoder_q-layer.4": 1714.2607,
      "encoder_q-layer.5": 1766.5958,
      "encoder_q-layer.6": 1883.7856,
      "encoder_q-layer.7": 2075.2065,
      "encoder_q-layer.8": 2102.5415,
      "encoder_q-layer.9": 1484.486,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3575,
      "inbatch_pos_score": 1.0234,
      "learning_rate": 2.4111111111111113e-05,
      "loss": 3.226,
      "norm_diff": 0.0509,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2538.3023,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3577,
      "query_norm": 1.4185,
      "queue_k_norm": 1.4717,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0624,
      "sent_len_1": 66.7981,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.105,
      "stdk": 0.0487,
      "stdq": 0.0455,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56600
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2211,
      "doc_norm": 1.4721,
      "encoder_q-embeddings": 1889.5103,
      "encoder_q-layer.0": 1294.5557,
      "encoder_q-layer.1": 1542.1815,
      "encoder_q-layer.10": 1257.0962,
      "encoder_q-layer.11": 2835.8401,
      "encoder_q-layer.2": 1802.87,
      "encoder_q-layer.3": 1774.9858,
      "encoder_q-layer.4": 1897.3698,
      "encoder_q-layer.5": 2241.6592,
      "encoder_q-layer.6": 2451.1428,
      "encoder_q-layer.7": 2836.6235,
      "encoder_q-layer.8": 2697.3149,
      "encoder_q-layer.9": 1777.7114,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3583,
      "inbatch_pos_score": 1.043,
      "learning_rate": 2.4055555555555555e-05,
      "loss": 3.2211,
      "norm_diff": 0.0314,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3141.3918,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3589,
      "query_norm": 1.4407,
      "queue_k_norm": 1.4731,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0373,
      "sent_len_1": 66.7911,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7063,
      "stdk": 0.0488,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56700
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2362,
      "doc_norm": 1.4671,
      "encoder_q-embeddings": 1022.0084,
      "encoder_q-layer.0": 692.3594,
      "encoder_q-layer.1": 706.1815,
      "encoder_q-layer.10": 1114.4156,
      "encoder_q-layer.11": 2529.1321,
      "encoder_q-layer.2": 805.1382,
      "encoder_q-layer.3": 831.0999,
      "encoder_q-layer.4": 901.4658,
      "encoder_q-layer.5": 916.0894,
      "encoder_q-layer.6": 1032.7516,
      "encoder_q-layer.7": 1185.2068,
      "encoder_q-layer.8": 1344.0674,
      "encoder_q-layer.9": 1157.3574,
      "epoch": 0.55,
      "inbatch_neg_score": 0.3622,
      "inbatch_pos_score": 1.0029,
      "learning_rate": 2.4e-05,
      "loss": 3.2362,
      "norm_diff": 0.0452,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1803.9718,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3625,
      "query_norm": 1.4219,
      "queue_k_norm": 1.472,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0519,
      "sent_len_1": 66.7586,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.95,
      "stdk": 0.0486,
      "stdq": 0.0454,
      "stdqueue_k": 0.0488,
      "stdqueue_q": 0.0,
      "step": 56800
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2309,
      "doc_norm": 1.4739,
      "encoder_q-embeddings": 1585.8496,
      "encoder_q-layer.0": 1070.0491,
      "encoder_q-layer.1": 1150.5988,
      "encoder_q-layer.10": 1192.8466,
      "encoder_q-layer.11": 2553.2424,
      "encoder_q-layer.2": 1374.6466,
      "encoder_q-layer.3": 1394.5096,
      "encoder_q-layer.4": 1543.5957,
      "encoder_q-layer.5": 1528.7664,
      "encoder_q-layer.6": 1637.9849,
      "encoder_q-layer.7": 1781.6418,
      "encoder_q-layer.8": 1616.5585,
      "encoder_q-layer.9": 1202.2461,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3612,
      "inbatch_pos_score": 1.0117,
      "learning_rate": 2.3944444444444443e-05,
      "loss": 3.2309,
      "norm_diff": 0.0636,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2346.0515,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3604,
      "query_norm": 1.4103,
      "queue_k_norm": 1.4728,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2413,
      "sent_len_1": 66.9302,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0913,
      "stdk": 0.0488,
      "stdq": 0.0448,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 56900
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2616,
      "doc_norm": 1.4701,
      "encoder_q-embeddings": 1990.4347,
      "encoder_q-layer.0": 1418.0406,
      "encoder_q-layer.1": 1498.2249,
      "encoder_q-layer.10": 1191.0461,
      "encoder_q-layer.11": 2716.2476,
      "encoder_q-layer.2": 1675.337,
      "encoder_q-layer.3": 1690.751,
      "encoder_q-layer.4": 1829.911,
      "encoder_q-layer.5": 1858.556,
      "encoder_q-layer.6": 2055.2085,
      "encoder_q-layer.7": 2017.7617,
      "encoder_q-layer.8": 1873.3562,
      "encoder_q-layer.9": 1345.9382,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3613,
      "inbatch_pos_score": 1.0078,
      "learning_rate": 2.3888888888888892e-05,
      "loss": 3.2616,
      "norm_diff": 0.057,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2762.3642,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3608,
      "query_norm": 1.4131,
      "queue_k_norm": 1.4751,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7819,
      "sent_len_1": 66.763,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.88,
      "stdk": 0.0486,
      "stdq": 0.0449,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 57000
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2021,
      "doc_norm": 1.4722,
      "encoder_q-embeddings": 1882.9877,
      "encoder_q-layer.0": 1342.3016,
      "encoder_q-layer.1": 1435.2957,
      "encoder_q-layer.10": 1259.845,
      "encoder_q-layer.11": 2448.5249,
      "encoder_q-layer.2": 1665.7238,
      "encoder_q-layer.3": 1803.7357,
      "encoder_q-layer.4": 1995.1848,
      "encoder_q-layer.5": 2028.9752,
      "encoder_q-layer.6": 1976.2976,
      "encoder_q-layer.7": 2113.6604,
      "encoder_q-layer.8": 2044.2844,
      "encoder_q-layer.9": 1319.6027,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3635,
      "inbatch_pos_score": 1.0469,
      "learning_rate": 2.3833333333333334e-05,
      "loss": 3.2021,
      "norm_diff": 0.036,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2767.1173,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3635,
      "query_norm": 1.4363,
      "queue_k_norm": 1.4747,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9079,
      "sent_len_1": 66.5744,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6513,
      "stdk": 0.0487,
      "stdq": 0.0457,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 57100
    },
    {
      "accuracy": 58.9844,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2662,
      "doc_norm": 1.4785,
      "encoder_q-embeddings": 1689.5511,
      "encoder_q-layer.0": 1115.2408,
      "encoder_q-layer.1": 1284.0946,
      "encoder_q-layer.10": 1259.921,
      "encoder_q-layer.11": 2610.3503,
      "encoder_q-layer.2": 1464.1357,
      "encoder_q-layer.3": 1479.0961,
      "encoder_q-layer.4": 1553.5214,
      "encoder_q-layer.5": 1764.0682,
      "encoder_q-layer.6": 1847.0388,
      "encoder_q-layer.7": 1997.9785,
      "encoder_q-layer.8": 1867.5741,
      "encoder_q-layer.9": 1345.8416,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3679,
      "inbatch_pos_score": 1.0781,
      "learning_rate": 2.377777777777778e-05,
      "loss": 3.2662,
      "norm_diff": 0.0306,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2500.8051,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3674,
      "query_norm": 1.4478,
      "queue_k_norm": 1.4749,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8777,
      "sent_len_1": 66.5608,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7587,
      "stdk": 0.0489,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 57200
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2094,
      "doc_norm": 1.4762,
      "encoder_q-embeddings": 1398.6498,
      "encoder_q-layer.0": 967.5212,
      "encoder_q-layer.1": 1024.116,
      "encoder_q-layer.10": 1236.739,
      "encoder_q-layer.11": 2643.6528,
      "encoder_q-layer.2": 1221.7419,
      "encoder_q-layer.3": 1220.7189,
      "encoder_q-layer.4": 1344.1329,
      "encoder_q-layer.5": 1256.6411,
      "encoder_q-layer.6": 1434.6155,
      "encoder_q-layer.7": 1670.2131,
      "encoder_q-layer.8": 1798.4407,
      "encoder_q-layer.9": 1404.6104,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3678,
      "inbatch_pos_score": 1.0195,
      "learning_rate": 2.3722222222222222e-05,
      "loss": 3.2094,
      "norm_diff": 0.0458,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2260.7453,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3682,
      "query_norm": 1.4305,
      "queue_k_norm": 1.4759,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8131,
      "sent_len_1": 66.6776,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6788,
      "stdk": 0.0488,
      "stdq": 0.0454,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 57300
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2339,
      "doc_norm": 1.4811,
      "encoder_q-embeddings": 1406.8981,
      "encoder_q-layer.0": 940.3824,
      "encoder_q-layer.1": 1032.658,
      "encoder_q-layer.10": 1105.8925,
      "encoder_q-layer.11": 2628.7004,
      "encoder_q-layer.2": 1269.3998,
      "encoder_q-layer.3": 1275.3231,
      "encoder_q-layer.4": 1400.9695,
      "encoder_q-layer.5": 1297.3739,
      "encoder_q-layer.6": 1451.3796,
      "encoder_q-layer.7": 1428.373,
      "encoder_q-layer.8": 1495.0306,
      "encoder_q-layer.9": 1221.3386,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3746,
      "inbatch_pos_score": 1.0117,
      "learning_rate": 2.3666666666666668e-05,
      "loss": 3.2339,
      "norm_diff": 0.0682,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2169.1873,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3735,
      "query_norm": 1.4129,
      "queue_k_norm": 1.4781,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8301,
      "sent_len_1": 66.879,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3525,
      "stdk": 0.049,
      "stdq": 0.0448,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 57400
    },
    {
      "accuracy": 53.0273,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2502,
      "doc_norm": 1.4793,
      "encoder_q-embeddings": 2560.502,
      "encoder_q-layer.0": 1708.2919,
      "encoder_q-layer.1": 1908.1431,
      "encoder_q-layer.10": 2291.0767,
      "encoder_q-layer.11": 5382.4878,
      "encoder_q-layer.2": 2125.9587,
      "encoder_q-layer.3": 2118.2417,
      "encoder_q-layer.4": 2224.906,
      "encoder_q-layer.5": 2344.0801,
      "encoder_q-layer.6": 2548.8296,
      "encoder_q-layer.7": 2629.7319,
      "encoder_q-layer.8": 2846.8865,
      "encoder_q-layer.9": 2426.5562,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3737,
      "inbatch_pos_score": 1.0215,
      "learning_rate": 2.361111111111111e-05,
      "loss": 3.2502,
      "norm_diff": 0.0611,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4040.1044,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3728,
      "query_norm": 1.4182,
      "queue_k_norm": 1.4788,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8497,
      "sent_len_1": 66.8904,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5775,
      "stdk": 0.0489,
      "stdq": 0.045,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 57500
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2333,
      "doc_norm": 1.4802,
      "encoder_q-embeddings": 2829.7754,
      "encoder_q-layer.0": 1844.2151,
      "encoder_q-layer.1": 2227.9456,
      "encoder_q-layer.10": 2534.1079,
      "encoder_q-layer.11": 5530.6963,
      "encoder_q-layer.2": 2585.2915,
      "encoder_q-layer.3": 2710.5222,
      "encoder_q-layer.4": 2961.1777,
      "encoder_q-layer.5": 2924.7153,
      "encoder_q-layer.6": 3041.9534,
      "encoder_q-layer.7": 3525.104,
      "encoder_q-layer.8": 3328.5366,
      "encoder_q-layer.9": 2599.9705,
      "epoch": 0.56,
      "inbatch_neg_score": 0.369,
      "inbatch_pos_score": 1.0459,
      "learning_rate": 2.3555555555555556e-05,
      "loss": 3.2333,
      "norm_diff": 0.031,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4635.5641,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3696,
      "query_norm": 1.4492,
      "queue_k_norm": 1.4792,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8333,
      "sent_len_1": 66.8105,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6813,
      "stdk": 0.0489,
      "stdq": 0.0465,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 57600
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2545,
      "doc_norm": 1.4832,
      "encoder_q-embeddings": 2785.8882,
      "encoder_q-layer.0": 2021.7037,
      "encoder_q-layer.1": 2260.6204,
      "encoder_q-layer.10": 2405.5547,
      "encoder_q-layer.11": 5367.0312,
      "encoder_q-layer.2": 2529.4033,
      "encoder_q-layer.3": 2542.6443,
      "encoder_q-layer.4": 2651.0549,
      "encoder_q-layer.5": 2578.0908,
      "encoder_q-layer.6": 2841.6814,
      "encoder_q-layer.7": 3235.6157,
      "encoder_q-layer.8": 3125.2593,
      "encoder_q-layer.9": 2532.2778,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3728,
      "inbatch_pos_score": 1.0791,
      "learning_rate": 2.35e-05,
      "loss": 3.2545,
      "norm_diff": 0.0341,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4375.5643,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3733,
      "query_norm": 1.449,
      "queue_k_norm": 1.4816,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7784,
      "sent_len_1": 67.0284,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7262,
      "stdk": 0.049,
      "stdq": 0.0465,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 57700
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.236,
      "doc_norm": 1.4739,
      "encoder_q-embeddings": 2274.3547,
      "encoder_q-layer.0": 1495.259,
      "encoder_q-layer.1": 1597.5135,
      "encoder_q-layer.10": 2488.0962,
      "encoder_q-layer.11": 5229.4512,
      "encoder_q-layer.2": 1798.0956,
      "encoder_q-layer.3": 1884.2036,
      "encoder_q-layer.4": 2097.9409,
      "encoder_q-layer.5": 2123.1375,
      "encoder_q-layer.6": 2239.3013,
      "encoder_q-layer.7": 2496.3826,
      "encoder_q-layer.8": 2800.6245,
      "encoder_q-layer.9": 2440.5247,
      "epoch": 0.56,
      "inbatch_neg_score": 0.3785,
      "inbatch_pos_score": 1.0684,
      "learning_rate": 2.3444444444444448e-05,
      "loss": 3.236,
      "norm_diff": 0.0385,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3822.1988,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3792,
      "query_norm": 1.4354,
      "queue_k_norm": 1.4808,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9136,
      "sent_len_1": 66.7715,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.185,
      "stdk": 0.0486,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 57800
    },
    {
      "accuracy": 50.9766,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2314,
      "doc_norm": 1.4851,
      "encoder_q-embeddings": 2246.7974,
      "encoder_q-layer.0": 1501.1688,
      "encoder_q-layer.1": 1530.6754,
      "encoder_q-layer.10": 2284.792,
      "encoder_q-layer.11": 5251.3647,
      "encoder_q-layer.2": 1800.0486,
      "encoder_q-layer.3": 1910.7146,
      "encoder_q-layer.4": 2060.6382,
      "encoder_q-layer.5": 2136.5322,
      "encoder_q-layer.6": 2385.7773,
      "encoder_q-layer.7": 2694.835,
      "encoder_q-layer.8": 2891.0928,
      "encoder_q-layer.9": 2402.4233,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3741,
      "inbatch_pos_score": 1.0146,
      "learning_rate": 2.338888888888889e-05,
      "loss": 3.2314,
      "norm_diff": 0.055,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3859.7088,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.375,
      "query_norm": 1.4301,
      "queue_k_norm": 1.4805,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2674,
      "sent_len_1": 66.8367,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9487,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 57900
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2392,
      "doc_norm": 1.4779,
      "encoder_q-embeddings": 2111.0615,
      "encoder_q-layer.0": 1433.8735,
      "encoder_q-layer.1": 1551.6718,
      "encoder_q-layer.10": 2258.9329,
      "encoder_q-layer.11": 5258.7676,
      "encoder_q-layer.2": 1778.7115,
      "encoder_q-layer.3": 1793.0382,
      "encoder_q-layer.4": 1876.6757,
      "encoder_q-layer.5": 2027.6704,
      "encoder_q-layer.6": 2163.843,
      "encoder_q-layer.7": 2474.1528,
      "encoder_q-layer.8": 2663.5017,
      "encoder_q-layer.9": 2345.3064,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3727,
      "inbatch_pos_score": 1.0615,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 3.2392,
      "norm_diff": 0.0313,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3721.6473,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3728,
      "query_norm": 1.4466,
      "queue_k_norm": 1.4816,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.814,
      "sent_len_1": 66.7058,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.68,
      "stdk": 0.0487,
      "stdq": 0.0465,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58000
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2115,
      "doc_norm": 1.4836,
      "encoder_q-embeddings": 2129.3435,
      "encoder_q-layer.0": 1374.0822,
      "encoder_q-layer.1": 1514.7676,
      "encoder_q-layer.10": 2279.9531,
      "encoder_q-layer.11": 5323.4619,
      "encoder_q-layer.2": 1764.7936,
      "encoder_q-layer.3": 1831.2867,
      "encoder_q-layer.4": 1935.9261,
      "encoder_q-layer.5": 2139.7332,
      "encoder_q-layer.6": 2204.0898,
      "encoder_q-layer.7": 2563.1211,
      "encoder_q-layer.8": 2944.6265,
      "encoder_q-layer.9": 2434.7625,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3721,
      "inbatch_pos_score": 1.043,
      "learning_rate": 2.3277777777777778e-05,
      "loss": 3.2115,
      "norm_diff": 0.0557,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3809.5045,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3726,
      "query_norm": 1.4279,
      "queue_k_norm": 1.4832,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0499,
      "sent_len_1": 66.6467,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1962,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58100
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2161,
      "doc_norm": 1.481,
      "encoder_q-embeddings": 1807.8369,
      "encoder_q-layer.0": 1242.4614,
      "encoder_q-layer.1": 1430.5211,
      "encoder_q-layer.10": 1256.3737,
      "encoder_q-layer.11": 2623.9595,
      "encoder_q-layer.2": 1686.5399,
      "encoder_q-layer.3": 1754.0078,
      "encoder_q-layer.4": 1867.3206,
      "encoder_q-layer.5": 1844.3683,
      "encoder_q-layer.6": 1865.9663,
      "encoder_q-layer.7": 2046.4343,
      "encoder_q-layer.8": 1925.2169,
      "encoder_q-layer.9": 1346.9208,
      "epoch": 0.57,
      "inbatch_neg_score": 0.375,
      "inbatch_pos_score": 1.0342,
      "learning_rate": 2.3222222222222224e-05,
      "loss": 3.2161,
      "norm_diff": 0.0765,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2674.5121,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3752,
      "query_norm": 1.4046,
      "queue_k_norm": 1.4815,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2143,
      "sent_len_1": 67.0368,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5037,
      "stdk": 0.0488,
      "stdq": 0.0448,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58200
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2729,
      "doc_norm": 1.485,
      "encoder_q-embeddings": 1234.8131,
      "encoder_q-layer.0": 840.254,
      "encoder_q-layer.1": 1021.2996,
      "encoder_q-layer.10": 1307.8557,
      "encoder_q-layer.11": 2770.2949,
      "encoder_q-layer.2": 1118.8516,
      "encoder_q-layer.3": 1211.8312,
      "encoder_q-layer.4": 1249.9968,
      "encoder_q-layer.5": 1145.2371,
      "encoder_q-layer.6": 1261.4974,
      "encoder_q-layer.7": 1368.9902,
      "encoder_q-layer.8": 1532.0051,
      "encoder_q-layer.9": 1262.4761,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3731,
      "inbatch_pos_score": 1.0322,
      "learning_rate": 2.3166666666666666e-05,
      "loss": 3.2729,
      "norm_diff": 0.0683,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2098.4774,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3733,
      "query_norm": 1.4167,
      "queue_k_norm": 1.4851,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8897,
      "sent_len_1": 66.8773,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1738,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 58300
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2269,
      "doc_norm": 1.4747,
      "encoder_q-embeddings": 1165.3013,
      "encoder_q-layer.0": 782.6828,
      "encoder_q-layer.1": 882.038,
      "encoder_q-layer.10": 1266.0754,
      "encoder_q-layer.11": 2701.8633,
      "encoder_q-layer.2": 984.8092,
      "encoder_q-layer.3": 1058.887,
      "encoder_q-layer.4": 1119.1152,
      "encoder_q-layer.5": 1286.0273,
      "encoder_q-layer.6": 1387.5836,
      "encoder_q-layer.7": 1553.72,
      "encoder_q-layer.8": 1611.3322,
      "encoder_q-layer.9": 1382.2734,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3708,
      "inbatch_pos_score": 1.0215,
      "learning_rate": 2.3111111111111112e-05,
      "loss": 3.2269,
      "norm_diff": 0.0631,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2098.5018,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3706,
      "query_norm": 1.4115,
      "queue_k_norm": 1.483,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0208,
      "sent_len_1": 66.7304,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1813,
      "stdk": 0.0486,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58400
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2365,
      "doc_norm": 1.481,
      "encoder_q-embeddings": 1296.2507,
      "encoder_q-layer.0": 839.9404,
      "encoder_q-layer.1": 940.9912,
      "encoder_q-layer.10": 1315.7197,
      "encoder_q-layer.11": 2653.3301,
      "encoder_q-layer.2": 1063.17,
      "encoder_q-layer.3": 1148.7961,
      "encoder_q-layer.4": 1197.1791,
      "encoder_q-layer.5": 1294.4727,
      "encoder_q-layer.6": 1389.9193,
      "encoder_q-layer.7": 1494.5872,
      "encoder_q-layer.8": 1543.0814,
      "encoder_q-layer.9": 1190.5802,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3733,
      "inbatch_pos_score": 1.0273,
      "learning_rate": 2.3055555555555558e-05,
      "loss": 3.2365,
      "norm_diff": 0.0693,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2114.7149,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3733,
      "query_norm": 1.4117,
      "queue_k_norm": 1.4842,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.099,
      "sent_len_1": 66.5761,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9812,
      "stdk": 0.0488,
      "stdq": 0.0451,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58500
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2343,
      "doc_norm": 1.4811,
      "encoder_q-embeddings": 6741.2349,
      "encoder_q-layer.0": 4982.4414,
      "encoder_q-layer.1": 5326.1895,
      "encoder_q-layer.10": 1236.051,
      "encoder_q-layer.11": 2738.7795,
      "encoder_q-layer.2": 6098.9258,
      "encoder_q-layer.3": 7158.2446,
      "encoder_q-layer.4": 7763.5508,
      "encoder_q-layer.5": 8164.394,
      "encoder_q-layer.6": 7681.416,
      "encoder_q-layer.7": 6992.5737,
      "encoder_q-layer.8": 5510.6802,
      "encoder_q-layer.9": 1714.272,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3693,
      "inbatch_pos_score": 1.0244,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 3.2343,
      "norm_diff": 0.0672,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9072.5265,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3699,
      "query_norm": 1.4139,
      "queue_k_norm": 1.4824,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9241,
      "sent_len_1": 66.7294,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.785,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58600
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2425,
      "doc_norm": 1.4782,
      "encoder_q-embeddings": 2108.739,
      "encoder_q-layer.0": 1430.9731,
      "encoder_q-layer.1": 1657.4059,
      "encoder_q-layer.10": 1294.033,
      "encoder_q-layer.11": 2643.9075,
      "encoder_q-layer.2": 2083.916,
      "encoder_q-layer.3": 2316.9341,
      "encoder_q-layer.4": 2375.9473,
      "encoder_q-layer.5": 2478.2507,
      "encoder_q-layer.6": 2079.1797,
      "encoder_q-layer.7": 2011.8567,
      "encoder_q-layer.8": 1840.9968,
      "encoder_q-layer.9": 1321.1821,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3695,
      "inbatch_pos_score": 1.0566,
      "learning_rate": 2.2944444444444446e-05,
      "loss": 3.2425,
      "norm_diff": 0.0406,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2989.1616,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3701,
      "query_norm": 1.4376,
      "queue_k_norm": 1.4838,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.008,
      "sent_len_1": 66.7881,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.77,
      "stdk": 0.0487,
      "stdq": 0.0462,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 58700
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.246,
      "doc_norm": 1.482,
      "encoder_q-embeddings": 1426.3964,
      "encoder_q-layer.0": 943.2488,
      "encoder_q-layer.1": 1036.9214,
      "encoder_q-layer.10": 1266.7295,
      "encoder_q-layer.11": 2673.0183,
      "encoder_q-layer.2": 1252.7764,
      "encoder_q-layer.3": 1256.1726,
      "encoder_q-layer.4": 1378.7083,
      "encoder_q-layer.5": 1540.4275,
      "encoder_q-layer.6": 1831.3196,
      "encoder_q-layer.7": 2224.5322,
      "encoder_q-layer.8": 1970.309,
      "encoder_q-layer.9": 1238.3535,
      "epoch": 0.57,
      "inbatch_neg_score": 0.3676,
      "inbatch_pos_score": 1.0244,
      "learning_rate": 2.288888888888889e-05,
      "loss": 3.246,
      "norm_diff": 0.0563,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2425.4736,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3667,
      "query_norm": 1.4257,
      "queue_k_norm": 1.4819,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9281,
      "sent_len_1": 66.6198,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.63,
      "stdk": 0.0488,
      "stdq": 0.0457,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 58800
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2341,
      "doc_norm": 1.4845,
      "encoder_q-embeddings": 2705.8613,
      "encoder_q-layer.0": 1828.7273,
      "encoder_q-layer.1": 1965.0187,
      "encoder_q-layer.10": 1326.0631,
      "encoder_q-layer.11": 2662.0981,
      "encoder_q-layer.2": 2359.1328,
      "encoder_q-layer.3": 2489.8345,
      "encoder_q-layer.4": 2592.4551,
      "encoder_q-layer.5": 2535.875,
      "encoder_q-layer.6": 2343.9319,
      "encoder_q-layer.7": 1920.0669,
      "encoder_q-layer.8": 1701.1536,
      "encoder_q-layer.9": 1322.4008,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3729,
      "inbatch_pos_score": 1.0527,
      "learning_rate": 2.2833333333333334e-05,
      "loss": 3.2341,
      "norm_diff": 0.0451,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3265.5412,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3728,
      "query_norm": 1.4394,
      "queue_k_norm": 1.4817,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9412,
      "sent_len_1": 66.6572,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9437,
      "stdk": 0.0489,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 58900
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2481,
      "doc_norm": 1.4834,
      "encoder_q-embeddings": 951.6925,
      "encoder_q-layer.0": 609.4625,
      "encoder_q-layer.1": 629.248,
      "encoder_q-layer.10": 1215.9781,
      "encoder_q-layer.11": 2663.083,
      "encoder_q-layer.2": 694.0403,
      "encoder_q-layer.3": 719.4877,
      "encoder_q-layer.4": 772.3863,
      "encoder_q-layer.5": 810.8699,
      "encoder_q-layer.6": 955.1356,
      "encoder_q-layer.7": 1122.6117,
      "encoder_q-layer.8": 1312.9832,
      "encoder_q-layer.9": 1159.7871,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3723,
      "inbatch_pos_score": 1.041,
      "learning_rate": 2.277777777777778e-05,
      "loss": 3.2481,
      "norm_diff": 0.0514,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1783.2123,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3718,
      "query_norm": 1.432,
      "queue_k_norm": 1.4833,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0052,
      "sent_len_1": 66.6221,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5538,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 59000
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2456,
      "doc_norm": 1.4894,
      "encoder_q-embeddings": 2328.9441,
      "encoder_q-layer.0": 1706.8611,
      "encoder_q-layer.1": 1992.8762,
      "encoder_q-layer.10": 1246.041,
      "encoder_q-layer.11": 2766.1099,
      "encoder_q-layer.2": 2393.1824,
      "encoder_q-layer.3": 2425.9644,
      "encoder_q-layer.4": 2716.353,
      "encoder_q-layer.5": 2572.9836,
      "encoder_q-layer.6": 2740.2017,
      "encoder_q-layer.7": 3095.4944,
      "encoder_q-layer.8": 2534.9001,
      "encoder_q-layer.9": 1299.7183,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3697,
      "inbatch_pos_score": 1.0518,
      "learning_rate": 2.2722222222222222e-05,
      "loss": 3.2456,
      "norm_diff": 0.0636,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3575.4678,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3689,
      "query_norm": 1.4257,
      "queue_k_norm": 1.4834,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9533,
      "sent_len_1": 66.776,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.3638,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 59100
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2409,
      "doc_norm": 1.475,
      "encoder_q-embeddings": 1776.7966,
      "encoder_q-layer.0": 1249.5449,
      "encoder_q-layer.1": 1346.5496,
      "encoder_q-layer.10": 1193.3099,
      "encoder_q-layer.11": 2656.3909,
      "encoder_q-layer.2": 1609.2906,
      "encoder_q-layer.3": 1701.645,
      "encoder_q-layer.4": 1974.1188,
      "encoder_q-layer.5": 1910.5322,
      "encoder_q-layer.6": 2027.8569,
      "encoder_q-layer.7": 2061.4094,
      "encoder_q-layer.8": 1800.928,
      "encoder_q-layer.9": 1321.6943,
      "epoch": 0.58,
      "inbatch_neg_score": 0.372,
      "inbatch_pos_score": 1.0068,
      "learning_rate": 2.2666666666666668e-05,
      "loss": 3.2409,
      "norm_diff": 0.0629,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2671.9452,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3723,
      "query_norm": 1.4121,
      "queue_k_norm": 1.4823,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7839,
      "sent_len_1": 66.5178,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4712,
      "stdk": 0.0485,
      "stdq": 0.0449,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 59200
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2437,
      "doc_norm": 1.4803,
      "encoder_q-embeddings": 1678.2751,
      "encoder_q-layer.0": 1152.5555,
      "encoder_q-layer.1": 1262.5835,
      "encoder_q-layer.10": 1259.549,
      "encoder_q-layer.11": 2698.4316,
      "encoder_q-layer.2": 1416.7975,
      "encoder_q-layer.3": 1567.031,
      "encoder_q-layer.4": 1614.1057,
      "encoder_q-layer.5": 1691.6161,
      "encoder_q-layer.6": 1771.9006,
      "encoder_q-layer.7": 2091.0752,
      "encoder_q-layer.8": 1797.8201,
      "encoder_q-layer.9": 1212.7612,
      "epoch": 0.58,
      "inbatch_neg_score": 0.373,
      "inbatch_pos_score": 1.0371,
      "learning_rate": 2.2611111111111113e-05,
      "loss": 3.2437,
      "norm_diff": 0.0574,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2508.3799,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3726,
      "query_norm": 1.4229,
      "queue_k_norm": 1.4818,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9721,
      "sent_len_1": 66.5892,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1463,
      "stdk": 0.0487,
      "stdq": 0.0453,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 59300
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2208,
      "doc_norm": 1.4736,
      "encoder_q-embeddings": 1100.4573,
      "encoder_q-layer.0": 693.9564,
      "encoder_q-layer.1": 763.4878,
      "encoder_q-layer.10": 1211.9359,
      "encoder_q-layer.11": 2610.5527,
      "encoder_q-layer.2": 887.7753,
      "encoder_q-layer.3": 877.621,
      "encoder_q-layer.4": 925.3334,
      "encoder_q-layer.5": 1006.9426,
      "encoder_q-layer.6": 1057.1116,
      "encoder_q-layer.7": 1244.0792,
      "encoder_q-layer.8": 1334.5861,
      "encoder_q-layer.9": 1190.9852,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3735,
      "inbatch_pos_score": 1.084,
      "learning_rate": 2.255555555555556e-05,
      "loss": 3.2208,
      "norm_diff": 0.0111,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1854.6283,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3726,
      "query_norm": 1.4669,
      "queue_k_norm": 1.482,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8148,
      "sent_len_1": 66.871,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0513,
      "stdk": 0.0485,
      "stdq": 0.0469,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 59400
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2342,
      "doc_norm": 1.4815,
      "encoder_q-embeddings": 1292.8199,
      "encoder_q-layer.0": 867.569,
      "encoder_q-layer.1": 905.5621,
      "encoder_q-layer.10": 1218.9939,
      "encoder_q-layer.11": 2686.3599,
      "encoder_q-layer.2": 1066.9955,
      "encoder_q-layer.3": 1168.7935,
      "encoder_q-layer.4": 1214.636,
      "encoder_q-layer.5": 1335.2168,
      "encoder_q-layer.6": 1350.2461,
      "encoder_q-layer.7": 1456.5557,
      "encoder_q-layer.8": 1554.0271,
      "encoder_q-layer.9": 1257.1686,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3737,
      "inbatch_pos_score": 1.043,
      "learning_rate": 2.25e-05,
      "loss": 3.2342,
      "norm_diff": 0.0371,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2123.2992,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3735,
      "query_norm": 1.4444,
      "queue_k_norm": 1.4836,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8757,
      "sent_len_1": 67.022,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.875,
      "stdk": 0.0488,
      "stdq": 0.046,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 59500
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2193,
      "doc_norm": 1.4848,
      "encoder_q-embeddings": 1382.3181,
      "encoder_q-layer.0": 919.9025,
      "encoder_q-layer.1": 989.8084,
      "encoder_q-layer.10": 1343.5951,
      "encoder_q-layer.11": 2863.3235,
      "encoder_q-layer.2": 1120.0734,
      "encoder_q-layer.3": 1237.583,
      "encoder_q-layer.4": 1276.7999,
      "encoder_q-layer.5": 1358.321,
      "encoder_q-layer.6": 1544.3029,
      "encoder_q-layer.7": 1628.8215,
      "encoder_q-layer.8": 1637.9274,
      "encoder_q-layer.9": 1326.272,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3773,
      "inbatch_pos_score": 1.0371,
      "learning_rate": 2.2444444444444447e-05,
      "loss": 3.2193,
      "norm_diff": 0.0554,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2257.0988,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3774,
      "query_norm": 1.4293,
      "queue_k_norm": 1.4825,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.956,
      "sent_len_1": 66.7029,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4338,
      "stdk": 0.0489,
      "stdq": 0.0453,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 59600
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2154,
      "doc_norm": 1.4796,
      "encoder_q-embeddings": 6987.8022,
      "encoder_q-layer.0": 5000.0127,
      "encoder_q-layer.1": 6008.4004,
      "encoder_q-layer.10": 1294.0751,
      "encoder_q-layer.11": 2693.0283,
      "encoder_q-layer.2": 6661.1401,
      "encoder_q-layer.3": 7780.8237,
      "encoder_q-layer.4": 8896.7012,
      "encoder_q-layer.5": 11413.7227,
      "encoder_q-layer.6": 12818.9053,
      "encoder_q-layer.7": 12924.4307,
      "encoder_q-layer.8": 11276.8164,
      "encoder_q-layer.9": 4820.6514,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3782,
      "inbatch_pos_score": 1.0293,
      "learning_rate": 2.238888888888889e-05,
      "loss": 3.2154,
      "norm_diff": 0.0318,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12561.0169,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.3782,
      "query_norm": 1.4477,
      "queue_k_norm": 1.4828,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9107,
      "sent_len_1": 66.7106,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6438,
      "stdk": 0.0487,
      "stdq": 0.0461,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 59700
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2105,
      "doc_norm": 1.4812,
      "encoder_q-embeddings": 1201.7867,
      "encoder_q-layer.0": 815.59,
      "encoder_q-layer.1": 890.6713,
      "encoder_q-layer.10": 1280.1959,
      "encoder_q-layer.11": 2720.9807,
      "encoder_q-layer.2": 998.9742,
      "encoder_q-layer.3": 1089.7461,
      "encoder_q-layer.4": 1167.8387,
      "encoder_q-layer.5": 1241.303,
      "encoder_q-layer.6": 1370.0525,
      "encoder_q-layer.7": 1567.6742,
      "encoder_q-layer.8": 1665.712,
      "encoder_q-layer.9": 1286.3778,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3828,
      "inbatch_pos_score": 1.04,
      "learning_rate": 2.2333333333333335e-05,
      "loss": 3.2105,
      "norm_diff": 0.0463,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2135.8815,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3828,
      "query_norm": 1.435,
      "queue_k_norm": 1.4834,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9763,
      "sent_len_1": 66.781,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8487,
      "stdk": 0.0488,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 59800
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2106,
      "doc_norm": 1.4859,
      "encoder_q-embeddings": 984.3124,
      "encoder_q-layer.0": 649.7353,
      "encoder_q-layer.1": 686.4484,
      "encoder_q-layer.10": 1319.3224,
      "encoder_q-layer.11": 2672.3264,
      "encoder_q-layer.2": 774.4699,
      "encoder_q-layer.3": 810.8926,
      "encoder_q-layer.4": 875.1655,
      "encoder_q-layer.5": 956.8505,
      "encoder_q-layer.6": 1092.3892,
      "encoder_q-layer.7": 1333.5803,
      "encoder_q-layer.8": 1555.5552,
      "encoder_q-layer.9": 1385.8132,
      "epoch": 0.58,
      "inbatch_neg_score": 0.3868,
      "inbatch_pos_score": 1.043,
      "learning_rate": 2.2277777777777778e-05,
      "loss": 3.2106,
      "norm_diff": 0.0781,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1887.7647,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3872,
      "query_norm": 1.4078,
      "queue_k_norm": 1.4836,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9694,
      "sent_len_1": 66.8533,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.34,
      "stdk": 0.049,
      "stdq": 0.0444,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 59900
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2265,
      "doc_norm": 1.4871,
      "encoder_q-embeddings": 647.5739,
      "encoder_q-layer.0": 453.8141,
      "encoder_q-layer.1": 496.2095,
      "encoder_q-layer.10": 601.7897,
      "encoder_q-layer.11": 1414.9572,
      "encoder_q-layer.2": 577.6119,
      "encoder_q-layer.3": 595.0921,
      "encoder_q-layer.4": 663.7764,
      "encoder_q-layer.5": 711.8868,
      "encoder_q-layer.6": 723.6525,
      "encoder_q-layer.7": 722.7355,
      "encoder_q-layer.8": 742.2968,
      "encoder_q-layer.9": 619.955,
      "epoch": 0.59,
      "inbatch_neg_score": 0.3913,
      "inbatch_pos_score": 1.084,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 3.2265,
      "norm_diff": 0.0365,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1092.5661,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3916,
      "query_norm": 1.4506,
      "queue_k_norm": 1.4828,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.823,
      "sent_len_1": 66.6466,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9162,
      "stdk": 0.049,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60000
    },
    {
      "dev_runtime": 38.5299,
      "dev_samples_per_second": 1.661,
      "dev_steps_per_second": 0.026,
      "epoch": 0.59,
      "step": 60000,
      "test_accuracy": 94.4580078125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3176065683364868,
      "test_doc_norm": 1.4854023456573486,
      "test_inbatch_neg_score": 0.759244441986084,
      "test_inbatch_pos_score": 1.731167197227478,
      "test_loss": 0.3176065683364868,
      "test_loss_align": 1.0831103324890137,
      "test_loss_unif": 3.7100939750671387,
      "test_loss_unif_q@queue": 3.7100939750671387,
      "test_norm_diff": 0.03488502651453018,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.3748781085014343,
      "test_query_norm": 1.520287275314331,
      "test_queue_k_norm": 1.482785940170288,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04331099987030029,
      "test_stdq": 0.04362639784812927,
      "test_stdqueue_k": 0.04889556020498276,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 38.5299,
      "dev_samples_per_second": 1.661,
      "dev_steps_per_second": 0.026,
      "epoch": 0.59,
      "eval_beir-arguana_ndcg@10": 0.37413,
      "eval_beir-arguana_recall@10": 0.63585,
      "eval_beir-arguana_recall@100": 0.9239,
      "eval_beir-arguana_recall@20": 0.77383,
      "eval_beir-avg_ndcg@10": 0.3619894166666667,
      "eval_beir-avg_recall@10": 0.4269369166666667,
      "eval_beir-avg_recall@100": 0.6093983333333333,
      "eval_beir-avg_recall@20": 0.48618683333333335,
      "eval_beir-cqadupstack_ndcg@10": 0.26613416666666667,
      "eval_beir-cqadupstack_recall@10": 0.3563691666666667,
      "eval_beir-cqadupstack_recall@100": 0.5823433333333333,
      "eval_beir-cqadupstack_recall@20": 0.42005833333333337,
      "eval_beir-fiqa_ndcg@10": 0.20731,
      "eval_beir-fiqa_recall@10": 0.26291,
      "eval_beir-fiqa_recall@100": 0.52885,
      "eval_beir-fiqa_recall@20": 0.32106,
      "eval_beir-nfcorpus_ndcg@10": 0.28619,
      "eval_beir-nfcorpus_recall@10": 0.14512,
      "eval_beir-nfcorpus_recall@100": 0.25994,
      "eval_beir-nfcorpus_recall@20": 0.17284,
      "eval_beir-nq_ndcg@10": 0.26354,
      "eval_beir-nq_recall@10": 0.42862,
      "eval_beir-nq_recall@100": 0.77851,
      "eval_beir-nq_recall@20": 0.55552,
      "eval_beir-quora_ndcg@10": 0.82317,
      "eval_beir-quora_recall@10": 0.91571,
      "eval_beir-quora_recall@100": 0.9848,
      "eval_beir-quora_recall@20": 0.94957,
      "eval_beir-scidocs_ndcg@10": 0.14061,
      "eval_beir-scidocs_recall@10": 0.14773,
      "eval_beir-scidocs_recall@100": 0.34017,
      "eval_beir-scidocs_recall@20": 0.19728,
      "eval_beir-scifact_ndcg@10": 0.60123,
      "eval_beir-scifact_recall@10": 0.74639,
      "eval_beir-scifact_recall@100": 0.89656,
      "eval_beir-scifact_recall@20": 0.79483,
      "eval_beir-trec-covid_ndcg@10": 0.47732,
      "eval_beir-trec-covid_recall@10": 0.502,
      "eval_beir-trec-covid_recall@100": 0.383,
      "eval_beir-trec-covid_recall@20": 0.478,
      "eval_beir-webis-touche2020_ndcg@10": 0.18026,
      "eval_beir-webis-touche2020_recall@10": 0.12867,
      "eval_beir-webis-touche2020_recall@100": 0.41591,
      "eval_beir-webis-touche2020_recall@20": 0.19888,
      "eval_senteval-avg_sts": 0.7327010099677189,
      "eval_senteval-sickr_spearman": 0.7026664835653339,
      "eval_senteval-stsb_spearman": 0.7627355363701038,
      "step": 60000,
      "test_accuracy": 94.4580078125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3176065683364868,
      "test_doc_norm": 1.4854023456573486,
      "test_inbatch_neg_score": 0.759244441986084,
      "test_inbatch_pos_score": 1.731167197227478,
      "test_loss": 0.3176065683364868,
      "test_loss_align": 1.0831103324890137,
      "test_loss_unif": 3.7100939750671387,
      "test_loss_unif_q@queue": 3.7100939750671387,
      "test_norm_diff": 0.03488502651453018,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.3748781085014343,
      "test_query_norm": 1.520287275314331,
      "test_queue_k_norm": 1.482785940170288,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04331099987030029,
      "test_stdq": 0.04362639784812927,
      "test_stdqueue_k": 0.04889556020498276,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2325,
      "doc_norm": 1.4858,
      "encoder_q-embeddings": 530.4781,
      "encoder_q-layer.0": 355.0197,
      "encoder_q-layer.1": 399.8825,
      "encoder_q-layer.10": 591.3457,
      "encoder_q-layer.11": 1344.3657,
      "encoder_q-layer.2": 471.4402,
      "encoder_q-layer.3": 500.9124,
      "encoder_q-layer.4": 531.5427,
      "encoder_q-layer.5": 564.7137,
      "encoder_q-layer.6": 600.617,
      "encoder_q-layer.7": 666.225,
      "encoder_q-layer.8": 702.0323,
      "encoder_q-layer.9": 610.3096,
      "epoch": 0.59,
      "inbatch_neg_score": 0.3979,
      "inbatch_pos_score": 1.1025,
      "learning_rate": 2.216666666666667e-05,
      "loss": 3.2325,
      "norm_diff": 0.0382,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 975.2374,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3979,
      "query_norm": 1.4475,
      "queue_k_norm": 1.4848,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9699,
      "sent_len_1": 66.7978,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7713,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60100
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2171,
      "doc_norm": 1.4798,
      "encoder_q-embeddings": 653.9945,
      "encoder_q-layer.0": 435.5611,
      "encoder_q-layer.1": 495.7063,
      "encoder_q-layer.10": 572.1553,
      "encoder_q-layer.11": 1337.6598,
      "encoder_q-layer.2": 570.6015,
      "encoder_q-layer.3": 574.9038,
      "encoder_q-layer.4": 627.9908,
      "encoder_q-layer.5": 702.2622,
      "encoder_q-layer.6": 710.2473,
      "encoder_q-layer.7": 714.8608,
      "encoder_q-layer.8": 732.435,
      "encoder_q-layer.9": 615.8098,
      "epoch": 0.59,
      "inbatch_neg_score": 0.4001,
      "inbatch_pos_score": 1.082,
      "learning_rate": 2.211111111111111e-05,
      "loss": 3.2171,
      "norm_diff": 0.0415,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1055.8863,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4014,
      "query_norm": 1.4383,
      "queue_k_norm": 1.4855,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9608,
      "sent_len_1": 67.0308,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3363,
      "stdk": 0.0486,
      "stdq": 0.0456,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60200
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2371,
      "doc_norm": 1.4826,
      "encoder_q-embeddings": 510.7079,
      "encoder_q-layer.0": 352.5789,
      "encoder_q-layer.1": 364.5461,
      "encoder_q-layer.10": 605.526,
      "encoder_q-layer.11": 1349.403,
      "encoder_q-layer.2": 408.748,
      "encoder_q-layer.3": 427.6019,
      "encoder_q-layer.4": 478.2571,
      "encoder_q-layer.5": 480.059,
      "encoder_q-layer.6": 554.2308,
      "encoder_q-layer.7": 642.988,
      "encoder_q-layer.8": 707.9402,
      "encoder_q-layer.9": 618.1706,
      "epoch": 0.59,
      "inbatch_neg_score": 0.3982,
      "inbatch_pos_score": 1.0762,
      "learning_rate": 2.2055555555555557e-05,
      "loss": 3.2371,
      "norm_diff": 0.0436,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 940.7401,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3989,
      "query_norm": 1.439,
      "queue_k_norm": 1.485,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0682,
      "sent_len_1": 66.6661,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1887,
      "stdk": 0.0487,
      "stdq": 0.0459,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60300
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2376,
      "doc_norm": 1.4857,
      "encoder_q-embeddings": 493.6274,
      "encoder_q-layer.0": 324.3079,
      "encoder_q-layer.1": 341.9132,
      "encoder_q-layer.10": 639.0984,
      "encoder_q-layer.11": 1385.0693,
      "encoder_q-layer.2": 385.2509,
      "encoder_q-layer.3": 403.1235,
      "encoder_q-layer.4": 424.8866,
      "encoder_q-layer.5": 472.687,
      "encoder_q-layer.6": 523.8005,
      "encoder_q-layer.7": 633.8199,
      "encoder_q-layer.8": 704.7192,
      "encoder_q-layer.9": 625.6462,
      "epoch": 0.59,
      "inbatch_neg_score": 0.403,
      "inbatch_pos_score": 1.0928,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 3.2376,
      "norm_diff": 0.0357,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 944.7979,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4048,
      "query_norm": 1.45,
      "queue_k_norm": 1.485,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0087,
      "sent_len_1": 66.7212,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2575,
      "stdk": 0.0488,
      "stdq": 0.0461,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60400
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2405,
      "doc_norm": 1.4857,
      "encoder_q-embeddings": 1453.0599,
      "encoder_q-layer.0": 1002.1303,
      "encoder_q-layer.1": 1034.2115,
      "encoder_q-layer.10": 623.3882,
      "encoder_q-layer.11": 1345.3962,
      "encoder_q-layer.2": 1136.8206,
      "encoder_q-layer.3": 1281.3127,
      "encoder_q-layer.4": 1491.3202,
      "encoder_q-layer.5": 1544.5057,
      "encoder_q-layer.6": 1665.1313,
      "encoder_q-layer.7": 1954.3557,
      "encoder_q-layer.8": 1770.9849,
      "encoder_q-layer.9": 955.7221,
      "epoch": 0.59,
      "inbatch_neg_score": 0.3997,
      "inbatch_pos_score": 1.0664,
      "learning_rate": 2.1944444444444445e-05,
      "loss": 3.2405,
      "norm_diff": 0.0433,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2085.6963,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.3994,
      "query_norm": 1.4424,
      "queue_k_norm": 1.4866,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9525,
      "sent_len_1": 66.699,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9725,
      "stdk": 0.0488,
      "stdq": 0.0459,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60500
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2077,
      "doc_norm": 1.4866,
      "encoder_q-embeddings": 614.1601,
      "encoder_q-layer.0": 404.1742,
      "encoder_q-layer.1": 424.0622,
      "encoder_q-layer.10": 641.1713,
      "encoder_q-layer.11": 1399.4604,
      "encoder_q-layer.2": 507.6924,
      "encoder_q-layer.3": 540.2646,
      "encoder_q-layer.4": 592.9575,
      "encoder_q-layer.5": 633.5609,
      "encoder_q-layer.6": 604.8401,
      "encoder_q-layer.7": 702.7231,
      "encoder_q-layer.8": 743.9575,
      "encoder_q-layer.9": 626.9386,
      "epoch": 0.59,
      "inbatch_neg_score": 0.4041,
      "inbatch_pos_score": 1.0566,
      "learning_rate": 2.188888888888889e-05,
      "loss": 3.2077,
      "norm_diff": 0.0574,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1046.3221,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4055,
      "query_norm": 1.4292,
      "queue_k_norm": 1.4897,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.942,
      "sent_len_1": 67.0051,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.055,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 60600
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2203,
      "doc_norm": 1.4877,
      "encoder_q-embeddings": 560.2883,
      "encoder_q-layer.0": 345.4669,
      "encoder_q-layer.1": 367.6927,
      "encoder_q-layer.10": 595.5017,
      "encoder_q-layer.11": 1308.3966,
      "encoder_q-layer.2": 433.157,
      "encoder_q-layer.3": 455.5255,
      "encoder_q-layer.4": 484.6642,
      "encoder_q-layer.5": 492.1579,
      "encoder_q-layer.6": 543.5474,
      "encoder_q-layer.7": 620.3129,
      "encoder_q-layer.8": 720.0184,
      "encoder_q-layer.9": 622.7245,
      "epoch": 0.59,
      "inbatch_neg_score": 0.4076,
      "inbatch_pos_score": 1.0947,
      "learning_rate": 2.1833333333333333e-05,
      "loss": 3.2203,
      "norm_diff": 0.0391,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 952.4538,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4082,
      "query_norm": 1.4486,
      "queue_k_norm": 1.4885,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0906,
      "sent_len_1": 66.987,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7312,
      "stdk": 0.0488,
      "stdq": 0.046,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60700
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2407,
      "doc_norm": 1.4864,
      "encoder_q-embeddings": 574.1752,
      "encoder_q-layer.0": 389.5114,
      "encoder_q-layer.1": 450.4989,
      "encoder_q-layer.10": 588.3186,
      "encoder_q-layer.11": 1413.1317,
      "encoder_q-layer.2": 512.5,
      "encoder_q-layer.3": 539.125,
      "encoder_q-layer.4": 604.0533,
      "encoder_q-layer.5": 593.9445,
      "encoder_q-layer.6": 590.8491,
      "encoder_q-layer.7": 658.3473,
      "encoder_q-layer.8": 707.8538,
      "encoder_q-layer.9": 641.3834,
      "epoch": 0.59,
      "inbatch_neg_score": 0.4083,
      "inbatch_pos_score": 1.0752,
      "learning_rate": 2.177777777777778e-05,
      "loss": 3.2407,
      "norm_diff": 0.0424,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1022.6137,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4082,
      "query_norm": 1.4439,
      "queue_k_norm": 1.49,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.918,
      "sent_len_1": 66.7307,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1175,
      "stdk": 0.0487,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 60800
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2262,
      "doc_norm": 1.4936,
      "encoder_q-embeddings": 1277.3589,
      "encoder_q-layer.0": 876.8021,
      "encoder_q-layer.1": 1020.5884,
      "encoder_q-layer.10": 616.385,
      "encoder_q-layer.11": 1345.793,
      "encoder_q-layer.2": 1208.593,
      "encoder_q-layer.3": 1374.6139,
      "encoder_q-layer.4": 1490.0122,
      "encoder_q-layer.5": 1817.6262,
      "encoder_q-layer.6": 1771.2329,
      "encoder_q-layer.7": 1392.684,
      "encoder_q-layer.8": 1078.6316,
      "encoder_q-layer.9": 635.0426,
      "epoch": 0.59,
      "inbatch_neg_score": 0.4063,
      "inbatch_pos_score": 1.0615,
      "learning_rate": 2.1722222222222225e-05,
      "loss": 3.2262,
      "norm_diff": 0.0658,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1864.8315,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4072,
      "query_norm": 1.4278,
      "queue_k_norm": 1.4894,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9605,
      "sent_len_1": 66.3643,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.0513,
      "stdk": 0.049,
      "stdq": 0.0452,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 60900
    },
    {
      "accuracy": 57.3242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2252,
      "doc_norm": 1.4845,
      "encoder_q-embeddings": 484.8877,
      "encoder_q-layer.0": 318.5356,
      "encoder_q-layer.1": 344.3434,
      "encoder_q-layer.10": 567.2136,
      "encoder_q-layer.11": 1309.7268,
      "encoder_q-layer.2": 390.9054,
      "encoder_q-layer.3": 409.8229,
      "encoder_q-layer.4": 450.6829,
      "encoder_q-layer.5": 485.9786,
      "encoder_q-layer.6": 531.7285,
      "encoder_q-layer.7": 580.5843,
      "encoder_q-layer.8": 679.1954,
      "encoder_q-layer.9": 598.4219,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4077,
      "inbatch_pos_score": 1.0986,
      "learning_rate": 2.1666666666666667e-05,
      "loss": 3.2252,
      "norm_diff": 0.0484,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 918.7627,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.407,
      "query_norm": 1.4361,
      "queue_k_norm": 1.4914,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9454,
      "sent_len_1": 66.7018,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.97,
      "stdk": 0.0486,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 61000
    },
    {
      "accuracy": 59.375,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2145,
      "doc_norm": 1.4894,
      "encoder_q-embeddings": 506.3975,
      "encoder_q-layer.0": 314.7016,
      "encoder_q-layer.1": 341.4273,
      "encoder_q-layer.10": 607.3353,
      "encoder_q-layer.11": 1356.3552,
      "encoder_q-layer.2": 390.8464,
      "encoder_q-layer.3": 403.01,
      "encoder_q-layer.4": 434.1795,
      "encoder_q-layer.5": 461.9469,
      "encoder_q-layer.6": 510.4001,
      "encoder_q-layer.7": 621.5029,
      "encoder_q-layer.8": 699.9742,
      "encoder_q-layer.9": 612.8306,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4051,
      "inbatch_pos_score": 1.126,
      "learning_rate": 2.1611111111111113e-05,
      "loss": 3.2145,
      "norm_diff": 0.0391,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 925.8093,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4065,
      "query_norm": 1.4503,
      "queue_k_norm": 1.491,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8971,
      "sent_len_1": 66.6262,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2775,
      "stdk": 0.0488,
      "stdq": 0.0462,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 61100
    },
    {
      "accuracy": 52.3438,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2422,
      "doc_norm": 1.4854,
      "encoder_q-embeddings": 557.4802,
      "encoder_q-layer.0": 376.6745,
      "encoder_q-layer.1": 394.8343,
      "encoder_q-layer.10": 717.3892,
      "encoder_q-layer.11": 1486.9265,
      "encoder_q-layer.2": 433.4803,
      "encoder_q-layer.3": 467.5325,
      "encoder_q-layer.4": 489.344,
      "encoder_q-layer.5": 493.2342,
      "encoder_q-layer.6": 535.9312,
      "encoder_q-layer.7": 637.6556,
      "encoder_q-layer.8": 735.0128,
      "encoder_q-layer.9": 686.1616,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4107,
      "inbatch_pos_score": 1.0625,
      "learning_rate": 2.1555555555555555e-05,
      "loss": 3.2422,
      "norm_diff": 0.0393,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1014.3018,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4114,
      "query_norm": 1.4464,
      "queue_k_norm": 1.492,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8698,
      "sent_len_1": 66.6163,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.71,
      "stdk": 0.0486,
      "stdq": 0.0459,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 61200
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2295,
      "doc_norm": 1.494,
      "encoder_q-embeddings": 621.9669,
      "encoder_q-layer.0": 415.8045,
      "encoder_q-layer.1": 461.6447,
      "encoder_q-layer.10": 616.7747,
      "encoder_q-layer.11": 1280.5724,
      "encoder_q-layer.2": 531.1612,
      "encoder_q-layer.3": 571.1211,
      "encoder_q-layer.4": 622.1876,
      "encoder_q-layer.5": 634.5808,
      "encoder_q-layer.6": 675.7037,
      "encoder_q-layer.7": 743.5679,
      "encoder_q-layer.8": 800.166,
      "encoder_q-layer.9": 642.1469,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4078,
      "inbatch_pos_score": 1.1338,
      "learning_rate": 2.15e-05,
      "loss": 3.2295,
      "norm_diff": 0.0252,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1028.6666,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.408,
      "query_norm": 1.4688,
      "queue_k_norm": 1.4911,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9564,
      "sent_len_1": 66.7887,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6775,
      "stdk": 0.0489,
      "stdq": 0.0469,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 61300
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2301,
      "doc_norm": 1.4866,
      "encoder_q-embeddings": 599.0319,
      "encoder_q-layer.0": 410.1107,
      "encoder_q-layer.1": 458.7542,
      "encoder_q-layer.10": 653.601,
      "encoder_q-layer.11": 1391.1235,
      "encoder_q-layer.2": 571.0105,
      "encoder_q-layer.3": 581.2432,
      "encoder_q-layer.4": 638.358,
      "encoder_q-layer.5": 631.2519,
      "encoder_q-layer.6": 637.5406,
      "encoder_q-layer.7": 731.2791,
      "encoder_q-layer.8": 798.1645,
      "encoder_q-layer.9": 676.6208,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4021,
      "inbatch_pos_score": 1.083,
      "learning_rate": 2.1444444444444443e-05,
      "loss": 3.2301,
      "norm_diff": 0.0516,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1080.6652,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4033,
      "query_norm": 1.4349,
      "queue_k_norm": 1.4944,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8257,
      "sent_len_1": 66.7181,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1675,
      "stdk": 0.0487,
      "stdq": 0.0457,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 61400
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.241,
      "doc_norm": 1.4949,
      "encoder_q-embeddings": 536.0862,
      "encoder_q-layer.0": 370.6507,
      "encoder_q-layer.1": 392.9736,
      "encoder_q-layer.10": 628.5786,
      "encoder_q-layer.11": 1275.9366,
      "encoder_q-layer.2": 477.4033,
      "encoder_q-layer.3": 528.0164,
      "encoder_q-layer.4": 578.7834,
      "encoder_q-layer.5": 637.3351,
      "encoder_q-layer.6": 667.1328,
      "encoder_q-layer.7": 762.3666,
      "encoder_q-layer.8": 851.7795,
      "encoder_q-layer.9": 667.1727,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4073,
      "inbatch_pos_score": 1.1201,
      "learning_rate": 2.138888888888889e-05,
      "loss": 3.241,
      "norm_diff": 0.0383,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1027.5928,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4067,
      "query_norm": 1.4566,
      "queue_k_norm": 1.4916,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0553,
      "sent_len_1": 66.7543,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8988,
      "stdk": 0.0489,
      "stdq": 0.0465,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 61500
    },
    {
      "accuracy": 58.1055,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2343,
      "doc_norm": 1.4919,
      "encoder_q-embeddings": 667.0848,
      "encoder_q-layer.0": 437.1859,
      "encoder_q-layer.1": 490.4369,
      "encoder_q-layer.10": 694.2335,
      "encoder_q-layer.11": 1401.502,
      "encoder_q-layer.2": 601.8382,
      "encoder_q-layer.3": 664.1978,
      "encoder_q-layer.4": 725.0717,
      "encoder_q-layer.5": 794.6619,
      "encoder_q-layer.6": 857.9032,
      "encoder_q-layer.7": 1071.8683,
      "encoder_q-layer.8": 1138.5554,
      "encoder_q-layer.9": 817.1213,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4091,
      "inbatch_pos_score": 1.1221,
      "learning_rate": 2.1333333333333335e-05,
      "loss": 3.2343,
      "norm_diff": 0.0331,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1260.2492,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4097,
      "query_norm": 1.4588,
      "queue_k_norm": 1.4933,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8349,
      "sent_len_1": 66.7645,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.835,
      "stdk": 0.0488,
      "stdq": 0.0465,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 61600
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2158,
      "doc_norm": 1.4923,
      "encoder_q-embeddings": 1083.8667,
      "encoder_q-layer.0": 758.9235,
      "encoder_q-layer.1": 910.6014,
      "encoder_q-layer.10": 591.7983,
      "encoder_q-layer.11": 1288.4182,
      "encoder_q-layer.2": 1150.1748,
      "encoder_q-layer.3": 1272.949,
      "encoder_q-layer.4": 1459.0439,
      "encoder_q-layer.5": 1364.908,
      "encoder_q-layer.6": 1161.2164,
      "encoder_q-layer.7": 1049.7815,
      "encoder_q-layer.8": 951.7767,
      "encoder_q-layer.9": 672.9375,
      "epoch": 0.6,
      "inbatch_neg_score": 0.41,
      "inbatch_pos_score": 1.1133,
      "learning_rate": 2.127777777777778e-05,
      "loss": 3.2158,
      "norm_diff": 0.0362,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1661.6763,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4102,
      "query_norm": 1.4561,
      "queue_k_norm": 1.4929,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9259,
      "sent_len_1": 66.8487,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.115,
      "stdk": 0.0488,
      "stdq": 0.0464,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 61700
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2212,
      "doc_norm": 1.4916,
      "encoder_q-embeddings": 475.0765,
      "encoder_q-layer.0": 301.861,
      "encoder_q-layer.1": 317.2011,
      "encoder_q-layer.10": 632.4028,
      "encoder_q-layer.11": 1306.8694,
      "encoder_q-layer.2": 352.614,
      "encoder_q-layer.3": 378.7502,
      "encoder_q-layer.4": 387.0456,
      "encoder_q-layer.5": 410.9973,
      "encoder_q-layer.6": 463.8963,
      "encoder_q-layer.7": 550.1902,
      "encoder_q-layer.8": 639.121,
      "encoder_q-layer.9": 594.9974,
      "epoch": 0.6,
      "inbatch_neg_score": 0.4126,
      "inbatch_pos_score": 1.1016,
      "learning_rate": 2.1222222222222223e-05,
      "loss": 3.2212,
      "norm_diff": 0.0453,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 880.2749,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4126,
      "query_norm": 1.4463,
      "queue_k_norm": 1.4931,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8579,
      "sent_len_1": 66.7185,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5488,
      "stdk": 0.0488,
      "stdq": 0.0459,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 61800
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.229,
      "doc_norm": 1.4868,
      "encoder_q-embeddings": 610.4066,
      "encoder_q-layer.0": 381.1989,
      "encoder_q-layer.1": 442.1959,
      "encoder_q-layer.10": 640.3755,
      "encoder_q-layer.11": 1377.1462,
      "encoder_q-layer.2": 517.2002,
      "encoder_q-layer.3": 550.1354,
      "encoder_q-layer.4": 577.9149,
      "encoder_q-layer.5": 594.0143,
      "encoder_q-layer.6": 664.7084,
      "encoder_q-layer.7": 730.6105,
      "encoder_q-layer.8": 770.4783,
      "encoder_q-layer.9": 663.2803,
      "epoch": 0.6,
      "inbatch_neg_score": 0.412,
      "inbatch_pos_score": 1.0654,
      "learning_rate": 2.116666666666667e-05,
      "loss": 3.229,
      "norm_diff": 0.0647,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1036.7934,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4106,
      "query_norm": 1.4221,
      "queue_k_norm": 1.4947,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.707,
      "sent_len_1": 66.7267,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.975,
      "stdk": 0.0486,
      "stdq": 0.045,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 61900
    },
    {
      "accuracy": 59.7656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2138,
      "doc_norm": 1.4904,
      "encoder_q-embeddings": 1226.7847,
      "encoder_q-layer.0": 826.5023,
      "encoder_q-layer.1": 916.6356,
      "encoder_q-layer.10": 1145.8489,
      "encoder_q-layer.11": 2553.5193,
      "encoder_q-layer.2": 1082.7368,
      "encoder_q-layer.3": 1203.5801,
      "encoder_q-layer.4": 1319.2076,
      "encoder_q-layer.5": 1371.1425,
      "encoder_q-layer.6": 1416.1914,
      "encoder_q-layer.7": 1504.1034,
      "encoder_q-layer.8": 1591.259,
      "encoder_q-layer.9": 1257.5867,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4128,
      "inbatch_pos_score": 1.1191,
      "learning_rate": 2.111111111111111e-05,
      "loss": 3.2138,
      "norm_diff": 0.0461,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2091.3718,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4141,
      "query_norm": 1.4444,
      "queue_k_norm": 1.4938,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1047,
      "sent_len_1": 66.7884,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8175,
      "stdk": 0.0487,
      "stdq": 0.0458,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 62000
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2332,
      "doc_norm": 1.4945,
      "encoder_q-embeddings": 1254.5432,
      "encoder_q-layer.0": 841.0117,
      "encoder_q-layer.1": 930.2258,
      "encoder_q-layer.10": 1183.9987,
      "encoder_q-layer.11": 2748.7329,
      "encoder_q-layer.2": 1083.6008,
      "encoder_q-layer.3": 1142.5891,
      "encoder_q-layer.4": 1235.5709,
      "encoder_q-layer.5": 1318.4954,
      "encoder_q-layer.6": 1374.0624,
      "encoder_q-layer.7": 1463.4292,
      "encoder_q-layer.8": 1490.2526,
      "encoder_q-layer.9": 1275.7299,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4169,
      "inbatch_pos_score": 1.0693,
      "learning_rate": 2.1055555555555556e-05,
      "loss": 3.2332,
      "norm_diff": 0.0557,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2138.6191,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4165,
      "query_norm": 1.4387,
      "queue_k_norm": 1.4955,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0998,
      "sent_len_1": 66.5155,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9888,
      "stdk": 0.0488,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 62100
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1969,
      "doc_norm": 1.4985,
      "encoder_q-embeddings": 1754.0603,
      "encoder_q-layer.0": 1217.9141,
      "encoder_q-layer.1": 1302.5131,
      "encoder_q-layer.10": 1234.6783,
      "encoder_q-layer.11": 2631.7412,
      "encoder_q-layer.2": 1397.5811,
      "encoder_q-layer.3": 1398.1086,
      "encoder_q-layer.4": 1355.611,
      "encoder_q-layer.5": 1365.6409,
      "encoder_q-layer.6": 1422.6722,
      "encoder_q-layer.7": 1522.9847,
      "encoder_q-layer.8": 1515.7716,
      "encoder_q-layer.9": 1326.7992,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4181,
      "inbatch_pos_score": 1.0977,
      "learning_rate": 2.1e-05,
      "loss": 3.1969,
      "norm_diff": 0.0448,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2334.5142,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.417,
      "query_norm": 1.4537,
      "queue_k_norm": 1.4941,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.953,
      "sent_len_1": 66.5143,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.62,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 62200
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.22,
      "doc_norm": 1.4985,
      "encoder_q-embeddings": 1016.699,
      "encoder_q-layer.0": 657.991,
      "encoder_q-layer.1": 736.7773,
      "encoder_q-layer.10": 1311.6809,
      "encoder_q-layer.11": 2730.5933,
      "encoder_q-layer.2": 830.534,
      "encoder_q-layer.3": 837.9756,
      "encoder_q-layer.4": 899.2057,
      "encoder_q-layer.5": 961.806,
      "encoder_q-layer.6": 1078.0076,
      "encoder_q-layer.7": 1258.1823,
      "encoder_q-layer.8": 1453.8092,
      "encoder_q-layer.9": 1271.5115,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4173,
      "inbatch_pos_score": 1.0957,
      "learning_rate": 2.0944444444444445e-05,
      "loss": 3.22,
      "norm_diff": 0.0421,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1922.2442,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4172,
      "query_norm": 1.4564,
      "queue_k_norm": 1.4944,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9377,
      "sent_len_1": 66.7552,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3063,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 62300
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2147,
      "doc_norm": 1.4962,
      "encoder_q-embeddings": 1863.5292,
      "encoder_q-layer.0": 1404.6775,
      "encoder_q-layer.1": 1433.7124,
      "encoder_q-layer.10": 1222.6865,
      "encoder_q-layer.11": 2620.7529,
      "encoder_q-layer.2": 1501.8065,
      "encoder_q-layer.3": 1495.6311,
      "encoder_q-layer.4": 1649.4762,
      "encoder_q-layer.5": 1643.7604,
      "encoder_q-layer.6": 1615.4962,
      "encoder_q-layer.7": 1886.8806,
      "encoder_q-layer.8": 1688.4178,
      "encoder_q-layer.9": 1218.3755,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4199,
      "inbatch_pos_score": 1.0986,
      "learning_rate": 2.088888888888889e-05,
      "loss": 3.2147,
      "norm_diff": 0.0408,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2527.1922,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4199,
      "query_norm": 1.4554,
      "queue_k_norm": 1.495,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9907,
      "sent_len_1": 66.8622,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4338,
      "stdk": 0.0488,
      "stdq": 0.046,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 62400
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2117,
      "doc_norm": 1.4964,
      "encoder_q-embeddings": 1284.9346,
      "encoder_q-layer.0": 888.4434,
      "encoder_q-layer.1": 973.6556,
      "encoder_q-layer.10": 1151.1753,
      "encoder_q-layer.11": 2691.5386,
      "encoder_q-layer.2": 1115.3248,
      "encoder_q-layer.3": 1163.0298,
      "encoder_q-layer.4": 1317.2156,
      "encoder_q-layer.5": 1408.1738,
      "encoder_q-layer.6": 1404.9594,
      "encoder_q-layer.7": 1536.6235,
      "encoder_q-layer.8": 1497.9653,
      "encoder_q-layer.9": 1264.4813,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4215,
      "inbatch_pos_score": 1.0889,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 3.2117,
      "norm_diff": 0.0543,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2155.1555,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4214,
      "query_norm": 1.4422,
      "queue_k_norm": 1.4961,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9142,
      "sent_len_1": 66.6321,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.815,
      "stdk": 0.0488,
      "stdq": 0.0455,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 62500
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1974,
      "doc_norm": 1.5024,
      "encoder_q-embeddings": 972.1968,
      "encoder_q-layer.0": 604.8071,
      "encoder_q-layer.1": 626.9871,
      "encoder_q-layer.10": 1267.0361,
      "encoder_q-layer.11": 2852.6484,
      "encoder_q-layer.2": 698.5737,
      "encoder_q-layer.3": 747.9437,
      "encoder_q-layer.4": 821.0933,
      "encoder_q-layer.5": 857.2014,
      "encoder_q-layer.6": 975.0208,
      "encoder_q-layer.7": 1155.129,
      "encoder_q-layer.8": 1382.7961,
      "encoder_q-layer.9": 1276.4304,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4251,
      "inbatch_pos_score": 1.1191,
      "learning_rate": 2.077777777777778e-05,
      "loss": 3.1974,
      "norm_diff": 0.043,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1860.4602,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.425,
      "query_norm": 1.4594,
      "queue_k_norm": 1.4965,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7652,
      "sent_len_1": 66.78,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.15,
      "stdk": 0.0491,
      "stdq": 0.046,
      "stdqueue_k": 0.0489,
      "stdqueue_q": 0.0,
      "step": 62600
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2333,
      "doc_norm": 1.5016,
      "encoder_q-embeddings": 2287.4858,
      "encoder_q-layer.0": 1551.5148,
      "encoder_q-layer.1": 1715.923,
      "encoder_q-layer.10": 1175.1272,
      "encoder_q-layer.11": 2816.4749,
      "encoder_q-layer.2": 2148.8792,
      "encoder_q-layer.3": 2292.4656,
      "encoder_q-layer.4": 2581.2595,
      "encoder_q-layer.5": 2500.4285,
      "encoder_q-layer.6": 2404.1299,
      "encoder_q-layer.7": 2555.3555,
      "encoder_q-layer.8": 2204.3013,
      "encoder_q-layer.9": 1307.9495,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4303,
      "inbatch_pos_score": 1.0791,
      "learning_rate": 2.0722222222222224e-05,
      "loss": 3.2333,
      "norm_diff": 0.0718,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3331.9063,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4304,
      "query_norm": 1.4298,
      "queue_k_norm": 1.4982,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8321,
      "sent_len_1": 66.7063,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.44,
      "stdk": 0.049,
      "stdq": 0.0446,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 62700
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2345,
      "doc_norm": 1.4982,
      "encoder_q-embeddings": 1729.9048,
      "encoder_q-layer.0": 1197.7823,
      "encoder_q-layer.1": 1493.4011,
      "encoder_q-layer.10": 1212.3589,
      "encoder_q-layer.11": 2664.459,
      "encoder_q-layer.2": 1738.3627,
      "encoder_q-layer.3": 1985.0439,
      "encoder_q-layer.4": 2204.0933,
      "encoder_q-layer.5": 2195.3096,
      "encoder_q-layer.6": 1984.2692,
      "encoder_q-layer.7": 1922.6267,
      "encoder_q-layer.8": 1577.6254,
      "encoder_q-layer.9": 1205.1179,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4344,
      "inbatch_pos_score": 1.124,
      "learning_rate": 2.0666666666666666e-05,
      "loss": 3.2345,
      "norm_diff": 0.0416,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2757.0295,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4341,
      "query_norm": 1.4566,
      "queue_k_norm": 1.4996,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9678,
      "sent_len_1": 66.958,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8975,
      "stdk": 0.0488,
      "stdq": 0.0457,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 62800
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2015,
      "doc_norm": 1.4999,
      "encoder_q-embeddings": 1326.7324,
      "encoder_q-layer.0": 859.2421,
      "encoder_q-layer.1": 964.8327,
      "encoder_q-layer.10": 1247.1624,
      "encoder_q-layer.11": 2750.4072,
      "encoder_q-layer.2": 1058.7524,
      "encoder_q-layer.3": 1097.8202,
      "encoder_q-layer.4": 1165.1726,
      "encoder_q-layer.5": 1359.0452,
      "encoder_q-layer.6": 1264.8317,
      "encoder_q-layer.7": 1442.1046,
      "encoder_q-layer.8": 1473.5322,
      "encoder_q-layer.9": 1300.3636,
      "epoch": 0.61,
      "inbatch_neg_score": 0.4384,
      "inbatch_pos_score": 1.1416,
      "learning_rate": 2.0611111111111112e-05,
      "loss": 3.2015,
      "norm_diff": 0.0223,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2147.8098,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4375,
      "query_norm": 1.4776,
      "queue_k_norm": 1.5002,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7547,
      "sent_len_1": 66.724,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5112,
      "stdk": 0.0489,
      "stdq": 0.0465,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 62900
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2297,
      "doc_norm": 1.5033,
      "encoder_q-embeddings": 1064.373,
      "encoder_q-layer.0": 705.0458,
      "encoder_q-layer.1": 765.2467,
      "encoder_q-layer.10": 1254.6095,
      "encoder_q-layer.11": 2679.2847,
      "encoder_q-layer.2": 927.4606,
      "encoder_q-layer.3": 998.3497,
      "encoder_q-layer.4": 1073.304,
      "encoder_q-layer.5": 1054.4171,
      "encoder_q-layer.6": 1096.2102,
      "encoder_q-layer.7": 1212.5695,
      "encoder_q-layer.8": 1386.8589,
      "encoder_q-layer.9": 1246.0183,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4431,
      "inbatch_pos_score": 1.1172,
      "learning_rate": 2.0555555555555555e-05,
      "loss": 3.2297,
      "norm_diff": 0.0397,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1931.7619,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4417,
      "query_norm": 1.4636,
      "queue_k_norm": 1.5005,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7349,
      "sent_len_1": 66.5433,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0263,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63000
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2204,
      "doc_norm": 1.5114,
      "encoder_q-embeddings": 2261.2258,
      "encoder_q-layer.0": 1650.2686,
      "encoder_q-layer.1": 2124.5676,
      "encoder_q-layer.10": 1344.9431,
      "encoder_q-layer.11": 2778.3564,
      "encoder_q-layer.2": 2336.3079,
      "encoder_q-layer.3": 2567.873,
      "encoder_q-layer.4": 2779.667,
      "encoder_q-layer.5": 3090.8225,
      "encoder_q-layer.6": 2814.8445,
      "encoder_q-layer.7": 2709.7156,
      "encoder_q-layer.8": 1807.3397,
      "encoder_q-layer.9": 1322.5043,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4418,
      "inbatch_pos_score": 1.1045,
      "learning_rate": 2.05e-05,
      "loss": 3.2204,
      "norm_diff": 0.0604,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3481.8541,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4409,
      "query_norm": 1.451,
      "queue_k_norm": 1.5015,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1385,
      "sent_len_1": 67.0171,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.015,
      "stdk": 0.0493,
      "stdq": 0.0454,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63100
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1959,
      "doc_norm": 1.5041,
      "encoder_q-embeddings": 480.6876,
      "encoder_q-layer.0": 303.5064,
      "encoder_q-layer.1": 322.6266,
      "encoder_q-layer.10": 692.1949,
      "encoder_q-layer.11": 1428.5521,
      "encoder_q-layer.2": 361.5616,
      "encoder_q-layer.3": 370.0077,
      "encoder_q-layer.4": 409.437,
      "encoder_q-layer.5": 424.8144,
      "encoder_q-layer.6": 491.5599,
      "encoder_q-layer.7": 580.7375,
      "encoder_q-layer.8": 679.5379,
      "encoder_q-layer.9": 623.8641,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4447,
      "inbatch_pos_score": 1.1084,
      "learning_rate": 2.0444444444444446e-05,
      "loss": 3.1959,
      "norm_diff": 0.0485,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 942.4431,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4456,
      "query_norm": 1.4556,
      "queue_k_norm": 1.5028,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8059,
      "sent_len_1": 66.7886,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4025,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63200
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2132,
      "doc_norm": 1.5063,
      "encoder_q-embeddings": 523.9493,
      "encoder_q-layer.0": 327.3377,
      "encoder_q-layer.1": 351.7188,
      "encoder_q-layer.10": 580.0222,
      "encoder_q-layer.11": 1315.1227,
      "encoder_q-layer.2": 390.2125,
      "encoder_q-layer.3": 397.5548,
      "encoder_q-layer.4": 423.2117,
      "encoder_q-layer.5": 443.4863,
      "encoder_q-layer.6": 504.9615,
      "encoder_q-layer.7": 547.2618,
      "encoder_q-layer.8": 658.8247,
      "encoder_q-layer.9": 589.1041,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4488,
      "inbatch_pos_score": 1.1523,
      "learning_rate": 2.0388888888888892e-05,
      "loss": 3.2132,
      "norm_diff": 0.025,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 908.8394,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4497,
      "query_norm": 1.4813,
      "queue_k_norm": 1.5029,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9335,
      "sent_len_1": 66.9307,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2525,
      "stdk": 0.0491,
      "stdq": 0.0467,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63300
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2201,
      "doc_norm": 1.5096,
      "encoder_q-embeddings": 843.6217,
      "encoder_q-layer.0": 566.1784,
      "encoder_q-layer.1": 656.5072,
      "encoder_q-layer.10": 639.5671,
      "encoder_q-layer.11": 1327.6091,
      "encoder_q-layer.2": 768.0009,
      "encoder_q-layer.3": 773.4526,
      "encoder_q-layer.4": 863.4234,
      "encoder_q-layer.5": 887.1375,
      "encoder_q-layer.6": 894.7137,
      "encoder_q-layer.7": 887.4065,
      "encoder_q-layer.8": 866.2281,
      "encoder_q-layer.9": 625.6284,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4545,
      "inbatch_pos_score": 1.168,
      "learning_rate": 2.0333333333333334e-05,
      "loss": 3.2201,
      "norm_diff": 0.027,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1258.7335,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4548,
      "query_norm": 1.4826,
      "queue_k_norm": 1.5035,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9733,
      "sent_len_1": 66.6381,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.455,
      "stdk": 0.0491,
      "stdq": 0.0467,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63400
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2069,
      "doc_norm": 1.4964,
      "encoder_q-embeddings": 556.9018,
      "encoder_q-layer.0": 350.787,
      "encoder_q-layer.1": 368.4693,
      "encoder_q-layer.10": 561.0217,
      "encoder_q-layer.11": 1307.9109,
      "encoder_q-layer.2": 411.245,
      "encoder_q-layer.3": 442.6608,
      "encoder_q-layer.4": 488.1336,
      "encoder_q-layer.5": 491.3095,
      "encoder_q-layer.6": 532.9182,
      "encoder_q-layer.7": 616.727,
      "encoder_q-layer.8": 656.4955,
      "encoder_q-layer.9": 575.3683,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4567,
      "inbatch_pos_score": 1.0928,
      "learning_rate": 2.027777777777778e-05,
      "loss": 3.2069,
      "norm_diff": 0.0586,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 941.7969,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4556,
      "query_norm": 1.4378,
      "queue_k_norm": 1.5052,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8882,
      "sent_len_1": 66.7186,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.875,
      "stdk": 0.0486,
      "stdq": 0.0449,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63500
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1986,
      "doc_norm": 1.5064,
      "encoder_q-embeddings": 911.6083,
      "encoder_q-layer.0": 640.2222,
      "encoder_q-layer.1": 677.6674,
      "encoder_q-layer.10": 667.524,
      "encoder_q-layer.11": 1528.8448,
      "encoder_q-layer.2": 771.4194,
      "encoder_q-layer.3": 814.3874,
      "encoder_q-layer.4": 885.0862,
      "encoder_q-layer.5": 902.5883,
      "encoder_q-layer.6": 1075.2175,
      "encoder_q-layer.7": 1099.9139,
      "encoder_q-layer.8": 964.2654,
      "encoder_q-layer.9": 690.4788,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4563,
      "inbatch_pos_score": 1.0977,
      "learning_rate": 2.0222222222222222e-05,
      "loss": 3.1986,
      "norm_diff": 0.0585,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1393.8876,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4561,
      "query_norm": 1.4479,
      "queue_k_norm": 1.5064,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8718,
      "sent_len_1": 66.7826,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.3462,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63600
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2007,
      "doc_norm": 1.5052,
      "encoder_q-embeddings": 674.227,
      "encoder_q-layer.0": 422.6924,
      "encoder_q-layer.1": 476.8215,
      "encoder_q-layer.10": 589.3983,
      "encoder_q-layer.11": 1401.9501,
      "encoder_q-layer.2": 541.6241,
      "encoder_q-layer.3": 576.86,
      "encoder_q-layer.4": 657.3491,
      "encoder_q-layer.5": 653.478,
      "encoder_q-layer.6": 703.026,
      "encoder_q-layer.7": 835.7457,
      "encoder_q-layer.8": 793.9247,
      "encoder_q-layer.9": 635.5317,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4565,
      "inbatch_pos_score": 1.1562,
      "learning_rate": 2.0166666666666668e-05,
      "loss": 3.2007,
      "norm_diff": 0.0401,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1092.3994,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4558,
      "query_norm": 1.4652,
      "queue_k_norm": 1.5067,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8207,
      "sent_len_1": 66.6387,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0037,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63700
    },
    {
      "accuracy": 52.5391,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2089,
      "doc_norm": 1.5083,
      "encoder_q-embeddings": 600.0281,
      "encoder_q-layer.0": 394.0482,
      "encoder_q-layer.1": 419.2438,
      "encoder_q-layer.10": 609.1521,
      "encoder_q-layer.11": 1391.829,
      "encoder_q-layer.2": 473.6771,
      "encoder_q-layer.3": 522.102,
      "encoder_q-layer.4": 546.5981,
      "encoder_q-layer.5": 578.3675,
      "encoder_q-layer.6": 620.463,
      "encoder_q-layer.7": 681.126,
      "encoder_q-layer.8": 765.634,
      "encoder_q-layer.9": 634.3867,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4613,
      "inbatch_pos_score": 1.1074,
      "learning_rate": 2.011111111111111e-05,
      "loss": 3.2089,
      "norm_diff": 0.0584,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1029.0174,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4614,
      "query_norm": 1.4499,
      "queue_k_norm": 1.5084,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6669,
      "sent_len_1": 66.5641,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9963,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 63800
    },
    {
      "accuracy": 59.375,
      "active_queue_size": 16384.0,
      "cl_loss": 3.211,
      "doc_norm": 1.5094,
      "encoder_q-embeddings": 711.7576,
      "encoder_q-layer.0": 488.8303,
      "encoder_q-layer.1": 564.585,
      "encoder_q-layer.10": 609.0189,
      "encoder_q-layer.11": 1321.2111,
      "encoder_q-layer.2": 672.1862,
      "encoder_q-layer.3": 758.0822,
      "encoder_q-layer.4": 834.1042,
      "encoder_q-layer.5": 816.5254,
      "encoder_q-layer.6": 838.6954,
      "encoder_q-layer.7": 898.8941,
      "encoder_q-layer.8": 930.4572,
      "encoder_q-layer.9": 690.8707,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4624,
      "inbatch_pos_score": 1.1748,
      "learning_rate": 2.0055555555555556e-05,
      "loss": 3.211,
      "norm_diff": 0.0476,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1205.3439,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4622,
      "query_norm": 1.4618,
      "queue_k_norm": 1.5061,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0139,
      "sent_len_1": 66.7167,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0387,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 63900
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.212,
      "doc_norm": 1.5077,
      "encoder_q-embeddings": 1269.094,
      "encoder_q-layer.0": 858.0757,
      "encoder_q-layer.1": 927.2702,
      "encoder_q-layer.10": 664.4972,
      "encoder_q-layer.11": 1392.3123,
      "encoder_q-layer.2": 1104.6333,
      "encoder_q-layer.3": 1166.6978,
      "encoder_q-layer.4": 1207.4374,
      "encoder_q-layer.5": 1229.6377,
      "encoder_q-layer.6": 1181.5341,
      "encoder_q-layer.7": 1171.1167,
      "encoder_q-layer.8": 923.1675,
      "encoder_q-layer.9": 661.6976,
      "epoch": 0.62,
      "inbatch_neg_score": 0.4686,
      "inbatch_pos_score": 1.1387,
      "learning_rate": 2e-05,
      "loss": 3.212,
      "norm_diff": 0.0485,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1645.4711,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4675,
      "query_norm": 1.4592,
      "queue_k_norm": 1.5109,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.158,
      "sent_len_1": 67.0109,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.71,
      "stdk": 0.0489,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 64000
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.224,
      "doc_norm": 1.5097,
      "encoder_q-embeddings": 771.7272,
      "encoder_q-layer.0": 543.55,
      "encoder_q-layer.1": 656.2346,
      "encoder_q-layer.10": 688.826,
      "encoder_q-layer.11": 1427.542,
      "encoder_q-layer.2": 827.6564,
      "encoder_q-layer.3": 879.9137,
      "encoder_q-layer.4": 931.7577,
      "encoder_q-layer.5": 877.5795,
      "encoder_q-layer.6": 847.3301,
      "encoder_q-layer.7": 865.303,
      "encoder_q-layer.8": 809.3295,
      "encoder_q-layer.9": 657.1165,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4671,
      "inbatch_pos_score": 1.1328,
      "learning_rate": 1.9944444444444447e-05,
      "loss": 3.224,
      "norm_diff": 0.0547,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1302.0413,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4666,
      "query_norm": 1.455,
      "queue_k_norm": 1.5103,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9727,
      "sent_len_1": 66.7153,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8125,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 64100
    },
    {
      "accuracy": 59.9609,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2053,
      "doc_norm": 1.5191,
      "encoder_q-embeddings": 1944.0,
      "encoder_q-layer.0": 1337.2644,
      "encoder_q-layer.1": 1520.7407,
      "encoder_q-layer.10": 723.5491,
      "encoder_q-layer.11": 1433.5845,
      "encoder_q-layer.2": 1743.6276,
      "encoder_q-layer.3": 1823.7217,
      "encoder_q-layer.4": 1807.4767,
      "encoder_q-layer.5": 1755.0422,
      "encoder_q-layer.6": 1698.0742,
      "encoder_q-layer.7": 2009.9667,
      "encoder_q-layer.8": 1675.8463,
      "encoder_q-layer.9": 855.4949,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4673,
      "inbatch_pos_score": 1.1699,
      "learning_rate": 1.988888888888889e-05,
      "loss": 3.2053,
      "norm_diff": 0.052,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2436.5761,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4668,
      "query_norm": 1.4671,
      "queue_k_norm": 1.5116,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7739,
      "sent_len_1": 66.6704,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.2337,
      "stdk": 0.0493,
      "stdq": 0.046,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 64200
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2089,
      "doc_norm": 1.5104,
      "encoder_q-embeddings": 1257.9249,
      "encoder_q-layer.0": 882.6318,
      "encoder_q-layer.1": 960.139,
      "encoder_q-layer.10": 633.6817,
      "encoder_q-layer.11": 1405.3759,
      "encoder_q-layer.2": 1145.9072,
      "encoder_q-layer.3": 1210.4412,
      "encoder_q-layer.4": 1365.4729,
      "encoder_q-layer.5": 1385.9846,
      "encoder_q-layer.6": 1297.0131,
      "encoder_q-layer.7": 1589.5129,
      "encoder_q-layer.8": 1214.0079,
      "encoder_q-layer.9": 766.0934,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4675,
      "inbatch_pos_score": 1.0859,
      "learning_rate": 1.9833333333333335e-05,
      "loss": 3.2089,
      "norm_diff": 0.0729,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1779.7827,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4663,
      "query_norm": 1.4375,
      "queue_k_norm": 1.5107,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8741,
      "sent_len_1": 66.7565,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0662,
      "stdk": 0.0489,
      "stdq": 0.0449,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 64300
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2032,
      "doc_norm": 1.5109,
      "encoder_q-embeddings": 550.6813,
      "encoder_q-layer.0": 361.2469,
      "encoder_q-layer.1": 392.1038,
      "encoder_q-layer.10": 629.7963,
      "encoder_q-layer.11": 1381.0402,
      "encoder_q-layer.2": 450.0116,
      "encoder_q-layer.3": 489.3995,
      "encoder_q-layer.4": 536.3806,
      "encoder_q-layer.5": 566.444,
      "encoder_q-layer.6": 633.9871,
      "encoder_q-layer.7": 712.8771,
      "encoder_q-layer.8": 798.7063,
      "encoder_q-layer.9": 660.7139,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4707,
      "inbatch_pos_score": 1.1309,
      "learning_rate": 1.9777777777777778e-05,
      "loss": 3.2032,
      "norm_diff": 0.0615,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1003.9123,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4705,
      "query_norm": 1.4493,
      "queue_k_norm": 1.5112,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9581,
      "sent_len_1": 66.8672,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.365,
      "stdk": 0.0489,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 64400
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.203,
      "doc_norm": 1.5096,
      "encoder_q-embeddings": 1212.8392,
      "encoder_q-layer.0": 895.9315,
      "encoder_q-layer.1": 1070.6514,
      "encoder_q-layer.10": 617.8488,
      "encoder_q-layer.11": 1364.6188,
      "encoder_q-layer.2": 1239.0651,
      "encoder_q-layer.3": 1251.7141,
      "encoder_q-layer.4": 1319.8737,
      "encoder_q-layer.5": 1438.1315,
      "encoder_q-layer.6": 1318.3118,
      "encoder_q-layer.7": 1294.7728,
      "encoder_q-layer.8": 1062.6543,
      "encoder_q-layer.9": 642.1147,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4585,
      "inbatch_pos_score": 1.1094,
      "learning_rate": 1.9722222222222224e-05,
      "loss": 3.203,
      "norm_diff": 0.0537,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1738.5892,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4587,
      "query_norm": 1.4559,
      "queue_k_norm": 1.5118,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.85,
      "sent_len_1": 66.6649,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2363,
      "stdk": 0.0488,
      "stdq": 0.0459,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 64500
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.187,
      "doc_norm": 1.5124,
      "encoder_q-embeddings": 593.7913,
      "encoder_q-layer.0": 399.5384,
      "encoder_q-layer.1": 434.3231,
      "encoder_q-layer.10": 656.9315,
      "encoder_q-layer.11": 1479.5264,
      "encoder_q-layer.2": 493.2334,
      "encoder_q-layer.3": 540.7809,
      "encoder_q-layer.4": 618.0341,
      "encoder_q-layer.5": 653.454,
      "encoder_q-layer.6": 715.1303,
      "encoder_q-layer.7": 744.5489,
      "encoder_q-layer.8": 801.8606,
      "encoder_q-layer.9": 667.8266,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4573,
      "inbatch_pos_score": 1.1035,
      "learning_rate": 1.9666666666666666e-05,
      "loss": 3.187,
      "norm_diff": 0.072,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1095.4615,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4585,
      "query_norm": 1.4404,
      "queue_k_norm": 1.5128,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9137,
      "sent_len_1": 66.5645,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.3825,
      "stdk": 0.049,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 64600
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2008,
      "doc_norm": 1.5149,
      "encoder_q-embeddings": 567.1237,
      "encoder_q-layer.0": 381.5062,
      "encoder_q-layer.1": 435.4225,
      "encoder_q-layer.10": 652.3917,
      "encoder_q-layer.11": 1310.0652,
      "encoder_q-layer.2": 528.6222,
      "encoder_q-layer.3": 551.6496,
      "encoder_q-layer.4": 582.7476,
      "encoder_q-layer.5": 625.6315,
      "encoder_q-layer.6": 640.0785,
      "encoder_q-layer.7": 682.5546,
      "encoder_q-layer.8": 753.3686,
      "encoder_q-layer.9": 601.1358,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4639,
      "inbatch_pos_score": 1.1396,
      "learning_rate": 1.9611111111111115e-05,
      "loss": 3.2008,
      "norm_diff": 0.0684,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1017.4339,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4631,
      "query_norm": 1.4464,
      "queue_k_norm": 1.5104,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0073,
      "sent_len_1": 66.764,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4963,
      "stdk": 0.049,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 64700
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2055,
      "doc_norm": 1.517,
      "encoder_q-embeddings": 487.0407,
      "encoder_q-layer.0": 316.7354,
      "encoder_q-layer.1": 324.119,
      "encoder_q-layer.10": 580.6937,
      "encoder_q-layer.11": 1329.3837,
      "encoder_q-layer.2": 367.2446,
      "encoder_q-layer.3": 398.5673,
      "encoder_q-layer.4": 424.7337,
      "encoder_q-layer.5": 435.5526,
      "encoder_q-layer.6": 509.0636,
      "encoder_q-layer.7": 647.3821,
      "encoder_q-layer.8": 697.4836,
      "encoder_q-layer.9": 602.6177,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4584,
      "inbatch_pos_score": 1.1494,
      "learning_rate": 1.9555555555555557e-05,
      "loss": 3.2055,
      "norm_diff": 0.0512,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 922.9053,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4592,
      "query_norm": 1.4659,
      "queue_k_norm": 1.5133,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9706,
      "sent_len_1": 66.8002,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.445,
      "stdk": 0.0491,
      "stdq": 0.0462,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 64800
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1913,
      "doc_norm": 1.5149,
      "encoder_q-embeddings": 808.1993,
      "encoder_q-layer.0": 535.0325,
      "encoder_q-layer.1": 574.3803,
      "encoder_q-layer.10": 614.5933,
      "encoder_q-layer.11": 1334.4319,
      "encoder_q-layer.2": 634.9188,
      "encoder_q-layer.3": 683.2802,
      "encoder_q-layer.4": 694.9596,
      "encoder_q-layer.5": 705.1801,
      "encoder_q-layer.6": 703.7319,
      "encoder_q-layer.7": 767.8749,
      "encoder_q-layer.8": 798.6143,
      "encoder_q-layer.9": 625.6681,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4587,
      "inbatch_pos_score": 1.126,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 3.1913,
      "norm_diff": 0.0641,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1145.8072,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4607,
      "query_norm": 1.4509,
      "queue_k_norm": 1.5122,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8277,
      "sent_len_1": 66.711,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.385,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 64900
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2264,
      "doc_norm": 1.5144,
      "encoder_q-embeddings": 2018.0134,
      "encoder_q-layer.0": 1277.7849,
      "encoder_q-layer.1": 1403.9498,
      "encoder_q-layer.10": 622.3005,
      "encoder_q-layer.11": 1358.184,
      "encoder_q-layer.2": 1753.6144,
      "encoder_q-layer.3": 1818.0968,
      "encoder_q-layer.4": 1963.3871,
      "encoder_q-layer.5": 1638.2202,
      "encoder_q-layer.6": 1883.2385,
      "encoder_q-layer.7": 1607.8263,
      "encoder_q-layer.8": 1160.9039,
      "encoder_q-layer.9": 649.7268,
      "epoch": 0.63,
      "inbatch_neg_score": 0.4583,
      "inbatch_pos_score": 1.123,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 3.2264,
      "norm_diff": 0.0741,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2292.0188,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.459,
      "query_norm": 1.4403,
      "queue_k_norm": 1.5131,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0258,
      "sent_len_1": 66.7768,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5425,
      "stdk": 0.049,
      "stdq": 0.0451,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 65000
    },
    {
      "accuracy": 58.6914,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1921,
      "doc_norm": 1.5118,
      "encoder_q-embeddings": 455.7672,
      "encoder_q-layer.0": 298.7204,
      "encoder_q-layer.1": 314.0677,
      "encoder_q-layer.10": 564.5554,
      "encoder_q-layer.11": 1242.8503,
      "encoder_q-layer.2": 350.0052,
      "encoder_q-layer.3": 368.6246,
      "encoder_q-layer.4": 402.666,
      "encoder_q-layer.5": 418.4878,
      "encoder_q-layer.6": 482.6115,
      "encoder_q-layer.7": 558.2763,
      "encoder_q-layer.8": 661.2522,
      "encoder_q-layer.9": 584.7864,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4591,
      "inbatch_pos_score": 1.1738,
      "learning_rate": 1.938888888888889e-05,
      "loss": 3.1921,
      "norm_diff": 0.0459,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 862.6619,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4592,
      "query_norm": 1.4658,
      "queue_k_norm": 1.5132,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9041,
      "sent_len_1": 66.8826,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.895,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 65100
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2211,
      "doc_norm": 1.5096,
      "encoder_q-embeddings": 1086.205,
      "encoder_q-layer.0": 701.2986,
      "encoder_q-layer.1": 748.4805,
      "encoder_q-layer.10": 1249.3127,
      "encoder_q-layer.11": 2762.9712,
      "encoder_q-layer.2": 834.8705,
      "encoder_q-layer.3": 878.8942,
      "encoder_q-layer.4": 897.5848,
      "encoder_q-layer.5": 950.7848,
      "encoder_q-layer.6": 1083.2183,
      "encoder_q-layer.7": 1269.5105,
      "encoder_q-layer.8": 1397.3561,
      "encoder_q-layer.9": 1241.2644,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4623,
      "inbatch_pos_score": 1.1348,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 3.2211,
      "norm_diff": 0.0363,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1939.515,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4622,
      "query_norm": 1.4733,
      "queue_k_norm": 1.5144,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9738,
      "sent_len_1": 66.8409,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9288,
      "stdk": 0.0488,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 65200
    },
    {
      "accuracy": 52.9297,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2288,
      "doc_norm": 1.5191,
      "encoder_q-embeddings": 1118.6464,
      "encoder_q-layer.0": 741.4786,
      "encoder_q-layer.1": 739.3354,
      "encoder_q-layer.10": 1206.8665,
      "encoder_q-layer.11": 2817.3152,
      "encoder_q-layer.2": 848.1968,
      "encoder_q-layer.3": 881.0345,
      "encoder_q-layer.4": 951.6905,
      "encoder_q-layer.5": 1005.3029,
      "encoder_q-layer.6": 1086.5719,
      "encoder_q-layer.7": 1279.7164,
      "encoder_q-layer.8": 1498.7537,
      "encoder_q-layer.9": 1255.3851,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4615,
      "inbatch_pos_score": 1.1025,
      "learning_rate": 1.927777777777778e-05,
      "loss": 3.2288,
      "norm_diff": 0.067,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2004.2184,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4617,
      "query_norm": 1.4521,
      "queue_k_norm": 1.5141,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8397,
      "sent_len_1": 66.6238,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0712,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 65300
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1966,
      "doc_norm": 1.5147,
      "encoder_q-embeddings": 1020.0449,
      "encoder_q-layer.0": 676.9633,
      "encoder_q-layer.1": 701.8994,
      "encoder_q-layer.10": 1187.7896,
      "encoder_q-layer.11": 2600.6587,
      "encoder_q-layer.2": 751.5262,
      "encoder_q-layer.3": 800.4385,
      "encoder_q-layer.4": 837.0302,
      "encoder_q-layer.5": 877.6196,
      "encoder_q-layer.6": 988.828,
      "encoder_q-layer.7": 1171.6836,
      "encoder_q-layer.8": 1348.8466,
      "encoder_q-layer.9": 1222.1935,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4592,
      "inbatch_pos_score": 1.127,
      "learning_rate": 1.922222222222222e-05,
      "loss": 3.1966,
      "norm_diff": 0.0501,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1820.9512,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4583,
      "query_norm": 1.4645,
      "queue_k_norm": 1.5138,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6336,
      "sent_len_1": 66.912,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3313,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 65400
    },
    {
      "accuracy": 58.0078,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2329,
      "doc_norm": 1.5122,
      "encoder_q-embeddings": 1339.879,
      "encoder_q-layer.0": 967.7212,
      "encoder_q-layer.1": 1020.1147,
      "encoder_q-layer.10": 1268.7947,
      "encoder_q-layer.11": 2592.2009,
      "encoder_q-layer.2": 1147.6774,
      "encoder_q-layer.3": 1139.8834,
      "encoder_q-layer.4": 1217.2393,
      "encoder_q-layer.5": 1258.157,
      "encoder_q-layer.6": 1203.0472,
      "encoder_q-layer.7": 1516.632,
      "encoder_q-layer.8": 1521.3065,
      "encoder_q-layer.9": 1323.3409,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4606,
      "inbatch_pos_score": 1.1494,
      "learning_rate": 1.9166666666666667e-05,
      "loss": 3.2329,
      "norm_diff": 0.0583,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2120.8892,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4609,
      "query_norm": 1.4539,
      "queue_k_norm": 1.513,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8226,
      "sent_len_1": 66.488,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.125,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 65500
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1932,
      "doc_norm": 1.5138,
      "encoder_q-embeddings": 1292.3354,
      "encoder_q-layer.0": 870.5707,
      "encoder_q-layer.1": 999.8556,
      "encoder_q-layer.10": 1254.1738,
      "encoder_q-layer.11": 2659.9363,
      "encoder_q-layer.2": 1099.4171,
      "encoder_q-layer.3": 1072.0564,
      "encoder_q-layer.4": 1130.0287,
      "encoder_q-layer.5": 1153.483,
      "encoder_q-layer.6": 1173.7854,
      "encoder_q-layer.7": 1299.562,
      "encoder_q-layer.8": 1387.2288,
      "encoder_q-layer.9": 1158.787,
      "epoch": 0.64,
      "inbatch_neg_score": 0.462,
      "inbatch_pos_score": 1.1406,
      "learning_rate": 1.9111111111111113e-05,
      "loss": 3.1932,
      "norm_diff": 0.0533,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2055.8034,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4614,
      "query_norm": 1.4605,
      "queue_k_norm": 1.5132,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9809,
      "sent_len_1": 66.6673,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6887,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 65600
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1919,
      "doc_norm": 1.5138,
      "encoder_q-embeddings": 1038.4542,
      "encoder_q-layer.0": 680.0045,
      "encoder_q-layer.1": 719.8704,
      "encoder_q-layer.10": 1164.9222,
      "encoder_q-layer.11": 2636.7117,
      "encoder_q-layer.2": 795.0626,
      "encoder_q-layer.3": 805.3928,
      "encoder_q-layer.4": 875.8926,
      "encoder_q-layer.5": 864.9722,
      "encoder_q-layer.6": 1008.7755,
      "encoder_q-layer.7": 1178.8743,
      "encoder_q-layer.8": 1324.0789,
      "encoder_q-layer.9": 1226.0549,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4655,
      "inbatch_pos_score": 1.1807,
      "learning_rate": 1.905555555555556e-05,
      "loss": 3.1919,
      "norm_diff": 0.0379,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1832.82,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4663,
      "query_norm": 1.4759,
      "queue_k_norm": 1.5164,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0796,
      "sent_len_1": 67.0481,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8988,
      "stdk": 0.0489,
      "stdq": 0.0463,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 65700
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.191,
      "doc_norm": 1.5197,
      "encoder_q-embeddings": 1053.7256,
      "encoder_q-layer.0": 696.8502,
      "encoder_q-layer.1": 748.8439,
      "encoder_q-layer.10": 1172.1399,
      "encoder_q-layer.11": 2630.1079,
      "encoder_q-layer.2": 829.3132,
      "encoder_q-layer.3": 911.3969,
      "encoder_q-layer.4": 954.4534,
      "encoder_q-layer.5": 989.6992,
      "encoder_q-layer.6": 1115.7574,
      "encoder_q-layer.7": 1229.374,
      "encoder_q-layer.8": 1357.3462,
      "encoder_q-layer.9": 1178.4222,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4648,
      "inbatch_pos_score": 1.1211,
      "learning_rate": 1.9e-05,
      "loss": 3.191,
      "norm_diff": 0.0654,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1872.2253,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4651,
      "query_norm": 1.4544,
      "queue_k_norm": 1.5157,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8972,
      "sent_len_1": 66.6606,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.37,
      "stdk": 0.0492,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 65800
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2096,
      "doc_norm": 1.5133,
      "encoder_q-embeddings": 1259.8729,
      "encoder_q-layer.0": 866.8186,
      "encoder_q-layer.1": 980.8087,
      "encoder_q-layer.10": 1197.1774,
      "encoder_q-layer.11": 2756.1487,
      "encoder_q-layer.2": 1156.9851,
      "encoder_q-layer.3": 1193.9724,
      "encoder_q-layer.4": 1240.1199,
      "encoder_q-layer.5": 1273.9299,
      "encoder_q-layer.6": 1434.4025,
      "encoder_q-layer.7": 1562.8062,
      "encoder_q-layer.8": 1552.8109,
      "encoder_q-layer.9": 1285.5798,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4663,
      "inbatch_pos_score": 1.1494,
      "learning_rate": 1.8944444444444447e-05,
      "loss": 3.2096,
      "norm_diff": 0.0277,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2175.7082,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4675,
      "query_norm": 1.4856,
      "queue_k_norm": 1.5143,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9268,
      "sent_len_1": 66.6605,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7587,
      "stdk": 0.0489,
      "stdq": 0.0466,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 65900
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1955,
      "doc_norm": 1.5114,
      "encoder_q-embeddings": 1156.0897,
      "encoder_q-layer.0": 753.3243,
      "encoder_q-layer.1": 820.4401,
      "encoder_q-layer.10": 1363.2491,
      "encoder_q-layer.11": 2828.3379,
      "encoder_q-layer.2": 933.7761,
      "encoder_q-layer.3": 1011.9638,
      "encoder_q-layer.4": 1085.7136,
      "encoder_q-layer.5": 1119.6705,
      "encoder_q-layer.6": 1221.006,
      "encoder_q-layer.7": 1402.0835,
      "encoder_q-layer.8": 1547.3834,
      "encoder_q-layer.9": 1347.7157,
      "epoch": 0.64,
      "inbatch_neg_score": 0.4672,
      "inbatch_pos_score": 1.1504,
      "learning_rate": 1.888888888888889e-05,
      "loss": 3.1955,
      "norm_diff": 0.0367,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2069.1406,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4668,
      "query_norm": 1.4747,
      "queue_k_norm": 1.5145,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0198,
      "sent_len_1": 66.697,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7925,
      "stdk": 0.0488,
      "stdq": 0.0462,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 66000
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2427,
      "doc_norm": 1.5167,
      "encoder_q-embeddings": 1266.3644,
      "encoder_q-layer.0": 814.5183,
      "encoder_q-layer.1": 898.7255,
      "encoder_q-layer.10": 1141.4359,
      "encoder_q-layer.11": 2633.9773,
      "encoder_q-layer.2": 1046.2938,
      "encoder_q-layer.3": 1077.5872,
      "encoder_q-layer.4": 1131.343,
      "encoder_q-layer.5": 1196.8231,
      "encoder_q-layer.6": 1391.1794,
      "encoder_q-layer.7": 1652.8013,
      "encoder_q-layer.8": 1891.2009,
      "encoder_q-layer.9": 1503.0392,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4727,
      "inbatch_pos_score": 1.1348,
      "learning_rate": 1.8833333333333335e-05,
      "loss": 3.2427,
      "norm_diff": 0.0659,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2163.0003,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4722,
      "query_norm": 1.4508,
      "queue_k_norm": 1.5155,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0807,
      "sent_len_1": 66.5189,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6813,
      "stdk": 0.049,
      "stdq": 0.0452,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 66100
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2101,
      "doc_norm": 1.5163,
      "encoder_q-embeddings": 1055.0765,
      "encoder_q-layer.0": 706.0272,
      "encoder_q-layer.1": 783.0759,
      "encoder_q-layer.10": 1225.1342,
      "encoder_q-layer.11": 2851.8794,
      "encoder_q-layer.2": 870.0491,
      "encoder_q-layer.3": 936.9124,
      "encoder_q-layer.4": 1006.8162,
      "encoder_q-layer.5": 1025.9822,
      "encoder_q-layer.6": 1182.9404,
      "encoder_q-layer.7": 1355.1207,
      "encoder_q-layer.8": 1568.2595,
      "encoder_q-layer.9": 1302.6631,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4724,
      "inbatch_pos_score": 1.1562,
      "learning_rate": 1.8777777777777777e-05,
      "loss": 3.2101,
      "norm_diff": 0.0489,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2023.875,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4729,
      "query_norm": 1.4673,
      "queue_k_norm": 1.5159,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0451,
      "sent_len_1": 66.7758,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7788,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 66200
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2215,
      "doc_norm": 1.5193,
      "encoder_q-embeddings": 1572.7737,
      "encoder_q-layer.0": 1167.8043,
      "encoder_q-layer.1": 1255.9601,
      "encoder_q-layer.10": 1249.7006,
      "encoder_q-layer.11": 2794.7954,
      "encoder_q-layer.2": 1478.1169,
      "encoder_q-layer.3": 1541.8165,
      "encoder_q-layer.4": 1729.3955,
      "encoder_q-layer.5": 1772.5271,
      "encoder_q-layer.6": 1576.5217,
      "encoder_q-layer.7": 1864.0337,
      "encoder_q-layer.8": 1755.8661,
      "encoder_q-layer.9": 1380.7043,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4801,
      "inbatch_pos_score": 1.1348,
      "learning_rate": 1.8722222222222223e-05,
      "loss": 3.2215,
      "norm_diff": 0.0507,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2532.2,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4797,
      "query_norm": 1.4686,
      "queue_k_norm": 1.516,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9878,
      "sent_len_1": 66.553,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2525,
      "stdk": 0.0491,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 66300
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1962,
      "doc_norm": 1.5175,
      "encoder_q-embeddings": 992.4595,
      "encoder_q-layer.0": 657.1902,
      "encoder_q-layer.1": 687.545,
      "encoder_q-layer.10": 1264.3606,
      "encoder_q-layer.11": 2854.7883,
      "encoder_q-layer.2": 760.746,
      "encoder_q-layer.3": 810.6346,
      "encoder_q-layer.4": 850.6559,
      "encoder_q-layer.5": 891.6509,
      "encoder_q-layer.6": 1021.1628,
      "encoder_q-layer.7": 1241.6879,
      "encoder_q-layer.8": 1394.3967,
      "encoder_q-layer.9": 1260.8405,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4816,
      "inbatch_pos_score": 1.1396,
      "learning_rate": 1.866666666666667e-05,
      "loss": 3.1962,
      "norm_diff": 0.0592,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1912.8885,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4819,
      "query_norm": 1.4583,
      "queue_k_norm": 1.5169,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9594,
      "sent_len_1": 66.8159,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.155,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 66400
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1999,
      "doc_norm": 1.5157,
      "encoder_q-embeddings": 1248.8297,
      "encoder_q-layer.0": 861.6307,
      "encoder_q-layer.1": 860.9256,
      "encoder_q-layer.10": 1275.7094,
      "encoder_q-layer.11": 2742.6914,
      "encoder_q-layer.2": 954.6053,
      "encoder_q-layer.3": 1040.0419,
      "encoder_q-layer.4": 1097.6724,
      "encoder_q-layer.5": 1113.3016,
      "encoder_q-layer.6": 1225.957,
      "encoder_q-layer.7": 1462.163,
      "encoder_q-layer.8": 1482.375,
      "encoder_q-layer.9": 1269.1078,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4823,
      "inbatch_pos_score": 1.1826,
      "learning_rate": 1.861111111111111e-05,
      "loss": 3.1999,
      "norm_diff": 0.0447,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2082.7485,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4832,
      "query_norm": 1.471,
      "queue_k_norm": 1.516,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9221,
      "sent_len_1": 66.5961,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.685,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 66500
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2256,
      "doc_norm": 1.5171,
      "encoder_q-embeddings": 1321.7635,
      "encoder_q-layer.0": 894.8356,
      "encoder_q-layer.1": 987.1892,
      "encoder_q-layer.10": 1138.4751,
      "encoder_q-layer.11": 2558.7668,
      "encoder_q-layer.2": 1117.0123,
      "encoder_q-layer.3": 1202.5038,
      "encoder_q-layer.4": 1250.2643,
      "encoder_q-layer.5": 1289.3866,
      "encoder_q-layer.6": 1425.1698,
      "encoder_q-layer.7": 1434.7356,
      "encoder_q-layer.8": 1434.5081,
      "encoder_q-layer.9": 1155.6309,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4813,
      "inbatch_pos_score": 1.1348,
      "learning_rate": 1.8555555555555557e-05,
      "loss": 3.2256,
      "norm_diff": 0.07,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2094.8336,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4822,
      "query_norm": 1.4471,
      "queue_k_norm": 1.5188,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7597,
      "sent_len_1": 66.8082,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.1,
      "stdk": 0.0489,
      "stdq": 0.045,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 66600
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2012,
      "doc_norm": 1.5177,
      "encoder_q-embeddings": 2316.5149,
      "encoder_q-layer.0": 1561.311,
      "encoder_q-layer.1": 1710.1243,
      "encoder_q-layer.10": 1170.449,
      "encoder_q-layer.11": 2718.3623,
      "encoder_q-layer.2": 1943.188,
      "encoder_q-layer.3": 2014.5714,
      "encoder_q-layer.4": 2157.4951,
      "encoder_q-layer.5": 2399.5161,
      "encoder_q-layer.6": 2314.009,
      "encoder_q-layer.7": 2493.8472,
      "encoder_q-layer.8": 2547.9565,
      "encoder_q-layer.9": 1490.8112,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4843,
      "inbatch_pos_score": 1.1621,
      "learning_rate": 1.85e-05,
      "loss": 3.2012,
      "norm_diff": 0.0665,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3246.517,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4844,
      "query_norm": 1.4512,
      "queue_k_norm": 1.5202,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8128,
      "sent_len_1": 66.7817,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4525,
      "stdk": 0.0489,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 66700
    },
    {
      "accuracy": 59.1797,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2211,
      "doc_norm": 1.5199,
      "encoder_q-embeddings": 1428.8146,
      "encoder_q-layer.0": 989.7008,
      "encoder_q-layer.1": 1081.5602,
      "encoder_q-layer.10": 1270.5989,
      "encoder_q-layer.11": 2749.4551,
      "encoder_q-layer.2": 1239.8781,
      "encoder_q-layer.3": 1356.0129,
      "encoder_q-layer.4": 1469.5216,
      "encoder_q-layer.5": 1496.98,
      "encoder_q-layer.6": 1573.2417,
      "encoder_q-layer.7": 1903.303,
      "encoder_q-layer.8": 1800.1536,
      "encoder_q-layer.9": 1278.26,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4863,
      "inbatch_pos_score": 1.1855,
      "learning_rate": 1.8444444444444445e-05,
      "loss": 3.2211,
      "norm_diff": 0.0563,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2386.505,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4861,
      "query_norm": 1.4637,
      "queue_k_norm": 1.5174,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7789,
      "sent_len_1": 66.6603,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4162,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 66800
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1956,
      "doc_norm": 1.5099,
      "encoder_q-embeddings": 3822.5632,
      "encoder_q-layer.0": 2637.7986,
      "encoder_q-layer.1": 2691.1316,
      "encoder_q-layer.10": 1196.2104,
      "encoder_q-layer.11": 2809.7473,
      "encoder_q-layer.2": 2682.9868,
      "encoder_q-layer.3": 2521.876,
      "encoder_q-layer.4": 2379.7815,
      "encoder_q-layer.5": 2331.3298,
      "encoder_q-layer.6": 2068.3506,
      "encoder_q-layer.7": 1741.7122,
      "encoder_q-layer.8": 1489.7711,
      "encoder_q-layer.9": 1209.8713,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4884,
      "inbatch_pos_score": 1.1641,
      "learning_rate": 1.838888888888889e-05,
      "loss": 3.1956,
      "norm_diff": 0.0374,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3763.0106,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4883,
      "query_norm": 1.4725,
      "queue_k_norm": 1.5208,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9903,
      "sent_len_1": 66.7764,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2812,
      "stdk": 0.0485,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 66900
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1778,
      "doc_norm": 1.5203,
      "encoder_q-embeddings": 1338.9401,
      "encoder_q-layer.0": 841.085,
      "encoder_q-layer.1": 959.1176,
      "encoder_q-layer.10": 1502.5173,
      "encoder_q-layer.11": 3035.0972,
      "encoder_q-layer.2": 1057.7815,
      "encoder_q-layer.3": 1145.7677,
      "encoder_q-layer.4": 1134.692,
      "encoder_q-layer.5": 1160.4739,
      "encoder_q-layer.6": 1236.6545,
      "encoder_q-layer.7": 1422.3202,
      "encoder_q-layer.8": 1512.2552,
      "encoder_q-layer.9": 1347.3927,
      "epoch": 0.65,
      "inbatch_neg_score": 0.4892,
      "inbatch_pos_score": 1.167,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 3.1778,
      "norm_diff": 0.0462,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2181.5251,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4883,
      "query_norm": 1.4741,
      "queue_k_norm": 1.5215,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1984,
      "sent_len_1": 66.701,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6387,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 67000
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2197,
      "doc_norm": 1.5223,
      "encoder_q-embeddings": 1064.3405,
      "encoder_q-layer.0": 718.1384,
      "encoder_q-layer.1": 755.067,
      "encoder_q-layer.10": 1207.463,
      "encoder_q-layer.11": 2568.2585,
      "encoder_q-layer.2": 862.363,
      "encoder_q-layer.3": 949.7911,
      "encoder_q-layer.4": 1019.1638,
      "encoder_q-layer.5": 1047.7592,
      "encoder_q-layer.6": 1205.2594,
      "encoder_q-layer.7": 1403.2936,
      "encoder_q-layer.8": 1417.9128,
      "encoder_q-layer.9": 1201.0784,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4888,
      "inbatch_pos_score": 1.1318,
      "learning_rate": 1.827777777777778e-05,
      "loss": 3.2197,
      "norm_diff": 0.0745,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1902.9124,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4878,
      "query_norm": 1.4478,
      "queue_k_norm": 1.5217,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9276,
      "sent_len_1": 66.89,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1637,
      "stdk": 0.049,
      "stdq": 0.0451,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 67100
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1745,
      "doc_norm": 1.525,
      "encoder_q-embeddings": 1977.5953,
      "encoder_q-layer.0": 1313.5333,
      "encoder_q-layer.1": 1446.9448,
      "encoder_q-layer.10": 2220.4412,
      "encoder_q-layer.11": 5102.1475,
      "encoder_q-layer.2": 1612.3713,
      "encoder_q-layer.3": 1737.1284,
      "encoder_q-layer.4": 1805.67,
      "encoder_q-layer.5": 1949.3031,
      "encoder_q-layer.6": 2192.5417,
      "encoder_q-layer.7": 2550.4897,
      "encoder_q-layer.8": 2996.1558,
      "encoder_q-layer.9": 2455.6638,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4872,
      "inbatch_pos_score": 1.168,
      "learning_rate": 1.8222222222222224e-05,
      "loss": 3.1745,
      "norm_diff": 0.0583,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3663.749,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4878,
      "query_norm": 1.4667,
      "queue_k_norm": 1.5211,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0573,
      "sent_len_1": 67.0753,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7225,
      "stdk": 0.0491,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 67200
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.208,
      "doc_norm": 1.5211,
      "encoder_q-embeddings": 2695.3662,
      "encoder_q-layer.0": 1828.6154,
      "encoder_q-layer.1": 2081.3638,
      "encoder_q-layer.10": 2488.3694,
      "encoder_q-layer.11": 5690.0156,
      "encoder_q-layer.2": 2353.2563,
      "encoder_q-layer.3": 2545.2271,
      "encoder_q-layer.4": 2634.636,
      "encoder_q-layer.5": 2602.564,
      "encoder_q-layer.6": 2985.6328,
      "encoder_q-layer.7": 3173.1541,
      "encoder_q-layer.8": 3316.5251,
      "encoder_q-layer.9": 2723.6062,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4867,
      "inbatch_pos_score": 1.1689,
      "learning_rate": 1.8166666666666667e-05,
      "loss": 3.208,
      "norm_diff": 0.0505,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4472.3721,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4866,
      "query_norm": 1.4707,
      "queue_k_norm": 1.5209,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.922,
      "sent_len_1": 66.8213,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8413,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 67300
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2181,
      "doc_norm": 1.5184,
      "encoder_q-embeddings": 7531.5083,
      "encoder_q-layer.0": 5070.645,
      "encoder_q-layer.1": 5679.938,
      "encoder_q-layer.10": 2510.8306,
      "encoder_q-layer.11": 5366.2495,
      "encoder_q-layer.2": 5955.084,
      "encoder_q-layer.3": 6232.4814,
      "encoder_q-layer.4": 6146.2441,
      "encoder_q-layer.5": 4783.5112,
      "encoder_q-layer.6": 4827.7769,
      "encoder_q-layer.7": 5608.2124,
      "encoder_q-layer.8": 4741.1802,
      "encoder_q-layer.9": 3037.9404,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4864,
      "inbatch_pos_score": 1.165,
      "learning_rate": 1.8111111111111112e-05,
      "loss": 3.2181,
      "norm_diff": 0.0624,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 8250.8761,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4868,
      "query_norm": 1.456,
      "queue_k_norm": 1.52,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0575,
      "sent_len_1": 66.8381,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2175,
      "stdk": 0.0488,
      "stdq": 0.0454,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 67400
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1868,
      "doc_norm": 1.5146,
      "encoder_q-embeddings": 1043.674,
      "encoder_q-layer.0": 702.5586,
      "encoder_q-layer.1": 773.145,
      "encoder_q-layer.10": 1178.4487,
      "encoder_q-layer.11": 2655.771,
      "encoder_q-layer.2": 859.5906,
      "encoder_q-layer.3": 923.7656,
      "encoder_q-layer.4": 1046.6565,
      "encoder_q-layer.5": 1088.6423,
      "encoder_q-layer.6": 1197.8267,
      "encoder_q-layer.7": 1334.7672,
      "encoder_q-layer.8": 1451.6708,
      "encoder_q-layer.9": 1208.9156,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4875,
      "inbatch_pos_score": 1.1797,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 3.1868,
      "norm_diff": 0.0503,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1922.3962,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4883,
      "query_norm": 1.4644,
      "queue_k_norm": 1.5214,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8113,
      "sent_len_1": 66.9035,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8388,
      "stdk": 0.0487,
      "stdq": 0.0457,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 67500
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1805,
      "doc_norm": 1.5245,
      "encoder_q-embeddings": 1310.7788,
      "encoder_q-layer.0": 858.6702,
      "encoder_q-layer.1": 973.6287,
      "encoder_q-layer.10": 1175.6765,
      "encoder_q-layer.11": 2671.4602,
      "encoder_q-layer.2": 1150.9014,
      "encoder_q-layer.3": 1127.5564,
      "encoder_q-layer.4": 1189.6876,
      "encoder_q-layer.5": 1213.7334,
      "encoder_q-layer.6": 1187.0231,
      "encoder_q-layer.7": 1288.7046,
      "encoder_q-layer.8": 1435.9858,
      "encoder_q-layer.9": 1262.1703,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4892,
      "inbatch_pos_score": 1.1416,
      "learning_rate": 1.8e-05,
      "loss": 3.1805,
      "norm_diff": 0.0707,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2080.1081,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4897,
      "query_norm": 1.4538,
      "queue_k_norm": 1.5208,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7952,
      "sent_len_1": 66.5417,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0825,
      "stdk": 0.049,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 67600
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1626,
      "doc_norm": 1.5193,
      "encoder_q-embeddings": 1048.6641,
      "encoder_q-layer.0": 681.0511,
      "encoder_q-layer.1": 735.0033,
      "encoder_q-layer.10": 1235.0002,
      "encoder_q-layer.11": 2729.7825,
      "encoder_q-layer.2": 842.5024,
      "encoder_q-layer.3": 862.2263,
      "encoder_q-layer.4": 919.9046,
      "encoder_q-layer.5": 949.9571,
      "encoder_q-layer.6": 1079.3254,
      "encoder_q-layer.7": 1316.5399,
      "encoder_q-layer.8": 1430.3328,
      "encoder_q-layer.9": 1253.7642,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4894,
      "inbatch_pos_score": 1.1504,
      "learning_rate": 1.7944444444444443e-05,
      "loss": 3.1626,
      "norm_diff": 0.06,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1920.8833,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.49,
      "query_norm": 1.4593,
      "queue_k_norm": 1.5226,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8669,
      "sent_len_1": 66.828,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6975,
      "stdk": 0.0488,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 67700
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2043,
      "doc_norm": 1.5283,
      "encoder_q-embeddings": 7820.229,
      "encoder_q-layer.0": 5392.0093,
      "encoder_q-layer.1": 6152.2246,
      "encoder_q-layer.10": 1325.6246,
      "encoder_q-layer.11": 2770.4961,
      "encoder_q-layer.2": 6872.645,
      "encoder_q-layer.3": 8224.7021,
      "encoder_q-layer.4": 9767.6113,
      "encoder_q-layer.5": 10353.9561,
      "encoder_q-layer.6": 10802.168,
      "encoder_q-layer.7": 11023.6807,
      "encoder_q-layer.8": 7869.9146,
      "encoder_q-layer.9": 1911.5994,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4904,
      "inbatch_pos_score": 1.1562,
      "learning_rate": 1.788888888888889e-05,
      "loss": 3.2043,
      "norm_diff": 0.0569,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 11425.395,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4907,
      "query_norm": 1.4715,
      "queue_k_norm": 1.5238,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9145,
      "sent_len_1": 66.7561,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6275,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 67800
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2292,
      "doc_norm": 1.5269,
      "encoder_q-embeddings": 1073.1238,
      "encoder_q-layer.0": 681.0153,
      "encoder_q-layer.1": 721.3771,
      "encoder_q-layer.10": 1199.608,
      "encoder_q-layer.11": 2705.636,
      "encoder_q-layer.2": 813.5032,
      "encoder_q-layer.3": 843.981,
      "encoder_q-layer.4": 910.0774,
      "encoder_q-layer.5": 973.1188,
      "encoder_q-layer.6": 1146.8329,
      "encoder_q-layer.7": 1319.5917,
      "encoder_q-layer.8": 1510.2059,
      "encoder_q-layer.9": 1310.8134,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4882,
      "inbatch_pos_score": 1.1504,
      "learning_rate": 1.7833333333333334e-05,
      "loss": 3.2292,
      "norm_diff": 0.0674,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1940.6058,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4893,
      "query_norm": 1.4594,
      "queue_k_norm": 1.5236,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0403,
      "sent_len_1": 66.8463,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5213,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 67900
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2251,
      "doc_norm": 1.5244,
      "encoder_q-embeddings": 11545.7627,
      "encoder_q-layer.0": 8340.6953,
      "encoder_q-layer.1": 8057.3979,
      "encoder_q-layer.10": 1166.6963,
      "encoder_q-layer.11": 2525.6101,
      "encoder_q-layer.2": 9826.0244,
      "encoder_q-layer.3": 10166.0996,
      "encoder_q-layer.4": 10042.833,
      "encoder_q-layer.5": 11535.5117,
      "encoder_q-layer.6": 9478.0732,
      "encoder_q-layer.7": 6499.2412,
      "encoder_q-layer.8": 4688.373,
      "encoder_q-layer.9": 1761.1381,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4904,
      "inbatch_pos_score": 1.1689,
      "learning_rate": 1.777777777777778e-05,
      "loss": 3.2251,
      "norm_diff": 0.0603,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12367.9378,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.4897,
      "query_norm": 1.4641,
      "queue_k_norm": 1.5211,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8749,
      "sent_len_1": 66.7007,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.345,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 68000
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2085,
      "doc_norm": 1.5208,
      "encoder_q-embeddings": 1274.7114,
      "encoder_q-layer.0": 849.0226,
      "encoder_q-layer.1": 952.6102,
      "encoder_q-layer.10": 1261.4025,
      "encoder_q-layer.11": 2856.0095,
      "encoder_q-layer.2": 1091.8851,
      "encoder_q-layer.3": 1175.0961,
      "encoder_q-layer.4": 1335.9082,
      "encoder_q-layer.5": 1324.2708,
      "encoder_q-layer.6": 1464.1993,
      "encoder_q-layer.7": 1576.0826,
      "encoder_q-layer.8": 1644.9454,
      "encoder_q-layer.9": 1281.9733,
      "epoch": 0.66,
      "inbatch_neg_score": 0.4904,
      "inbatch_pos_score": 1.125,
      "learning_rate": 1.7722222222222222e-05,
      "loss": 3.2085,
      "norm_diff": 0.0722,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2218.3406,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4902,
      "query_norm": 1.4486,
      "queue_k_norm": 1.5246,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0656,
      "sent_len_1": 66.8031,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7738,
      "stdk": 0.0488,
      "stdq": 0.045,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 68100
    },
    {
      "accuracy": 58.3008,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1899,
      "doc_norm": 1.5326,
      "encoder_q-embeddings": 1371.7784,
      "encoder_q-layer.0": 880.1191,
      "encoder_q-layer.1": 1040.6588,
      "encoder_q-layer.10": 1248.3231,
      "encoder_q-layer.11": 2727.3687,
      "encoder_q-layer.2": 1173.6083,
      "encoder_q-layer.3": 1418.5712,
      "encoder_q-layer.4": 1424.6377,
      "encoder_q-layer.5": 1465.9413,
      "encoder_q-layer.6": 1540.0216,
      "encoder_q-layer.7": 1534.8785,
      "encoder_q-layer.8": 1493.0798,
      "encoder_q-layer.9": 1231.464,
      "epoch": 0.67,
      "inbatch_neg_score": 0.4953,
      "inbatch_pos_score": 1.1992,
      "learning_rate": 1.7666666666666668e-05,
      "loss": 3.1899,
      "norm_diff": 0.0671,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2224.4909,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4963,
      "query_norm": 1.4655,
      "queue_k_norm": 1.5244,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1192,
      "sent_len_1": 66.8282,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6538,
      "stdk": 0.0493,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 68200
    },
    {
      "accuracy": 58.1055,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1973,
      "doc_norm": 1.5263,
      "encoder_q-embeddings": 1249.9409,
      "encoder_q-layer.0": 842.6589,
      "encoder_q-layer.1": 1006.8672,
      "encoder_q-layer.10": 1152.9519,
      "encoder_q-layer.11": 2538.8555,
      "encoder_q-layer.2": 1305.3854,
      "encoder_q-layer.3": 1383.4764,
      "encoder_q-layer.4": 1330.645,
      "encoder_q-layer.5": 1186.66,
      "encoder_q-layer.6": 1365.9683,
      "encoder_q-layer.7": 1453.8704,
      "encoder_q-layer.8": 1414.6389,
      "encoder_q-layer.9": 1214.7428,
      "epoch": 0.67,
      "inbatch_neg_score": 0.499,
      "inbatch_pos_score": 1.1973,
      "learning_rate": 1.761111111111111e-05,
      "loss": 3.1973,
      "norm_diff": 0.0434,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2111.081,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.4993,
      "query_norm": 1.4829,
      "queue_k_norm": 1.5238,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0545,
      "sent_len_1": 66.7266,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3212,
      "stdk": 0.0491,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 68300
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2053,
      "doc_norm": 1.5199,
      "encoder_q-embeddings": 943.1119,
      "encoder_q-layer.0": 608.234,
      "encoder_q-layer.1": 643.598,
      "encoder_q-layer.10": 1196.2977,
      "encoder_q-layer.11": 2632.3943,
      "encoder_q-layer.2": 710.9733,
      "encoder_q-layer.3": 766.9286,
      "encoder_q-layer.4": 845.1102,
      "encoder_q-layer.5": 899.4209,
      "encoder_q-layer.6": 1054.3529,
      "encoder_q-layer.7": 1171.8704,
      "encoder_q-layer.8": 1382.4926,
      "encoder_q-layer.9": 1216.3771,
      "epoch": 0.67,
      "inbatch_neg_score": 0.5007,
      "inbatch_pos_score": 1.1748,
      "learning_rate": 1.7555555555555556e-05,
      "loss": 3.2053,
      "norm_diff": 0.051,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1831.6731,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5,
      "query_norm": 1.4688,
      "queue_k_norm": 1.525,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9818,
      "sent_len_1": 66.6791,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6175,
      "stdk": 0.0487,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 68400
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1983,
      "doc_norm": 1.5269,
      "encoder_q-embeddings": 8864.1426,
      "encoder_q-layer.0": 5931.0259,
      "encoder_q-layer.1": 6544.5781,
      "encoder_q-layer.10": 1318.2808,
      "encoder_q-layer.11": 2950.8213,
      "encoder_q-layer.2": 7724.0801,
      "encoder_q-layer.3": 8310.6914,
      "encoder_q-layer.4": 9652.8682,
      "encoder_q-layer.5": 10946.9561,
      "encoder_q-layer.6": 11348.334,
      "encoder_q-layer.7": 10398.1504,
      "encoder_q-layer.8": 6834.3618,
      "encoder_q-layer.9": 2022.1581,
      "epoch": 0.67,
      "inbatch_neg_score": 0.505,
      "inbatch_pos_score": 1.1914,
      "learning_rate": 1.75e-05,
      "loss": 3.1983,
      "norm_diff": 0.0471,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 11600.2654,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5039,
      "query_norm": 1.4798,
      "queue_k_norm": 1.5252,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9837,
      "sent_len_1": 66.7062,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 185.6087,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 68500
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2171,
      "doc_norm": 1.5337,
      "encoder_q-embeddings": 1002.3615,
      "encoder_q-layer.0": 659.6469,
      "encoder_q-layer.1": 698.1865,
      "encoder_q-layer.10": 1297.0663,
      "encoder_q-layer.11": 2786.0684,
      "encoder_q-layer.2": 794.8015,
      "encoder_q-layer.3": 860.4907,
      "encoder_q-layer.4": 928.982,
      "encoder_q-layer.5": 965.394,
      "encoder_q-layer.6": 1052.8906,
      "encoder_q-layer.7": 1222.5747,
      "encoder_q-layer.8": 1436.2412,
      "encoder_q-layer.9": 1314.4457,
      "epoch": 0.67,
      "inbatch_neg_score": 0.5033,
      "inbatch_pos_score": 1.1885,
      "learning_rate": 1.7444444444444448e-05,
      "loss": 3.2171,
      "norm_diff": 0.0629,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1903.6103,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5039,
      "query_norm": 1.4708,
      "queue_k_norm": 1.5251,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8307,
      "sent_len_1": 66.6683,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.545,
      "stdk": 0.0493,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 68600
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1719,
      "doc_norm": 1.5251,
      "encoder_q-embeddings": 1027.3361,
      "encoder_q-layer.0": 662.5479,
      "encoder_q-layer.1": 698.8363,
      "encoder_q-layer.10": 1215.6469,
      "encoder_q-layer.11": 2827.7136,
      "encoder_q-layer.2": 813.6992,
      "encoder_q-layer.3": 841.1902,
      "encoder_q-layer.4": 925.2224,
      "encoder_q-layer.5": 950.0133,
      "encoder_q-layer.6": 1121.2323,
      "encoder_q-layer.7": 1293.2194,
      "encoder_q-layer.8": 1452.264,
      "encoder_q-layer.9": 1271.0665,
      "epoch": 0.67,
      "inbatch_neg_score": 0.506,
      "inbatch_pos_score": 1.1504,
      "learning_rate": 1.738888888888889e-05,
      "loss": 3.1719,
      "norm_diff": 0.0525,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1964.817,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5063,
      "query_norm": 1.4726,
      "queue_k_norm": 1.5256,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8819,
      "sent_len_1": 66.7434,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0863,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 68700
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1721,
      "doc_norm": 1.5273,
      "encoder_q-embeddings": 1192.0872,
      "encoder_q-layer.0": 789.7538,
      "encoder_q-layer.1": 884.5311,
      "encoder_q-layer.10": 1150.0106,
      "encoder_q-layer.11": 2653.0049,
      "encoder_q-layer.2": 997.0336,
      "encoder_q-layer.3": 1078.4352,
      "encoder_q-layer.4": 1147.7887,
      "encoder_q-layer.5": 1237.3641,
      "encoder_q-layer.6": 1395.551,
      "encoder_q-layer.7": 1593.4426,
      "encoder_q-layer.8": 1762.6409,
      "encoder_q-layer.9": 1180.917,
      "epoch": 0.67,
      "inbatch_neg_score": 0.5113,
      "inbatch_pos_score": 1.1816,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 3.1721,
      "norm_diff": 0.0477,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2122.4622,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5112,
      "query_norm": 1.4796,
      "queue_k_norm": 1.5266,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8802,
      "sent_len_1": 66.7103,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6138,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 68800
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1526,
      "doc_norm": 1.5216,
      "encoder_q-embeddings": 1042.5646,
      "encoder_q-layer.0": 674.2281,
      "encoder_q-layer.1": 714.0071,
      "encoder_q-layer.10": 1234.6494,
      "encoder_q-layer.11": 2611.9165,
      "encoder_q-layer.2": 818.027,
      "encoder_q-layer.3": 871.0714,
      "encoder_q-layer.4": 972.5583,
      "encoder_q-layer.5": 999.8356,
      "encoder_q-layer.6": 1152.4741,
      "encoder_q-layer.7": 1268.1172,
      "encoder_q-layer.8": 1442.5642,
      "encoder_q-layer.9": 1296.2848,
      "epoch": 0.67,
      "inbatch_neg_score": 0.5135,
      "inbatch_pos_score": 1.1758,
      "learning_rate": 1.7277777777777778e-05,
      "loss": 3.1526,
      "norm_diff": 0.0419,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1891.9781,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5132,
      "query_norm": 1.4797,
      "queue_k_norm": 1.5276,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9197,
      "sent_len_1": 66.8487,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2713,
      "stdk": 0.0487,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 68900
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1886,
      "doc_norm": 1.5246,
      "encoder_q-embeddings": 910.4836,
      "encoder_q-layer.0": 593.6825,
      "encoder_q-layer.1": 631.8937,
      "encoder_q-layer.10": 1175.0714,
      "encoder_q-layer.11": 2531.6089,
      "encoder_q-layer.2": 724.4766,
      "encoder_q-layer.3": 732.6523,
      "encoder_q-layer.4": 770.4797,
      "encoder_q-layer.5": 815.086,
      "encoder_q-layer.6": 993.97,
      "encoder_q-layer.7": 1197.702,
      "encoder_q-layer.8": 1356.6261,
      "encoder_q-layer.9": 1209.7686,
      "epoch": 0.67,
      "inbatch_neg_score": 0.5147,
      "inbatch_pos_score": 1.2197,
      "learning_rate": 1.7222222222222224e-05,
      "loss": 3.1886,
      "norm_diff": 0.0254,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1758.5526,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5137,
      "query_norm": 1.4991,
      "queue_k_norm": 1.5266,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1288,
      "sent_len_1": 66.6093,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.865,
      "stdk": 0.0488,
      "stdq": 0.0464,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69000
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.204,
      "doc_norm": 1.5249,
      "encoder_q-embeddings": 2508.3132,
      "encoder_q-layer.0": 1695.6292,
      "encoder_q-layer.1": 1855.4183,
      "encoder_q-layer.10": 1257.9952,
      "encoder_q-layer.11": 2678.6606,
      "encoder_q-layer.2": 2158.4463,
      "encoder_q-layer.3": 2355.9932,
      "encoder_q-layer.4": 2813.8875,
      "encoder_q-layer.5": 3183.4236,
      "encoder_q-layer.6": 3355.6299,
      "encoder_q-layer.7": 3653.6331,
      "encoder_q-layer.8": 2581.7874,
      "encoder_q-layer.9": 1430.7206,
      "epoch": 0.67,
      "inbatch_neg_score": 0.5165,
      "inbatch_pos_score": 1.165,
      "learning_rate": 1.7166666666666666e-05,
      "loss": 3.204,
      "norm_diff": 0.0447,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3738.4383,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5161,
      "query_norm": 1.4802,
      "queue_k_norm": 1.5266,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9629,
      "sent_len_1": 66.8261,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1475,
      "stdk": 0.0488,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69100
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1934,
      "doc_norm": 1.5278,
      "encoder_q-embeddings": 899.2848,
      "encoder_q-layer.0": 605.0774,
      "encoder_q-layer.1": 627.5343,
      "encoder_q-layer.10": 1210.9093,
      "encoder_q-layer.11": 2694.4802,
      "encoder_q-layer.2": 735.9399,
      "encoder_q-layer.3": 787.6987,
      "encoder_q-layer.4": 814.3285,
      "encoder_q-layer.5": 874.9877,
      "encoder_q-layer.6": 990.7902,
      "encoder_q-layer.7": 1214.6355,
      "encoder_q-layer.8": 1395.687,
      "encoder_q-layer.9": 1194.7007,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5166,
      "inbatch_pos_score": 1.2148,
      "learning_rate": 1.7111111111111112e-05,
      "loss": 3.1934,
      "norm_diff": 0.0332,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1835.34,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5161,
      "query_norm": 1.4946,
      "queue_k_norm": 1.5271,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1926,
      "sent_len_1": 66.9804,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4437,
      "stdk": 0.0489,
      "stdq": 0.0463,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69200
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.214,
      "doc_norm": 1.5235,
      "encoder_q-embeddings": 503.0397,
      "encoder_q-layer.0": 327.3791,
      "encoder_q-layer.1": 356.7862,
      "encoder_q-layer.10": 694.6742,
      "encoder_q-layer.11": 1456.4427,
      "encoder_q-layer.2": 401.3344,
      "encoder_q-layer.3": 425.9633,
      "encoder_q-layer.4": 444.0031,
      "encoder_q-layer.5": 452.0669,
      "encoder_q-layer.6": 541.3746,
      "encoder_q-layer.7": 617.7626,
      "encoder_q-layer.8": 687.8088,
      "encoder_q-layer.9": 623.4868,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5223,
      "inbatch_pos_score": 1.1611,
      "learning_rate": 1.7055555555555554e-05,
      "loss": 3.214,
      "norm_diff": 0.0581,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 981.0727,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5229,
      "query_norm": 1.4653,
      "queue_k_norm": 1.5296,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.894,
      "sent_len_1": 66.5787,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8562,
      "stdk": 0.0487,
      "stdq": 0.045,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 69300
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1995,
      "doc_norm": 1.5346,
      "encoder_q-embeddings": 474.9575,
      "encoder_q-layer.0": 319.4057,
      "encoder_q-layer.1": 333.4863,
      "encoder_q-layer.10": 588.4682,
      "encoder_q-layer.11": 1374.3879,
      "encoder_q-layer.2": 375.0822,
      "encoder_q-layer.3": 396.8225,
      "encoder_q-layer.4": 431.2452,
      "encoder_q-layer.5": 448.2468,
      "encoder_q-layer.6": 506.0506,
      "encoder_q-layer.7": 573.7466,
      "encoder_q-layer.8": 664.8459,
      "encoder_q-layer.9": 593.0918,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5234,
      "inbatch_pos_score": 1.2217,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 3.1995,
      "norm_diff": 0.0465,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 928.6183,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5234,
      "query_norm": 1.4881,
      "queue_k_norm": 1.5294,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7643,
      "sent_len_1": 66.4279,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4875,
      "stdk": 0.0491,
      "stdq": 0.046,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69400
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2027,
      "doc_norm": 1.5206,
      "encoder_q-embeddings": 500.6015,
      "encoder_q-layer.0": 331.5562,
      "encoder_q-layer.1": 346.3316,
      "encoder_q-layer.10": 645.4955,
      "encoder_q-layer.11": 1459.1542,
      "encoder_q-layer.2": 388.9737,
      "encoder_q-layer.3": 409.1449,
      "encoder_q-layer.4": 456.3963,
      "encoder_q-layer.5": 478.7166,
      "encoder_q-layer.6": 555.383,
      "encoder_q-layer.7": 668.4579,
      "encoder_q-layer.8": 745.601,
      "encoder_q-layer.9": 630.4598,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5239,
      "inbatch_pos_score": 1.166,
      "learning_rate": 1.6944444444444446e-05,
      "loss": 3.2027,
      "norm_diff": 0.0634,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 978.7598,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5234,
      "query_norm": 1.4572,
      "queue_k_norm": 1.5292,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.947,
      "sent_len_1": 66.674,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4525,
      "stdk": 0.0485,
      "stdq": 0.0449,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69500
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.187,
      "doc_norm": 1.534,
      "encoder_q-embeddings": 957.8161,
      "encoder_q-layer.0": 635.778,
      "encoder_q-layer.1": 737.5353,
      "encoder_q-layer.10": 669.6371,
      "encoder_q-layer.11": 1444.5139,
      "encoder_q-layer.2": 851.9166,
      "encoder_q-layer.3": 954.955,
      "encoder_q-layer.4": 1048.824,
      "encoder_q-layer.5": 1265.3629,
      "encoder_q-layer.6": 1153.6866,
      "encoder_q-layer.7": 1164.2694,
      "encoder_q-layer.8": 1065.024,
      "encoder_q-layer.9": 655.3311,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5248,
      "inbatch_pos_score": 1.2227,
      "learning_rate": 1.688888888888889e-05,
      "loss": 3.187,
      "norm_diff": 0.0475,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1507.6127,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5239,
      "query_norm": 1.4865,
      "queue_k_norm": 1.5313,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9703,
      "sent_len_1": 66.5988,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.885,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 69600
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2032,
      "doc_norm": 1.5275,
      "encoder_q-embeddings": 736.4041,
      "encoder_q-layer.0": 485.0319,
      "encoder_q-layer.1": 562.1609,
      "encoder_q-layer.10": 625.3538,
      "encoder_q-layer.11": 1394.5526,
      "encoder_q-layer.2": 649.6409,
      "encoder_q-layer.3": 690.0334,
      "encoder_q-layer.4": 740.6035,
      "encoder_q-layer.5": 797.5995,
      "encoder_q-layer.6": 807.9946,
      "encoder_q-layer.7": 811.5798,
      "encoder_q-layer.8": 821.8193,
      "encoder_q-layer.9": 641.6797,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5257,
      "inbatch_pos_score": 1.1729,
      "learning_rate": 1.6833333333333334e-05,
      "loss": 3.2032,
      "norm_diff": 0.0652,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1190.7113,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5259,
      "query_norm": 1.4623,
      "queue_k_norm": 1.5317,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9209,
      "sent_len_1": 66.925,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4462,
      "stdk": 0.0488,
      "stdq": 0.0451,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69700
    },
    {
      "accuracy": 55.1758,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2133,
      "doc_norm": 1.5329,
      "encoder_q-embeddings": 911.0137,
      "encoder_q-layer.0": 607.2457,
      "encoder_q-layer.1": 725.8464,
      "encoder_q-layer.10": 645.4068,
      "encoder_q-layer.11": 1389.1228,
      "encoder_q-layer.2": 832.4066,
      "encoder_q-layer.3": 885.9224,
      "encoder_q-layer.4": 935.8718,
      "encoder_q-layer.5": 968.6967,
      "encoder_q-layer.6": 963.5327,
      "encoder_q-layer.7": 844.6526,
      "encoder_q-layer.8": 801.3793,
      "encoder_q-layer.9": 642.5611,
      "epoch": 0.68,
      "inbatch_neg_score": 0.527,
      "inbatch_pos_score": 1.1855,
      "learning_rate": 1.677777777777778e-05,
      "loss": 3.2133,
      "norm_diff": 0.0583,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1319.0473,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5269,
      "query_norm": 1.4746,
      "queue_k_norm": 1.5325,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0478,
      "sent_len_1": 66.9045,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9412,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69800
    },
    {
      "accuracy": 50.8789,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2094,
      "doc_norm": 1.527,
      "encoder_q-embeddings": 645.037,
      "encoder_q-layer.0": 450.5551,
      "encoder_q-layer.1": 525.4329,
      "encoder_q-layer.10": 634.7334,
      "encoder_q-layer.11": 1472.4832,
      "encoder_q-layer.2": 629.398,
      "encoder_q-layer.3": 661.5403,
      "encoder_q-layer.4": 712.6177,
      "encoder_q-layer.5": 731.8926,
      "encoder_q-layer.6": 699.7377,
      "encoder_q-layer.7": 782.0722,
      "encoder_q-layer.8": 781.5135,
      "encoder_q-layer.9": 620.0786,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5311,
      "inbatch_pos_score": 1.1543,
      "learning_rate": 1.6722222222222222e-05,
      "loss": 3.2094,
      "norm_diff": 0.0682,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1145.0715,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5298,
      "query_norm": 1.4588,
      "queue_k_norm": 1.5306,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9129,
      "sent_len_1": 66.537,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5938,
      "stdk": 0.0487,
      "stdq": 0.0449,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 69900
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1962,
      "doc_norm": 1.5293,
      "encoder_q-embeddings": 926.1125,
      "encoder_q-layer.0": 668.567,
      "encoder_q-layer.1": 737.9895,
      "encoder_q-layer.10": 619.5527,
      "encoder_q-layer.11": 1431.3833,
      "encoder_q-layer.2": 911.4559,
      "encoder_q-layer.3": 1058.8562,
      "encoder_q-layer.4": 1242.4004,
      "encoder_q-layer.5": 1168.8914,
      "encoder_q-layer.6": 1363.1268,
      "encoder_q-layer.7": 1389.6298,
      "encoder_q-layer.8": 921.5632,
      "encoder_q-layer.9": 686.6312,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5307,
      "inbatch_pos_score": 1.2324,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 3.1962,
      "norm_diff": 0.0277,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1596.3432,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5293,
      "query_norm": 1.5017,
      "queue_k_norm": 1.5329,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8683,
      "sent_len_1": 66.8016,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8862,
      "stdk": 0.0488,
      "stdq": 0.0468,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 70000
    },
    {
      "dev_runtime": 38.5045,
      "dev_samples_per_second": 1.662,
      "dev_steps_per_second": 0.026,
      "epoch": 0.68,
      "step": 70000,
      "test_accuracy": 94.59228515625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.31344741582870483,
      "test_doc_norm": 1.5325968265533447,
      "test_inbatch_neg_score": 0.9010056853294373,
      "test_inbatch_pos_score": 1.8864071369171143,
      "test_loss": 0.31344741582870483,
      "test_loss_align": 1.0511610507965088,
      "test_loss_unif": 3.477783679962158,
      "test_loss_unif_q@queue": 3.477783679962158,
      "test_norm_diff": 0.03206353634595871,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.5119207501411438,
      "test_query_norm": 1.5646604299545288,
      "test_queue_k_norm": 1.5326721668243408,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04340749233961105,
      "test_stdq": 0.043892648071050644,
      "test_stdqueue_k": 0.04903004691004753,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 38.5045,
      "dev_samples_per_second": 1.662,
      "dev_steps_per_second": 0.026,
      "epoch": 0.68,
      "eval_beir-arguana_ndcg@10": 0.35943,
      "eval_beir-arguana_recall@10": 0.6138,
      "eval_beir-arguana_recall@100": 0.90967,
      "eval_beir-arguana_recall@20": 0.75036,
      "eval_beir-avg_ndcg@10": 0.36724750000000006,
      "eval_beir-avg_recall@10": 0.43349650000000006,
      "eval_beir-avg_recall@100": 0.6165515833333333,
      "eval_beir-avg_recall@20": 0.49311358333333344,
      "eval_beir-cqadupstack_ndcg@10": 0.26323500000000005,
      "eval_beir-cqadupstack_recall@10": 0.353265,
      "eval_beir-cqadupstack_recall@100": 0.5810358333333333,
      "eval_beir-cqadupstack_recall@20": 0.41737583333333333,
      "eval_beir-fiqa_ndcg@10": 0.20825,
      "eval_beir-fiqa_recall@10": 0.26453,
      "eval_beir-fiqa_recall@100": 0.52102,
      "eval_beir-fiqa_recall@20": 0.32776,
      "eval_beir-nfcorpus_ndcg@10": 0.29474,
      "eval_beir-nfcorpus_recall@10": 0.14505,
      "eval_beir-nfcorpus_recall@100": 0.27328,
      "eval_beir-nfcorpus_recall@20": 0.17854,
      "eval_beir-nq_ndcg@10": 0.26901,
      "eval_beir-nq_recall@10": 0.44457,
      "eval_beir-nq_recall@100": 0.78498,
      "eval_beir-nq_recall@20": 0.5624,
      "eval_beir-quora_ndcg@10": 0.82261,
      "eval_beir-quora_recall@10": 0.9152,
      "eval_beir-quora_recall@100": 0.98481,
      "eval_beir-quora_recall@20": 0.9495,
      "eval_beir-scidocs_ndcg@10": 0.14481,
      "eval_beir-scidocs_recall@10": 0.15133,
      "eval_beir-scidocs_recall@100": 0.35128,
      "eval_beir-scidocs_recall@20": 0.20582,
      "eval_beir-scifact_ndcg@10": 0.61668,
      "eval_beir-scifact_recall@10": 0.75483,
      "eval_beir-scifact_recall@100": 0.92422,
      "eval_beir-scifact_recall@20": 0.82567,
      "eval_beir-trec-covid_ndcg@10": 0.51145,
      "eval_beir-trec-covid_recall@10": 0.556,
      "eval_beir-trec-covid_recall@100": 0.404,
      "eval_beir-trec-covid_recall@20": 0.503,
      "eval_beir-webis-touche2020_ndcg@10": 0.18226,
      "eval_beir-webis-touche2020_recall@10": 0.13639,
      "eval_beir-webis-touche2020_recall@100": 0.43122,
      "eval_beir-webis-touche2020_recall@20": 0.21071,
      "eval_senteval-avg_sts": 0.7291967840801405,
      "eval_senteval-sickr_spearman": 0.6983638574985077,
      "eval_senteval-stsb_spearman": 0.7600297106617734,
      "step": 70000,
      "test_accuracy": 94.59228515625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.31344741582870483,
      "test_doc_norm": 1.5325968265533447,
      "test_inbatch_neg_score": 0.9010056853294373,
      "test_inbatch_pos_score": 1.8864071369171143,
      "test_loss": 0.31344741582870483,
      "test_loss_align": 1.0511610507965088,
      "test_loss_unif": 3.477783679962158,
      "test_loss_unif_q@queue": 3.477783679962158,
      "test_norm_diff": 0.03206353634595871,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.5119207501411438,
      "test_query_norm": 1.5646604299545288,
      "test_queue_k_norm": 1.5326721668243408,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04340749233961105,
      "test_stdq": 0.043892648071050644,
      "test_stdqueue_k": 0.04903004691004753,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 58.0078,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2004,
      "doc_norm": 1.5316,
      "encoder_q-embeddings": 515.7128,
      "encoder_q-layer.0": 316.4032,
      "encoder_q-layer.1": 340.3427,
      "encoder_q-layer.10": 630.8476,
      "encoder_q-layer.11": 1408.4774,
      "encoder_q-layer.2": 391.0073,
      "encoder_q-layer.3": 421.0928,
      "encoder_q-layer.4": 476.8719,
      "encoder_q-layer.5": 488.3036,
      "encoder_q-layer.6": 560.9354,
      "encoder_q-layer.7": 676.2326,
      "encoder_q-layer.8": 735.8165,
      "encoder_q-layer.9": 667.2932,
      "epoch": 0.68,
      "inbatch_neg_score": 0.5279,
      "inbatch_pos_score": 1.2344,
      "learning_rate": 1.661111111111111e-05,
      "loss": 3.2004,
      "norm_diff": 0.0491,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 988.976,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5283,
      "query_norm": 1.4825,
      "queue_k_norm": 1.5351,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8184,
      "sent_len_1": 66.7368,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.0112,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 70100
    },
    {
      "accuracy": 51.8555,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1836,
      "doc_norm": 1.5292,
      "encoder_q-embeddings": 753.8774,
      "encoder_q-layer.0": 544.5876,
      "encoder_q-layer.1": 594.0212,
      "encoder_q-layer.10": 768.0432,
      "encoder_q-layer.11": 1553.374,
      "encoder_q-layer.2": 651.5072,
      "encoder_q-layer.3": 670.739,
      "encoder_q-layer.4": 729.2429,
      "encoder_q-layer.5": 723.2966,
      "encoder_q-layer.6": 802.7032,
      "encoder_q-layer.7": 893.1661,
      "encoder_q-layer.8": 891.2643,
      "encoder_q-layer.9": 679.1931,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5323,
      "inbatch_pos_score": 1.1562,
      "learning_rate": 1.655555555555556e-05,
      "loss": 3.1836,
      "norm_diff": 0.0632,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1240.5823,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5303,
      "query_norm": 1.466,
      "queue_k_norm": 1.5333,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0653,
      "sent_len_1": 66.5055,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1925,
      "stdk": 0.0487,
      "stdq": 0.0452,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 70200
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1986,
      "doc_norm": 1.5372,
      "encoder_q-embeddings": 2087.2517,
      "encoder_q-layer.0": 1472.8663,
      "encoder_q-layer.1": 1765.2465,
      "encoder_q-layer.10": 662.0661,
      "encoder_q-layer.11": 1416.5288,
      "encoder_q-layer.2": 1935.0385,
      "encoder_q-layer.3": 1905.8081,
      "encoder_q-layer.4": 1767.7029,
      "encoder_q-layer.5": 1724.814,
      "encoder_q-layer.6": 1685.4497,
      "encoder_q-layer.7": 1508.4396,
      "encoder_q-layer.8": 1086.5974,
      "encoder_q-layer.9": 698.4948,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5315,
      "inbatch_pos_score": 1.209,
      "learning_rate": 1.65e-05,
      "loss": 3.1986,
      "norm_diff": 0.0705,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2377.9269,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5312,
      "query_norm": 1.4667,
      "queue_k_norm": 1.5358,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0075,
      "sent_len_1": 66.7621,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7587,
      "stdk": 0.0491,
      "stdq": 0.0453,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 70300
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.198,
      "doc_norm": 1.5301,
      "encoder_q-embeddings": 602.7047,
      "encoder_q-layer.0": 416.8262,
      "encoder_q-layer.1": 463.1029,
      "encoder_q-layer.10": 637.6121,
      "encoder_q-layer.11": 1413.7876,
      "encoder_q-layer.2": 570.2717,
      "encoder_q-layer.3": 598.4441,
      "encoder_q-layer.4": 636.6793,
      "encoder_q-layer.5": 657.051,
      "encoder_q-layer.6": 694.2979,
      "encoder_q-layer.7": 741.4088,
      "encoder_q-layer.8": 746.6188,
      "encoder_q-layer.9": 657.2516,
      "epoch": 0.69,
      "inbatch_neg_score": 0.529,
      "inbatch_pos_score": 1.166,
      "learning_rate": 1.6444444444444447e-05,
      "loss": 3.198,
      "norm_diff": 0.0722,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1085.3205,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5283,
      "query_norm": 1.4579,
      "queue_k_norm": 1.535,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0845,
      "sent_len_1": 66.9094,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5725,
      "stdk": 0.0487,
      "stdq": 0.0449,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 70400
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.201,
      "doc_norm": 1.5434,
      "encoder_q-embeddings": 484.9706,
      "encoder_q-layer.0": 324.8301,
      "encoder_q-layer.1": 341.9171,
      "encoder_q-layer.10": 622.782,
      "encoder_q-layer.11": 1391.5027,
      "encoder_q-layer.2": 389.8105,
      "encoder_q-layer.3": 404.7368,
      "encoder_q-layer.4": 426.1296,
      "encoder_q-layer.5": 487.1759,
      "encoder_q-layer.6": 520.5556,
      "encoder_q-layer.7": 590.7158,
      "encoder_q-layer.8": 677.4696,
      "encoder_q-layer.9": 602.5097,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5314,
      "inbatch_pos_score": 1.2002,
      "learning_rate": 1.638888888888889e-05,
      "loss": 3.201,
      "norm_diff": 0.0669,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 944.3497,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5317,
      "query_norm": 1.4765,
      "queue_k_norm": 1.534,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0884,
      "sent_len_1": 66.872,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.39,
      "stdk": 0.0493,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 70500
    },
    {
      "accuracy": 58.6914,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1956,
      "doc_norm": 1.5373,
      "encoder_q-embeddings": 590.171,
      "encoder_q-layer.0": 431.7177,
      "encoder_q-layer.1": 462.8841,
      "encoder_q-layer.10": 633.7925,
      "encoder_q-layer.11": 1346.0123,
      "encoder_q-layer.2": 559.2192,
      "encoder_q-layer.3": 647.4349,
      "encoder_q-layer.4": 762.725,
      "encoder_q-layer.5": 775.5848,
      "encoder_q-layer.6": 843.2466,
      "encoder_q-layer.7": 881.8356,
      "encoder_q-layer.8": 992.3448,
      "encoder_q-layer.9": 702.7228,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5334,
      "inbatch_pos_score": 1.2471,
      "learning_rate": 1.6333333333333335e-05,
      "loss": 3.1956,
      "norm_diff": 0.0453,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1178.0942,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5322,
      "query_norm": 1.4921,
      "queue_k_norm": 1.5354,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2054,
      "sent_len_1": 66.8945,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.995,
      "stdk": 0.049,
      "stdq": 0.0462,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 70600
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1921,
      "doc_norm": 1.5376,
      "encoder_q-embeddings": 779.2072,
      "encoder_q-layer.0": 515.3904,
      "encoder_q-layer.1": 607.6839,
      "encoder_q-layer.10": 640.9584,
      "encoder_q-layer.11": 1374.0524,
      "encoder_q-layer.2": 730.933,
      "encoder_q-layer.3": 816.5665,
      "encoder_q-layer.4": 935.2493,
      "encoder_q-layer.5": 937.9548,
      "encoder_q-layer.6": 1015.7632,
      "encoder_q-layer.7": 868.7376,
      "encoder_q-layer.8": 777.7844,
      "encoder_q-layer.9": 608.5214,
      "epoch": 0.69,
      "inbatch_neg_score": 0.533,
      "inbatch_pos_score": 1.2207,
      "learning_rate": 1.6277777777777777e-05,
      "loss": 3.1921,
      "norm_diff": 0.0619,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1249.0474,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5332,
      "query_norm": 1.4756,
      "queue_k_norm": 1.5357,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1486,
      "sent_len_1": 66.7209,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.27,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 70700
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1811,
      "doc_norm": 1.5353,
      "encoder_q-embeddings": 482.6019,
      "encoder_q-layer.0": 315.6527,
      "encoder_q-layer.1": 339.0034,
      "encoder_q-layer.10": 567.5803,
      "encoder_q-layer.11": 1327.6178,
      "encoder_q-layer.2": 388.6182,
      "encoder_q-layer.3": 405.1554,
      "encoder_q-layer.4": 430.3468,
      "encoder_q-layer.5": 461.8609,
      "encoder_q-layer.6": 497.9529,
      "encoder_q-layer.7": 552.6882,
      "encoder_q-layer.8": 650.6743,
      "encoder_q-layer.9": 584.5176,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5339,
      "inbatch_pos_score": 1.2471,
      "learning_rate": 1.6222222222222223e-05,
      "loss": 3.1811,
      "norm_diff": 0.0459,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 914.7534,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5342,
      "query_norm": 1.4894,
      "queue_k_norm": 1.5366,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0927,
      "sent_len_1": 66.6019,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1012,
      "stdk": 0.0489,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 70800
    },
    {
      "accuracy": 57.8125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1984,
      "doc_norm": 1.5394,
      "encoder_q-embeddings": 1905.5875,
      "encoder_q-layer.0": 1358.2888,
      "encoder_q-layer.1": 1617.8768,
      "encoder_q-layer.10": 615.5238,
      "encoder_q-layer.11": 1362.2583,
      "encoder_q-layer.2": 2080.8726,
      "encoder_q-layer.3": 2170.4014,
      "encoder_q-layer.4": 2459.9766,
      "encoder_q-layer.5": 2346.6816,
      "encoder_q-layer.6": 2742.5527,
      "encoder_q-layer.7": 2934.394,
      "encoder_q-layer.8": 2525.5825,
      "encoder_q-layer.9": 1101.1082,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5317,
      "inbatch_pos_score": 1.2227,
      "learning_rate": 1.6166666666666665e-05,
      "loss": 3.1984,
      "norm_diff": 0.0625,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3128.8665,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5322,
      "query_norm": 1.4769,
      "queue_k_norm": 1.5356,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8584,
      "sent_len_1": 66.6066,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7688,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 70900
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1872,
      "doc_norm": 1.5316,
      "encoder_q-embeddings": 983.7383,
      "encoder_q-layer.0": 634.2096,
      "encoder_q-layer.1": 712.2656,
      "encoder_q-layer.10": 613.8813,
      "encoder_q-layer.11": 1377.4264,
      "encoder_q-layer.2": 826.2786,
      "encoder_q-layer.3": 766.6396,
      "encoder_q-layer.4": 821.8364,
      "encoder_q-layer.5": 828.7028,
      "encoder_q-layer.6": 811.2592,
      "encoder_q-layer.7": 824.8439,
      "encoder_q-layer.8": 860.9696,
      "encoder_q-layer.9": 644.8098,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5338,
      "inbatch_pos_score": 1.2236,
      "learning_rate": 1.6111111111111115e-05,
      "loss": 3.1872,
      "norm_diff": 0.04,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1274.5047,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5337,
      "query_norm": 1.4916,
      "queue_k_norm": 1.5366,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9825,
      "sent_len_1": 66.6328,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6687,
      "stdk": 0.0487,
      "stdq": 0.0462,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 71000
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1744,
      "doc_norm": 1.5417,
      "encoder_q-embeddings": 529.0616,
      "encoder_q-layer.0": 347.1833,
      "encoder_q-layer.1": 378.8297,
      "encoder_q-layer.10": 578.5715,
      "encoder_q-layer.11": 1315.9069,
      "encoder_q-layer.2": 424.1558,
      "encoder_q-layer.3": 438.1785,
      "encoder_q-layer.4": 503.0419,
      "encoder_q-layer.5": 498.591,
      "encoder_q-layer.6": 539.8322,
      "encoder_q-layer.7": 613.2227,
      "encoder_q-layer.8": 670.2361,
      "encoder_q-layer.9": 614.5303,
      "epoch": 0.69,
      "inbatch_neg_score": 0.5341,
      "inbatch_pos_score": 1.2578,
      "learning_rate": 1.6055555555555557e-05,
      "loss": 3.1744,
      "norm_diff": 0.0386,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 925.3141,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5342,
      "query_norm": 1.5031,
      "queue_k_norm": 1.5388,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1864,
      "sent_len_1": 66.9593,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1575,
      "stdk": 0.0491,
      "stdq": 0.0467,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71100
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1854,
      "doc_norm": 1.5364,
      "encoder_q-embeddings": 948.5603,
      "encoder_q-layer.0": 637.6948,
      "encoder_q-layer.1": 697.204,
      "encoder_q-layer.10": 600.7193,
      "encoder_q-layer.11": 1381.3829,
      "encoder_q-layer.2": 807.0599,
      "encoder_q-layer.3": 833.1315,
      "encoder_q-layer.4": 818.1985,
      "encoder_q-layer.5": 784.3917,
      "encoder_q-layer.6": 832.0074,
      "encoder_q-layer.7": 837.0083,
      "encoder_q-layer.8": 881.6041,
      "encoder_q-layer.9": 638.1913,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5323,
      "inbatch_pos_score": 1.2061,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 3.1854,
      "norm_diff": 0.0726,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1278.6828,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5327,
      "query_norm": 1.4638,
      "queue_k_norm": 1.5377,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7364,
      "sent_len_1": 66.5152,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.325,
      "stdk": 0.0489,
      "stdq": 0.0451,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71200
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1879,
      "doc_norm": 1.5438,
      "encoder_q-embeddings": 965.7256,
      "encoder_q-layer.0": 639.279,
      "encoder_q-layer.1": 690.4075,
      "encoder_q-layer.10": 1144.7428,
      "encoder_q-layer.11": 2476.3455,
      "encoder_q-layer.2": 769.8522,
      "encoder_q-layer.3": 833.4374,
      "encoder_q-layer.4": 894.1732,
      "encoder_q-layer.5": 877.6177,
      "encoder_q-layer.6": 999.1593,
      "encoder_q-layer.7": 1206.4763,
      "encoder_q-layer.8": 1383.5122,
      "encoder_q-layer.9": 1161.969,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5353,
      "inbatch_pos_score": 1.2451,
      "learning_rate": 1.5944444444444445e-05,
      "loss": 3.1879,
      "norm_diff": 0.0546,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1760.3388,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5352,
      "query_norm": 1.4892,
      "queue_k_norm": 1.5374,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9131,
      "sent_len_1": 66.8312,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.305,
      "stdk": 0.0492,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 71300
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2038,
      "doc_norm": 1.5429,
      "encoder_q-embeddings": 1230.0942,
      "encoder_q-layer.0": 846.2477,
      "encoder_q-layer.1": 920.123,
      "encoder_q-layer.10": 1285.4615,
      "encoder_q-layer.11": 2907.335,
      "encoder_q-layer.2": 1009.33,
      "encoder_q-layer.3": 1061.0415,
      "encoder_q-layer.4": 1117.7844,
      "encoder_q-layer.5": 1154.4823,
      "encoder_q-layer.6": 1236.4139,
      "encoder_q-layer.7": 1381.965,
      "encoder_q-layer.8": 1449.3081,
      "encoder_q-layer.9": 1248.3583,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5381,
      "inbatch_pos_score": 1.2188,
      "learning_rate": 1.588888888888889e-05,
      "loss": 3.2038,
      "norm_diff": 0.0436,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2130.7931,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5381,
      "query_norm": 1.4993,
      "queue_k_norm": 1.5401,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.777,
      "sent_len_1": 66.6317,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3113,
      "stdk": 0.0491,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71400
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1672,
      "doc_norm": 1.5398,
      "encoder_q-embeddings": 1472.0751,
      "encoder_q-layer.0": 1082.6639,
      "encoder_q-layer.1": 1186.0702,
      "encoder_q-layer.10": 1159.8062,
      "encoder_q-layer.11": 2612.5283,
      "encoder_q-layer.2": 1289.652,
      "encoder_q-layer.3": 1372.4067,
      "encoder_q-layer.4": 1460.2654,
      "encoder_q-layer.5": 1465.7311,
      "encoder_q-layer.6": 1572.6449,
      "encoder_q-layer.7": 1735.0814,
      "encoder_q-layer.8": 1647.5488,
      "encoder_q-layer.9": 1189.8567,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5372,
      "inbatch_pos_score": 1.2041,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 3.1672,
      "norm_diff": 0.0662,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2326.7663,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5371,
      "query_norm": 1.4736,
      "queue_k_norm": 1.5407,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0597,
      "sent_len_1": 66.8194,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.64,
      "stdk": 0.049,
      "stdq": 0.0453,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71500
    },
    {
      "accuracy": 58.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1634,
      "doc_norm": 1.5363,
      "encoder_q-embeddings": 932.7405,
      "encoder_q-layer.0": 601.7432,
      "encoder_q-layer.1": 632.7126,
      "encoder_q-layer.10": 1143.2515,
      "encoder_q-layer.11": 2442.6924,
      "encoder_q-layer.2": 722.0427,
      "encoder_q-layer.3": 749.0936,
      "encoder_q-layer.4": 802.7853,
      "encoder_q-layer.5": 851.2307,
      "encoder_q-layer.6": 969.6721,
      "encoder_q-layer.7": 1127.7356,
      "encoder_q-layer.8": 1316.3347,
      "encoder_q-layer.9": 1173.0963,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5357,
      "inbatch_pos_score": 1.25,
      "learning_rate": 1.577777777777778e-05,
      "loss": 3.1634,
      "norm_diff": 0.0355,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1716.2884,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5347,
      "query_norm": 1.501,
      "queue_k_norm": 1.5397,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9166,
      "sent_len_1": 66.6555,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2212,
      "stdk": 0.0489,
      "stdq": 0.0465,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71600
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1794,
      "doc_norm": 1.5418,
      "encoder_q-embeddings": 1068.6145,
      "encoder_q-layer.0": 712.7808,
      "encoder_q-layer.1": 796.6774,
      "encoder_q-layer.10": 1298.5408,
      "encoder_q-layer.11": 2676.0078,
      "encoder_q-layer.2": 922.3505,
      "encoder_q-layer.3": 941.6613,
      "encoder_q-layer.4": 1009.9976,
      "encoder_q-layer.5": 1065.5171,
      "encoder_q-layer.6": 1115.4706,
      "encoder_q-layer.7": 1266.5148,
      "encoder_q-layer.8": 1414.9316,
      "encoder_q-layer.9": 1237.6165,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5356,
      "inbatch_pos_score": 1.2324,
      "learning_rate": 1.5722222222222225e-05,
      "loss": 3.1794,
      "norm_diff": 0.0455,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1941.8542,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5361,
      "query_norm": 1.4962,
      "queue_k_norm": 1.5394,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2844,
      "sent_len_1": 66.8745,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9762,
      "stdk": 0.0491,
      "stdq": 0.0463,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71700
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.184,
      "doc_norm": 1.5333,
      "encoder_q-embeddings": 1026.6343,
      "encoder_q-layer.0": 658.447,
      "encoder_q-layer.1": 693.7869,
      "encoder_q-layer.10": 1191.3671,
      "encoder_q-layer.11": 2671.0923,
      "encoder_q-layer.2": 786.9291,
      "encoder_q-layer.3": 832.7916,
      "encoder_q-layer.4": 883.1344,
      "encoder_q-layer.5": 938.7551,
      "encoder_q-layer.6": 1018.2356,
      "encoder_q-layer.7": 1178.0758,
      "encoder_q-layer.8": 1322.8929,
      "encoder_q-layer.9": 1215.6926,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5361,
      "inbatch_pos_score": 1.2285,
      "learning_rate": 1.5666666666666667e-05,
      "loss": 3.184,
      "norm_diff": 0.0349,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1856.7294,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5356,
      "query_norm": 1.4984,
      "queue_k_norm": 1.5407,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8913,
      "sent_len_1": 66.8694,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9225,
      "stdk": 0.0487,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71800
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1745,
      "doc_norm": 1.5375,
      "encoder_q-embeddings": 1465.3086,
      "encoder_q-layer.0": 973.243,
      "encoder_q-layer.1": 1014.2963,
      "encoder_q-layer.10": 1291.4965,
      "encoder_q-layer.11": 2827.3767,
      "encoder_q-layer.2": 1180.2833,
      "encoder_q-layer.3": 1230.1434,
      "encoder_q-layer.4": 1249.8862,
      "encoder_q-layer.5": 1250.7499,
      "encoder_q-layer.6": 1341.8885,
      "encoder_q-layer.7": 1406.4265,
      "encoder_q-layer.8": 1354.6614,
      "encoder_q-layer.9": 1216.0846,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5343,
      "inbatch_pos_score": 1.21,
      "learning_rate": 1.5611111111111113e-05,
      "loss": 3.1745,
      "norm_diff": 0.0586,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2207.5096,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5347,
      "query_norm": 1.4788,
      "queue_k_norm": 1.5395,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0393,
      "sent_len_1": 66.773,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8537,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 71900
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.195,
      "doc_norm": 1.5381,
      "encoder_q-embeddings": 1199.856,
      "encoder_q-layer.0": 803.529,
      "encoder_q-layer.1": 926.9163,
      "encoder_q-layer.10": 1224.8711,
      "encoder_q-layer.11": 2793.5403,
      "encoder_q-layer.2": 1055.5001,
      "encoder_q-layer.3": 1070.9636,
      "encoder_q-layer.4": 1096.0175,
      "encoder_q-layer.5": 1137.4442,
      "encoder_q-layer.6": 1260.6923,
      "encoder_q-layer.7": 1365.7776,
      "encoder_q-layer.8": 1600.8005,
      "encoder_q-layer.9": 1326.332,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5357,
      "inbatch_pos_score": 1.168,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 3.195,
      "norm_diff": 0.0672,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2088.2895,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5356,
      "query_norm": 1.4709,
      "queue_k_norm": 1.5397,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1721,
      "sent_len_1": 67.0348,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4963,
      "stdk": 0.0489,
      "stdq": 0.0451,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72000
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2046,
      "doc_norm": 1.5439,
      "encoder_q-embeddings": 963.538,
      "encoder_q-layer.0": 624.4863,
      "encoder_q-layer.1": 638.0428,
      "encoder_q-layer.10": 1135.1042,
      "encoder_q-layer.11": 2671.5867,
      "encoder_q-layer.2": 710.6894,
      "encoder_q-layer.3": 751.0983,
      "encoder_q-layer.4": 801.1094,
      "encoder_q-layer.5": 879.6038,
      "encoder_q-layer.6": 959.3598,
      "encoder_q-layer.7": 1136.3459,
      "encoder_q-layer.8": 1297.0698,
      "encoder_q-layer.9": 1174.1433,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5377,
      "inbatch_pos_score": 1.21,
      "learning_rate": 1.55e-05,
      "loss": 3.2046,
      "norm_diff": 0.0623,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1822.1386,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5376,
      "query_norm": 1.4816,
      "queue_k_norm": 1.5398,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6019,
      "sent_len_1": 66.5596,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.405,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72100
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1852,
      "doc_norm": 1.5405,
      "encoder_q-embeddings": 1020.0333,
      "encoder_q-layer.0": 679.1085,
      "encoder_q-layer.1": 703.8406,
      "encoder_q-layer.10": 1291.4701,
      "encoder_q-layer.11": 2686.6746,
      "encoder_q-layer.2": 763.9773,
      "encoder_q-layer.3": 829.0822,
      "encoder_q-layer.4": 838.7519,
      "encoder_q-layer.5": 897.1908,
      "encoder_q-layer.6": 1046.0139,
      "encoder_q-layer.7": 1153.1458,
      "encoder_q-layer.8": 1313.1279,
      "encoder_q-layer.9": 1235.2695,
      "epoch": 0.7,
      "inbatch_neg_score": 0.5417,
      "inbatch_pos_score": 1.2314,
      "learning_rate": 1.5444444444444446e-05,
      "loss": 3.1852,
      "norm_diff": 0.0421,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1858.986,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.542,
      "query_norm": 1.4983,
      "queue_k_norm": 1.5389,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9828,
      "sent_len_1": 66.6382,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1287,
      "stdk": 0.049,
      "stdq": 0.0463,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 72200
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1914,
      "doc_norm": 1.5408,
      "encoder_q-embeddings": 1782.5144,
      "encoder_q-layer.0": 1231.6986,
      "encoder_q-layer.1": 1430.2997,
      "encoder_q-layer.10": 1239.2374,
      "encoder_q-layer.11": 2794.6853,
      "encoder_q-layer.2": 1710.9081,
      "encoder_q-layer.3": 1857.2864,
      "encoder_q-layer.4": 1912.1774,
      "encoder_q-layer.5": 1713.2714,
      "encoder_q-layer.6": 1629.1248,
      "encoder_q-layer.7": 1530.9055,
      "encoder_q-layer.8": 1483.1257,
      "encoder_q-layer.9": 1230.3801,
      "epoch": 0.71,
      "inbatch_neg_score": 0.548,
      "inbatch_pos_score": 1.2266,
      "learning_rate": 1.538888888888889e-05,
      "loss": 3.1914,
      "norm_diff": 0.0527,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2577.6464,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5474,
      "query_norm": 1.488,
      "queue_k_norm": 1.5399,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0431,
      "sent_len_1": 66.6913,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.165,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72300
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1735,
      "doc_norm": 1.5434,
      "encoder_q-embeddings": 1057.2794,
      "encoder_q-layer.0": 691.0752,
      "encoder_q-layer.1": 759.4733,
      "encoder_q-layer.10": 1326.1072,
      "encoder_q-layer.11": 2938.0374,
      "encoder_q-layer.2": 833.6299,
      "encoder_q-layer.3": 893.9981,
      "encoder_q-layer.4": 1011.2407,
      "encoder_q-layer.5": 1033.0869,
      "encoder_q-layer.6": 1146.9874,
      "encoder_q-layer.7": 1267.506,
      "encoder_q-layer.8": 1457.3942,
      "encoder_q-layer.9": 1284.1876,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5515,
      "inbatch_pos_score": 1.2314,
      "learning_rate": 1.5333333333333334e-05,
      "loss": 3.1735,
      "norm_diff": 0.0415,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2040.0857,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5513,
      "query_norm": 1.5019,
      "queue_k_norm": 1.5396,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.725,
      "sent_len_1": 66.544,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9325,
      "stdk": 0.0491,
      "stdq": 0.0462,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 72400
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1455,
      "doc_norm": 1.5367,
      "encoder_q-embeddings": 956.7857,
      "encoder_q-layer.0": 633.1376,
      "encoder_q-layer.1": 670.0874,
      "encoder_q-layer.10": 1200.874,
      "encoder_q-layer.11": 2664.6772,
      "encoder_q-layer.2": 768.853,
      "encoder_q-layer.3": 833.6395,
      "encoder_q-layer.4": 856.0808,
      "encoder_q-layer.5": 901.9776,
      "encoder_q-layer.6": 1063.5857,
      "encoder_q-layer.7": 1229.8795,
      "encoder_q-layer.8": 1332.2993,
      "encoder_q-layer.9": 1248.9725,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5552,
      "inbatch_pos_score": 1.2441,
      "learning_rate": 1.527777777777778e-05,
      "loss": 3.1455,
      "norm_diff": 0.0286,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1835.1272,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5537,
      "query_norm": 1.5082,
      "queue_k_norm": 1.5422,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9084,
      "sent_len_1": 67.0096,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1425,
      "stdk": 0.0488,
      "stdq": 0.0465,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72500
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1753,
      "doc_norm": 1.5413,
      "encoder_q-embeddings": 1366.1344,
      "encoder_q-layer.0": 897.4189,
      "encoder_q-layer.1": 1075.3844,
      "encoder_q-layer.10": 1400.9963,
      "encoder_q-layer.11": 2938.1736,
      "encoder_q-layer.2": 1327.1598,
      "encoder_q-layer.3": 1451.5048,
      "encoder_q-layer.4": 1668.7462,
      "encoder_q-layer.5": 1618.936,
      "encoder_q-layer.6": 1633.665,
      "encoder_q-layer.7": 1768.9148,
      "encoder_q-layer.8": 1680.6855,
      "encoder_q-layer.9": 1404.2271,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5547,
      "inbatch_pos_score": 1.25,
      "learning_rate": 1.5222222222222224e-05,
      "loss": 3.1753,
      "norm_diff": 0.0354,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2447.2976,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5547,
      "query_norm": 1.506,
      "queue_k_norm": 1.542,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0173,
      "sent_len_1": 66.9014,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.795,
      "stdk": 0.0489,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72600
    },
    {
      "accuracy": 62.1094,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1559,
      "doc_norm": 1.5443,
      "encoder_q-embeddings": 1382.0852,
      "encoder_q-layer.0": 939.1196,
      "encoder_q-layer.1": 1137.7272,
      "encoder_q-layer.10": 1333.1063,
      "encoder_q-layer.11": 2741.1362,
      "encoder_q-layer.2": 1524.5566,
      "encoder_q-layer.3": 1738.2177,
      "encoder_q-layer.4": 1823.5594,
      "encoder_q-layer.5": 2199.7002,
      "encoder_q-layer.6": 2265.1729,
      "encoder_q-layer.7": 2492.9666,
      "encoder_q-layer.8": 1905.3916,
      "encoder_q-layer.9": 1399.0282,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5571,
      "inbatch_pos_score": 1.2949,
      "learning_rate": 1.5166666666666668e-05,
      "loss": 3.1559,
      "norm_diff": 0.0386,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2696.5216,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5571,
      "query_norm": 1.5057,
      "queue_k_norm": 1.5424,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1229,
      "sent_len_1": 66.8711,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7912,
      "stdk": 0.0491,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72700
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1819,
      "doc_norm": 1.5321,
      "encoder_q-embeddings": 980.1714,
      "encoder_q-layer.0": 625.0706,
      "encoder_q-layer.1": 667.2089,
      "encoder_q-layer.10": 1319.1206,
      "encoder_q-layer.11": 3006.425,
      "encoder_q-layer.2": 754.1843,
      "encoder_q-layer.3": 788.7116,
      "encoder_q-layer.4": 851.3588,
      "encoder_q-layer.5": 860.2919,
      "encoder_q-layer.6": 1016.0336,
      "encoder_q-layer.7": 1230.1238,
      "encoder_q-layer.8": 1446.6981,
      "encoder_q-layer.9": 1284.7523,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5581,
      "inbatch_pos_score": 1.2354,
      "learning_rate": 1.5111111111111112e-05,
      "loss": 3.1819,
      "norm_diff": 0.0353,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1915.6816,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5581,
      "query_norm": 1.4968,
      "queue_k_norm": 1.5426,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9767,
      "sent_len_1": 66.5692,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4313,
      "stdk": 0.0486,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72800
    },
    {
      "accuracy": 51.3672,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1783,
      "doc_norm": 1.5412,
      "encoder_q-embeddings": 2256.7512,
      "encoder_q-layer.0": 1585.0378,
      "encoder_q-layer.1": 1908.9926,
      "encoder_q-layer.10": 1322.6052,
      "encoder_q-layer.11": 2923.0339,
      "encoder_q-layer.2": 2538.4495,
      "encoder_q-layer.3": 2642.4526,
      "encoder_q-layer.4": 2745.3403,
      "encoder_q-layer.5": 2429.6995,
      "encoder_q-layer.6": 2260.0715,
      "encoder_q-layer.7": 2308.7078,
      "encoder_q-layer.8": 1907.8826,
      "encoder_q-layer.9": 1301.8066,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5588,
      "inbatch_pos_score": 1.2021,
      "learning_rate": 1.5055555555555556e-05,
      "loss": 3.1783,
      "norm_diff": 0.0608,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3374.6156,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5581,
      "query_norm": 1.4804,
      "queue_k_norm": 1.5438,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8843,
      "sent_len_1": 66.8672,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2862,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 72900
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.2022,
      "doc_norm": 1.5465,
      "encoder_q-embeddings": 1776.5383,
      "encoder_q-layer.0": 1266.0554,
      "encoder_q-layer.1": 1305.1581,
      "encoder_q-layer.10": 1312.3555,
      "encoder_q-layer.11": 2767.3096,
      "encoder_q-layer.2": 1568.8806,
      "encoder_q-layer.3": 1632.4921,
      "encoder_q-layer.4": 1744.3792,
      "encoder_q-layer.5": 1841.4377,
      "encoder_q-layer.6": 2040.5695,
      "encoder_q-layer.7": 1948.2245,
      "encoder_q-layer.8": 1877.6686,
      "encoder_q-layer.9": 1432.355,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5564,
      "inbatch_pos_score": 1.2578,
      "learning_rate": 1.5e-05,
      "loss": 3.2022,
      "norm_diff": 0.0485,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2704.854,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5566,
      "query_norm": 1.4979,
      "queue_k_norm": 1.5436,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8898,
      "sent_len_1": 66.8137,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.96,
      "stdk": 0.0491,
      "stdq": 0.0463,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73000
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1779,
      "doc_norm": 1.5503,
      "encoder_q-embeddings": 1320.5981,
      "encoder_q-layer.0": 889.962,
      "encoder_q-layer.1": 1072.604,
      "encoder_q-layer.10": 1201.1991,
      "encoder_q-layer.11": 2713.6729,
      "encoder_q-layer.2": 1251.2932,
      "encoder_q-layer.3": 1383.3165,
      "encoder_q-layer.4": 1532.2915,
      "encoder_q-layer.5": 1609.4688,
      "encoder_q-layer.6": 1776.6464,
      "encoder_q-layer.7": 1679.5671,
      "encoder_q-layer.8": 1540.7937,
      "encoder_q-layer.9": 1220.278,
      "epoch": 0.71,
      "inbatch_neg_score": 0.56,
      "inbatch_pos_score": 1.2617,
      "learning_rate": 1.4944444444444444e-05,
      "loss": 3.1779,
      "norm_diff": 0.0601,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2323.6512,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5596,
      "query_norm": 1.4902,
      "queue_k_norm": 1.5448,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9164,
      "sent_len_1": 66.7928,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.6475,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73100
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1727,
      "doc_norm": 1.5453,
      "encoder_q-embeddings": 1105.4956,
      "encoder_q-layer.0": 744.0076,
      "encoder_q-layer.1": 807.7679,
      "encoder_q-layer.10": 1526.9027,
      "encoder_q-layer.11": 3126.978,
      "encoder_q-layer.2": 914.2983,
      "encoder_q-layer.3": 958.4001,
      "encoder_q-layer.4": 1053.8214,
      "encoder_q-layer.5": 1141.5128,
      "encoder_q-layer.6": 1201.1501,
      "encoder_q-layer.7": 1379.9718,
      "encoder_q-layer.8": 1585.1754,
      "encoder_q-layer.9": 1428.0548,
      "epoch": 0.71,
      "inbatch_neg_score": 0.5596,
      "inbatch_pos_score": 1.2461,
      "learning_rate": 1.4888888888888888e-05,
      "loss": 3.1727,
      "norm_diff": 0.0692,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2154.9034,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5605,
      "query_norm": 1.4761,
      "queue_k_norm": 1.542,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8558,
      "sent_len_1": 66.5763,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5513,
      "stdk": 0.049,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 73200
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.186,
      "doc_norm": 1.5456,
      "encoder_q-embeddings": 4775.9546,
      "encoder_q-layer.0": 3182.3481,
      "encoder_q-layer.1": 3373.8032,
      "encoder_q-layer.10": 2708.7681,
      "encoder_q-layer.11": 5716.1406,
      "encoder_q-layer.2": 3718.3706,
      "encoder_q-layer.3": 3984.1104,
      "encoder_q-layer.4": 4623.1948,
      "encoder_q-layer.5": 4692.0396,
      "encoder_q-layer.6": 4481.0938,
      "encoder_q-layer.7": 5305.48,
      "encoder_q-layer.8": 4883.3623,
      "encoder_q-layer.9": 2854.0037,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5601,
      "inbatch_pos_score": 1.25,
      "learning_rate": 1.4833333333333336e-05,
      "loss": 3.186,
      "norm_diff": 0.0601,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6497.3825,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5596,
      "query_norm": 1.4855,
      "queue_k_norm": 1.5458,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9192,
      "sent_len_1": 66.7742,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.1188,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73300
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1479,
      "doc_norm": 1.5503,
      "encoder_q-embeddings": 2135.3315,
      "encoder_q-layer.0": 1393.318,
      "encoder_q-layer.1": 1564.1238,
      "encoder_q-layer.10": 2588.0894,
      "encoder_q-layer.11": 5638.6821,
      "encoder_q-layer.2": 1788.6798,
      "encoder_q-layer.3": 1910.8817,
      "encoder_q-layer.4": 2114.97,
      "encoder_q-layer.5": 2129.5303,
      "encoder_q-layer.6": 2323.5862,
      "encoder_q-layer.7": 2724.7512,
      "encoder_q-layer.8": 2914.6274,
      "encoder_q-layer.9": 2613.0667,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5606,
      "inbatch_pos_score": 1.25,
      "learning_rate": 1.477777777777778e-05,
      "loss": 3.1479,
      "norm_diff": 0.0655,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4078.4348,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5605,
      "query_norm": 1.4848,
      "queue_k_norm": 1.5473,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.271,
      "sent_len_1": 66.9835,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3225,
      "stdk": 0.0492,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 73400
    },
    {
      "accuracy": 58.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1767,
      "doc_norm": 1.5478,
      "encoder_q-embeddings": 2879.199,
      "encoder_q-layer.0": 1976.3889,
      "encoder_q-layer.1": 2309.5571,
      "encoder_q-layer.10": 2542.9678,
      "encoder_q-layer.11": 5406.4795,
      "encoder_q-layer.2": 2609.2537,
      "encoder_q-layer.3": 2722.1729,
      "encoder_q-layer.4": 3044.2354,
      "encoder_q-layer.5": 3183.6204,
      "encoder_q-layer.6": 3554.7119,
      "encoder_q-layer.7": 4151.7007,
      "encoder_q-layer.8": 3868.1228,
      "encoder_q-layer.9": 2644.3853,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5606,
      "inbatch_pos_score": 1.2549,
      "learning_rate": 1.4722222222222224e-05,
      "loss": 3.1767,
      "norm_diff": 0.0644,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4946.6773,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5601,
      "query_norm": 1.4834,
      "queue_k_norm": 1.5486,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0967,
      "sent_len_1": 67.3128,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5687,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 73500
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1736,
      "doc_norm": 1.5508,
      "encoder_q-embeddings": 2637.3401,
      "encoder_q-layer.0": 1782.7277,
      "encoder_q-layer.1": 1913.7959,
      "encoder_q-layer.10": 2413.231,
      "encoder_q-layer.11": 5667.415,
      "encoder_q-layer.2": 2187.2498,
      "encoder_q-layer.3": 2260.1665,
      "encoder_q-layer.4": 2320.5464,
      "encoder_q-layer.5": 2527.384,
      "encoder_q-layer.6": 2613.2351,
      "encoder_q-layer.7": 2477.113,
      "encoder_q-layer.8": 2779.9819,
      "encoder_q-layer.9": 2473.6982,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5631,
      "inbatch_pos_score": 1.2441,
      "learning_rate": 1.4666666666666668e-05,
      "loss": 3.1736,
      "norm_diff": 0.0525,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4267.7951,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.563,
      "query_norm": 1.4984,
      "queue_k_norm": 1.5462,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1046,
      "sent_len_1": 66.8505,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4338,
      "stdk": 0.0492,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73600
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1873,
      "doc_norm": 1.5476,
      "encoder_q-embeddings": 9180.7822,
      "encoder_q-layer.0": 6834.5381,
      "encoder_q-layer.1": 9043.832,
      "encoder_q-layer.10": 2564.552,
      "encoder_q-layer.11": 5620.5962,
      "encoder_q-layer.2": 10681.5117,
      "encoder_q-layer.3": 10543.3027,
      "encoder_q-layer.4": 12531.3721,
      "encoder_q-layer.5": 12283.7656,
      "encoder_q-layer.6": 10125.8516,
      "encoder_q-layer.7": 6991.7002,
      "encoder_q-layer.8": 3606.9109,
      "encoder_q-layer.9": 2682.0381,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5619,
      "inbatch_pos_score": 1.2529,
      "learning_rate": 1.4611111111111112e-05,
      "loss": 3.1873,
      "norm_diff": 0.0574,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12673.5548,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.562,
      "query_norm": 1.4902,
      "queue_k_norm": 1.5481,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.857,
      "sent_len_1": 66.7291,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.5062,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73700
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1902,
      "doc_norm": 1.5461,
      "encoder_q-embeddings": 2945.0659,
      "encoder_q-layer.0": 1978.5757,
      "encoder_q-layer.1": 2218.1934,
      "encoder_q-layer.10": 2394.0613,
      "encoder_q-layer.11": 5491.7256,
      "encoder_q-layer.2": 2589.1882,
      "encoder_q-layer.3": 2821.7188,
      "encoder_q-layer.4": 3016.8611,
      "encoder_q-layer.5": 3632.8818,
      "encoder_q-layer.6": 3682.0234,
      "encoder_q-layer.7": 3763.4468,
      "encoder_q-layer.8": 3811.658,
      "encoder_q-layer.9": 2692.0144,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5665,
      "inbatch_pos_score": 1.2256,
      "learning_rate": 1.4555555555555556e-05,
      "loss": 3.1902,
      "norm_diff": 0.0687,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4933.289,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5659,
      "query_norm": 1.4774,
      "queue_k_norm": 1.5482,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9966,
      "sent_len_1": 66.843,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.64,
      "stdk": 0.0489,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73800
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.154,
      "doc_norm": 1.5449,
      "encoder_q-embeddings": 1963.1678,
      "encoder_q-layer.0": 1311.3064,
      "encoder_q-layer.1": 1438.8348,
      "encoder_q-layer.10": 2547.9529,
      "encoder_q-layer.11": 5531.4458,
      "encoder_q-layer.2": 1706.9778,
      "encoder_q-layer.3": 1828.1493,
      "encoder_q-layer.4": 1976.5031,
      "encoder_q-layer.5": 2008.2637,
      "encoder_q-layer.6": 2200.2373,
      "encoder_q-layer.7": 2480.9993,
      "encoder_q-layer.8": 2916.833,
      "encoder_q-layer.9": 2450.3875,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5692,
      "inbatch_pos_score": 1.2715,
      "learning_rate": 1.45e-05,
      "loss": 3.154,
      "norm_diff": 0.0325,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3831.5839,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5693,
      "query_norm": 1.5125,
      "queue_k_norm": 1.547,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1228,
      "sent_len_1": 66.5413,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9062,
      "stdk": 0.0489,
      "stdq": 0.0466,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 73900
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.165,
      "doc_norm": 1.551,
      "encoder_q-embeddings": 2045.7683,
      "encoder_q-layer.0": 1314.6774,
      "encoder_q-layer.1": 1404.4771,
      "encoder_q-layer.10": 2649.1448,
      "encoder_q-layer.11": 5719.7603,
      "encoder_q-layer.2": 1557.0284,
      "encoder_q-layer.3": 1630.3934,
      "encoder_q-layer.4": 1790.9945,
      "encoder_q-layer.5": 1846.1086,
      "encoder_q-layer.6": 2091.6504,
      "encoder_q-layer.7": 2533.0791,
      "encoder_q-layer.8": 2895.127,
      "encoder_q-layer.9": 2581.6055,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5676,
      "inbatch_pos_score": 1.25,
      "learning_rate": 1.4444444444444444e-05,
      "loss": 3.165,
      "norm_diff": 0.0632,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3960.5231,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5674,
      "query_norm": 1.4877,
      "queue_k_norm": 1.5466,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8725,
      "sent_len_1": 66.7215,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6838,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 74000
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1812,
      "doc_norm": 1.5499,
      "encoder_q-embeddings": 2057.2131,
      "encoder_q-layer.0": 1374.24,
      "encoder_q-layer.1": 1568.3915,
      "encoder_q-layer.10": 2744.8362,
      "encoder_q-layer.11": 5662.5532,
      "encoder_q-layer.2": 1720.4698,
      "encoder_q-layer.3": 1759.5288,
      "encoder_q-layer.4": 1868.1562,
      "encoder_q-layer.5": 2005.7847,
      "encoder_q-layer.6": 2241.0059,
      "encoder_q-layer.7": 2547.6477,
      "encoder_q-layer.8": 2844.1973,
      "encoder_q-layer.9": 2661.7791,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5707,
      "inbatch_pos_score": 1.252,
      "learning_rate": 1.438888888888889e-05,
      "loss": 3.1812,
      "norm_diff": 0.0653,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3956.1758,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5713,
      "query_norm": 1.4846,
      "queue_k_norm": 1.5484,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9154,
      "sent_len_1": 66.8857,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7775,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74100
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.181,
      "doc_norm": 1.5441,
      "encoder_q-embeddings": 2132.2742,
      "encoder_q-layer.0": 1445.2349,
      "encoder_q-layer.1": 1569.752,
      "encoder_q-layer.10": 2605.0398,
      "encoder_q-layer.11": 5749.0854,
      "encoder_q-layer.2": 1793.3547,
      "encoder_q-layer.3": 1937.5059,
      "encoder_q-layer.4": 2240.4143,
      "encoder_q-layer.5": 2297.1946,
      "encoder_q-layer.6": 2547.5247,
      "encoder_q-layer.7": 2891.6091,
      "encoder_q-layer.8": 3394.3682,
      "encoder_q-layer.9": 2703.5581,
      "epoch": 0.72,
      "inbatch_neg_score": 0.5738,
      "inbatch_pos_score": 1.2354,
      "learning_rate": 1.4333333333333334e-05,
      "loss": 3.181,
      "norm_diff": 0.056,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4215.2319,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5742,
      "query_norm": 1.4882,
      "queue_k_norm": 1.5484,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0381,
      "sent_len_1": 66.7226,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6025,
      "stdk": 0.0488,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74200
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1971,
      "doc_norm": 1.5513,
      "encoder_q-embeddings": 2507.4348,
      "encoder_q-layer.0": 1695.3728,
      "encoder_q-layer.1": 1903.7732,
      "encoder_q-layer.10": 2624.97,
      "encoder_q-layer.11": 5785.106,
      "encoder_q-layer.2": 2099.7253,
      "encoder_q-layer.3": 2184.3726,
      "encoder_q-layer.4": 2494.7854,
      "encoder_q-layer.5": 2576.0596,
      "encoder_q-layer.6": 2649.0005,
      "encoder_q-layer.7": 2879.0752,
      "encoder_q-layer.8": 3265.0156,
      "encoder_q-layer.9": 2762.2158,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5766,
      "inbatch_pos_score": 1.2432,
      "learning_rate": 1.427777777777778e-05,
      "loss": 3.1971,
      "norm_diff": 0.0408,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4346.5225,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5762,
      "query_norm": 1.5105,
      "queue_k_norm": 1.5484,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8849,
      "sent_len_1": 66.6012,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5188,
      "stdk": 0.049,
      "stdq": 0.0463,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 74300
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.17,
      "doc_norm": 1.5462,
      "encoder_q-embeddings": 2837.8813,
      "encoder_q-layer.0": 1806.2917,
      "encoder_q-layer.1": 2120.8926,
      "encoder_q-layer.10": 2584.8296,
      "encoder_q-layer.11": 5673.3184,
      "encoder_q-layer.2": 2476.1406,
      "encoder_q-layer.3": 2699.0688,
      "encoder_q-layer.4": 3133.8174,
      "encoder_q-layer.5": 3288.8823,
      "encoder_q-layer.6": 3508.032,
      "encoder_q-layer.7": 3580.135,
      "encoder_q-layer.8": 3341.1057,
      "encoder_q-layer.9": 2868.8608,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5744,
      "inbatch_pos_score": 1.2822,
      "learning_rate": 1.4222222222222224e-05,
      "loss": 3.17,
      "norm_diff": 0.0153,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4754.1507,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5747,
      "query_norm": 1.531,
      "queue_k_norm": 1.5508,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.902,
      "sent_len_1": 66.6944,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.34,
      "stdk": 0.0488,
      "stdq": 0.0472,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74400
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1695,
      "doc_norm": 1.5415,
      "encoder_q-embeddings": 1935.1073,
      "encoder_q-layer.0": 1276.0426,
      "encoder_q-layer.1": 1338.9579,
      "encoder_q-layer.10": 2692.7441,
      "encoder_q-layer.11": 6060.7949,
      "encoder_q-layer.2": 1545.8142,
      "encoder_q-layer.3": 1607.6183,
      "encoder_q-layer.4": 1723.0992,
      "encoder_q-layer.5": 1825.5347,
      "encoder_q-layer.6": 2018.031,
      "encoder_q-layer.7": 2319.1597,
      "encoder_q-layer.8": 2848.0645,
      "encoder_q-layer.9": 2534.0303,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5801,
      "inbatch_pos_score": 1.2451,
      "learning_rate": 1.4166666666666668e-05,
      "loss": 3.1695,
      "norm_diff": 0.0409,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3929.3242,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5811,
      "query_norm": 1.5006,
      "queue_k_norm": 1.5514,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7414,
      "sent_len_1": 66.5487,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1375,
      "stdk": 0.0486,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74500
    },
    {
      "accuracy": 53.125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1697,
      "doc_norm": 1.5583,
      "encoder_q-embeddings": 2496.3596,
      "encoder_q-layer.0": 1666.329,
      "encoder_q-layer.1": 1819.6046,
      "encoder_q-layer.10": 2319.0237,
      "encoder_q-layer.11": 5455.604,
      "encoder_q-layer.2": 2069.7515,
      "encoder_q-layer.3": 2318.2341,
      "encoder_q-layer.4": 2474.989,
      "encoder_q-layer.5": 2584.1257,
      "encoder_q-layer.6": 3024.9583,
      "encoder_q-layer.7": 3048.3445,
      "encoder_q-layer.8": 3196.4854,
      "encoder_q-layer.9": 2606.4011,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5861,
      "inbatch_pos_score": 1.2402,
      "learning_rate": 1.4111111111111112e-05,
      "loss": 3.1697,
      "norm_diff": 0.0618,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4321.0793,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5845,
      "query_norm": 1.4964,
      "queue_k_norm": 1.5533,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0889,
      "sent_len_1": 66.9753,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2488,
      "stdk": 0.0493,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 74600
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1931,
      "doc_norm": 1.5499,
      "encoder_q-embeddings": 1985.882,
      "encoder_q-layer.0": 1307.681,
      "encoder_q-layer.1": 1392.5277,
      "encoder_q-layer.10": 2680.355,
      "encoder_q-layer.11": 5641.0732,
      "encoder_q-layer.2": 1575.7742,
      "encoder_q-layer.3": 1573.6864,
      "encoder_q-layer.4": 1702.7737,
      "encoder_q-layer.5": 1771.4996,
      "encoder_q-layer.6": 2004.7928,
      "encoder_q-layer.7": 2315.2092,
      "encoder_q-layer.8": 2916.8472,
      "encoder_q-layer.9": 2619.9878,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5833,
      "inbatch_pos_score": 1.2666,
      "learning_rate": 1.4055555555555556e-05,
      "loss": 3.1931,
      "norm_diff": 0.0333,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3866.5448,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.583,
      "query_norm": 1.5166,
      "queue_k_norm": 1.5514,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7371,
      "sent_len_1": 66.6456,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.69,
      "stdk": 0.0489,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74700
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1592,
      "doc_norm": 1.551,
      "encoder_q-embeddings": 2239.324,
      "encoder_q-layer.0": 1556.635,
      "encoder_q-layer.1": 1636.0442,
      "encoder_q-layer.10": 2552.6133,
      "encoder_q-layer.11": 5657.3203,
      "encoder_q-layer.2": 1904.9974,
      "encoder_q-layer.3": 2028.551,
      "encoder_q-layer.4": 2148.2825,
      "encoder_q-layer.5": 2350.7375,
      "encoder_q-layer.6": 2316.5293,
      "encoder_q-layer.7": 2761.5635,
      "encoder_q-layer.8": 3012.98,
      "encoder_q-layer.9": 2530.4187,
      "epoch": 0.73,
      "inbatch_neg_score": 0.588,
      "inbatch_pos_score": 1.2598,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 3.1592,
      "norm_diff": 0.0511,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4113.5701,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5884,
      "query_norm": 1.4999,
      "queue_k_norm": 1.5532,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0023,
      "sent_len_1": 66.6952,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.695,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74800
    },
    {
      "accuracy": 59.375,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1696,
      "doc_norm": 1.5549,
      "encoder_q-embeddings": 2368.4841,
      "encoder_q-layer.0": 1494.657,
      "encoder_q-layer.1": 1646.26,
      "encoder_q-layer.10": 2376.613,
      "encoder_q-layer.11": 5147.5762,
      "encoder_q-layer.2": 1847.7101,
      "encoder_q-layer.3": 1933.8927,
      "encoder_q-layer.4": 2074.6189,
      "encoder_q-layer.5": 2270.6841,
      "encoder_q-layer.6": 2390.7878,
      "encoder_q-layer.7": 2545.4421,
      "encoder_q-layer.8": 2874.4338,
      "encoder_q-layer.9": 2361.8599,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5893,
      "inbatch_pos_score": 1.3066,
      "learning_rate": 1.3944444444444446e-05,
      "loss": 3.1696,
      "norm_diff": 0.037,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3877.1805,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5894,
      "query_norm": 1.5179,
      "queue_k_norm": 1.5527,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0135,
      "sent_len_1": 66.6425,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9162,
      "stdk": 0.0491,
      "stdq": 0.0463,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 74900
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1903,
      "doc_norm": 1.5533,
      "encoder_q-embeddings": 5004.8408,
      "encoder_q-layer.0": 3453.0413,
      "encoder_q-layer.1": 4017.7893,
      "encoder_q-layer.10": 2403.5061,
      "encoder_q-layer.11": 5478.8193,
      "encoder_q-layer.2": 4905.5273,
      "encoder_q-layer.3": 4915.8271,
      "encoder_q-layer.4": 5839.5869,
      "encoder_q-layer.5": 6254.4795,
      "encoder_q-layer.6": 6969.1807,
      "encoder_q-layer.7": 6610.6826,
      "encoder_q-layer.8": 5360.8354,
      "encoder_q-layer.9": 2952.3828,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5919,
      "inbatch_pos_score": 1.2812,
      "learning_rate": 1.388888888888889e-05,
      "loss": 3.1903,
      "norm_diff": 0.04,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7687.6038,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5923,
      "query_norm": 1.5133,
      "queue_k_norm": 1.5527,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8629,
      "sent_len_1": 66.6422,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5575,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 75000
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.17,
      "doc_norm": 1.5572,
      "encoder_q-embeddings": 2147.8879,
      "encoder_q-layer.0": 1399.3508,
      "encoder_q-layer.1": 1461.99,
      "encoder_q-layer.10": 2391.553,
      "encoder_q-layer.11": 5612.5649,
      "encoder_q-layer.2": 1660.1729,
      "encoder_q-layer.3": 1776.1384,
      "encoder_q-layer.4": 1902.4763,
      "encoder_q-layer.5": 2016.4128,
      "encoder_q-layer.6": 2282.6772,
      "encoder_q-layer.7": 2511.8677,
      "encoder_q-layer.8": 2949.3242,
      "encoder_q-layer.9": 2417.5254,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5964,
      "inbatch_pos_score": 1.252,
      "learning_rate": 1.3833333333333334e-05,
      "loss": 3.17,
      "norm_diff": 0.0583,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3925.9457,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5962,
      "query_norm": 1.4989,
      "queue_k_norm": 1.5547,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0272,
      "sent_len_1": 66.7479,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6863,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75100
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1739,
      "doc_norm": 1.5618,
      "encoder_q-embeddings": 3140.9307,
      "encoder_q-layer.0": 2115.4121,
      "encoder_q-layer.1": 2236.687,
      "encoder_q-layer.10": 2631.1016,
      "encoder_q-layer.11": 5286.0659,
      "encoder_q-layer.2": 2830.2698,
      "encoder_q-layer.3": 2934.0173,
      "encoder_q-layer.4": 3073.6458,
      "encoder_q-layer.5": 2904.4629,
      "encoder_q-layer.6": 3136.1614,
      "encoder_q-layer.7": 3234.2322,
      "encoder_q-layer.8": 3343.9675,
      "encoder_q-layer.9": 2486.5,
      "epoch": 0.73,
      "inbatch_neg_score": 0.5986,
      "inbatch_pos_score": 1.292,
      "learning_rate": 1.3777777777777778e-05,
      "loss": 3.1739,
      "norm_diff": 0.0437,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4732.9826,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.5981,
      "query_norm": 1.518,
      "queue_k_norm": 1.5565,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0229,
      "sent_len_1": 66.8028,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2425,
      "stdk": 0.0493,
      "stdq": 0.0462,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75200
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1727,
      "doc_norm": 1.5585,
      "encoder_q-embeddings": 8200.0557,
      "encoder_q-layer.0": 5321.5244,
      "encoder_q-layer.1": 6126.0962,
      "encoder_q-layer.10": 4967.2534,
      "encoder_q-layer.11": 11056.999,
      "encoder_q-layer.2": 7834.8369,
      "encoder_q-layer.3": 8595.9062,
      "encoder_q-layer.4": 9220.5566,
      "encoder_q-layer.5": 9783.0049,
      "encoder_q-layer.6": 9581.6934,
      "encoder_q-layer.7": 9118.8604,
      "encoder_q-layer.8": 7900.5327,
      "encoder_q-layer.9": 5249.5459,
      "epoch": 0.74,
      "inbatch_neg_score": 0.599,
      "inbatch_pos_score": 1.2852,
      "learning_rate": 1.3722222222222222e-05,
      "loss": 3.1727,
      "norm_diff": 0.0479,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 12299.9379,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.5986,
      "query_norm": 1.5105,
      "queue_k_norm": 1.5559,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9723,
      "sent_len_1": 66.6506,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6662,
      "stdk": 0.0491,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75300
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1696,
      "doc_norm": 1.5537,
      "encoder_q-embeddings": 7141.873,
      "encoder_q-layer.0": 5027.9131,
      "encoder_q-layer.1": 6302.9639,
      "encoder_q-layer.10": 5169.4668,
      "encoder_q-layer.11": 11498.4814,
      "encoder_q-layer.2": 7170.8823,
      "encoder_q-layer.3": 7928.3135,
      "encoder_q-layer.4": 9416.3799,
      "encoder_q-layer.5": 11293.3242,
      "encoder_q-layer.6": 12675.4258,
      "encoder_q-layer.7": 11556.8008,
      "encoder_q-layer.8": 10069.3203,
      "encoder_q-layer.9": 6133.7568,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6044,
      "inbatch_pos_score": 1.2559,
      "learning_rate": 1.3666666666666666e-05,
      "loss": 3.1696,
      "norm_diff": 0.0635,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 13186.9737,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.6045,
      "query_norm": 1.4901,
      "queue_k_norm": 1.5567,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8775,
      "sent_len_1": 66.6525,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1775,
      "stdk": 0.0489,
      "stdq": 0.0449,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75400
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1824,
      "doc_norm": 1.5591,
      "encoder_q-embeddings": 2161.3477,
      "encoder_q-layer.0": 1406.6367,
      "encoder_q-layer.1": 1513.7592,
      "encoder_q-layer.10": 2438.5886,
      "encoder_q-layer.11": 5440.3457,
      "encoder_q-layer.2": 1725.8319,
      "encoder_q-layer.3": 1827.8616,
      "encoder_q-layer.4": 1977.649,
      "encoder_q-layer.5": 2059.0391,
      "encoder_q-layer.6": 2393.4873,
      "encoder_q-layer.7": 2644.2095,
      "encoder_q-layer.8": 2926.1238,
      "encoder_q-layer.9": 2574.6382,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6037,
      "inbatch_pos_score": 1.293,
      "learning_rate": 1.3611111111111111e-05,
      "loss": 3.1824,
      "norm_diff": 0.0567,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4000.8086,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6035,
      "query_norm": 1.5024,
      "queue_k_norm": 1.5565,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0105,
      "sent_len_1": 66.7862,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4888,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75500
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1771,
      "doc_norm": 1.5598,
      "encoder_q-embeddings": 1857.0653,
      "encoder_q-layer.0": 1215.1851,
      "encoder_q-layer.1": 1275.8252,
      "encoder_q-layer.10": 2572.4373,
      "encoder_q-layer.11": 5751.29,
      "encoder_q-layer.2": 1458.463,
      "encoder_q-layer.3": 1598.4443,
      "encoder_q-layer.4": 1671.8458,
      "encoder_q-layer.5": 1858.9637,
      "encoder_q-layer.6": 2148.8347,
      "encoder_q-layer.7": 2624.7588,
      "encoder_q-layer.8": 2912.4653,
      "encoder_q-layer.9": 2556.8767,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6091,
      "inbatch_pos_score": 1.291,
      "learning_rate": 1.3555555555555557e-05,
      "loss": 3.1771,
      "norm_diff": 0.0568,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3868.6348,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6089,
      "query_norm": 1.5029,
      "queue_k_norm": 1.5581,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8549,
      "sent_len_1": 66.786,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7312,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75600
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1713,
      "doc_norm": 1.5585,
      "encoder_q-embeddings": 1465.4961,
      "encoder_q-layer.0": 950.424,
      "encoder_q-layer.1": 1143.38,
      "encoder_q-layer.10": 1334.3623,
      "encoder_q-layer.11": 2896.6162,
      "encoder_q-layer.2": 1433.5925,
      "encoder_q-layer.3": 1625.8075,
      "encoder_q-layer.4": 1970.6665,
      "encoder_q-layer.5": 2094.7869,
      "encoder_q-layer.6": 2106.1877,
      "encoder_q-layer.7": 2097.6926,
      "encoder_q-layer.8": 1815.2234,
      "encoder_q-layer.9": 1356.8777,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6101,
      "inbatch_pos_score": 1.291,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 3.1713,
      "norm_diff": 0.0589,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2711.3901,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6104,
      "query_norm": 1.4996,
      "queue_k_norm": 1.5562,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8917,
      "sent_len_1": 66.6638,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3775,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 75700
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1596,
      "doc_norm": 1.5591,
      "encoder_q-embeddings": 1024.3262,
      "encoder_q-layer.0": 665.9628,
      "encoder_q-layer.1": 750.1851,
      "encoder_q-layer.10": 1229.4347,
      "encoder_q-layer.11": 2781.9602,
      "encoder_q-layer.2": 868.2911,
      "encoder_q-layer.3": 929.0332,
      "encoder_q-layer.4": 972.0706,
      "encoder_q-layer.5": 1005.922,
      "encoder_q-layer.6": 1183.7092,
      "encoder_q-layer.7": 1261.6667,
      "encoder_q-layer.8": 1404.2211,
      "encoder_q-layer.9": 1231.803,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6125,
      "inbatch_pos_score": 1.2998,
      "learning_rate": 1.3444444444444445e-05,
      "loss": 3.1596,
      "norm_diff": 0.043,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1979.1145,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6123,
      "query_norm": 1.516,
      "queue_k_norm": 1.5589,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0991,
      "sent_len_1": 66.7472,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0725,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75800
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1873,
      "doc_norm": 1.5561,
      "encoder_q-embeddings": 1263.7653,
      "encoder_q-layer.0": 836.9243,
      "encoder_q-layer.1": 946.2392,
      "encoder_q-layer.10": 1189.983,
      "encoder_q-layer.11": 2791.854,
      "encoder_q-layer.2": 1149.2076,
      "encoder_q-layer.3": 1228.8594,
      "encoder_q-layer.4": 1352.5887,
      "encoder_q-layer.5": 1363.0895,
      "encoder_q-layer.6": 1334.312,
      "encoder_q-layer.7": 1469.1436,
      "encoder_q-layer.8": 1539.1296,
      "encoder_q-layer.9": 1294.8751,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6157,
      "inbatch_pos_score": 1.2773,
      "learning_rate": 1.338888888888889e-05,
      "loss": 3.1873,
      "norm_diff": 0.0539,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2203.7267,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6152,
      "query_norm": 1.5022,
      "queue_k_norm": 1.5598,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0633,
      "sent_len_1": 66.6232,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2775,
      "stdk": 0.0488,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 75900
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1594,
      "doc_norm": 1.5561,
      "encoder_q-embeddings": 1208.7668,
      "encoder_q-layer.0": 823.3502,
      "encoder_q-layer.1": 924.2937,
      "encoder_q-layer.10": 1246.4893,
      "encoder_q-layer.11": 2757.6956,
      "encoder_q-layer.2": 1116.6431,
      "encoder_q-layer.3": 1149.9817,
      "encoder_q-layer.4": 1174.375,
      "encoder_q-layer.5": 1299.7758,
      "encoder_q-layer.6": 1390.3964,
      "encoder_q-layer.7": 1361.884,
      "encoder_q-layer.8": 1393.5601,
      "encoder_q-layer.9": 1195.6016,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6189,
      "inbatch_pos_score": 1.2842,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 3.1594,
      "norm_diff": 0.0583,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2128.0162,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6191,
      "query_norm": 1.4979,
      "queue_k_norm": 1.5617,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1851,
      "sent_len_1": 66.8962,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.95,
      "stdk": 0.0488,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76000
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1434,
      "doc_norm": 1.555,
      "encoder_q-embeddings": 979.9028,
      "encoder_q-layer.0": 615.1132,
      "encoder_q-layer.1": 628.2957,
      "encoder_q-layer.10": 1432.2999,
      "encoder_q-layer.11": 2948.2053,
      "encoder_q-layer.2": 698.894,
      "encoder_q-layer.3": 736.2956,
      "encoder_q-layer.4": 781.2615,
      "encoder_q-layer.5": 810.0349,
      "encoder_q-layer.6": 954.7894,
      "encoder_q-layer.7": 1190.7322,
      "encoder_q-layer.8": 1446.9783,
      "encoder_q-layer.9": 1310.5179,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6199,
      "inbatch_pos_score": 1.3086,
      "learning_rate": 1.3277777777777777e-05,
      "loss": 3.1434,
      "norm_diff": 0.0379,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1943.1603,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6201,
      "query_norm": 1.5171,
      "queue_k_norm": 1.5606,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9196,
      "sent_len_1": 66.8198,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.745,
      "stdk": 0.0487,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76100
    },
    {
      "accuracy": 59.7656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1844,
      "doc_norm": 1.5602,
      "encoder_q-embeddings": 1052.3665,
      "encoder_q-layer.0": 684.2986,
      "encoder_q-layer.1": 736.3815,
      "encoder_q-layer.10": 1171.1365,
      "encoder_q-layer.11": 2734.2947,
      "encoder_q-layer.2": 840.4086,
      "encoder_q-layer.3": 888.8661,
      "encoder_q-layer.4": 967.0641,
      "encoder_q-layer.5": 980.3438,
      "encoder_q-layer.6": 1131.8843,
      "encoder_q-layer.7": 1283.1742,
      "encoder_q-layer.8": 1431.8995,
      "encoder_q-layer.9": 1268.5126,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6199,
      "inbatch_pos_score": 1.3193,
      "learning_rate": 1.3222222222222221e-05,
      "loss": 3.1844,
      "norm_diff": 0.0479,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1929.2667,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6206,
      "query_norm": 1.5123,
      "queue_k_norm": 1.5616,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0538,
      "sent_len_1": 66.9808,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1375,
      "stdk": 0.0489,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76200
    },
    {
      "accuracy": 52.2461,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1664,
      "doc_norm": 1.5579,
      "encoder_q-embeddings": 1019.8066,
      "encoder_q-layer.0": 670.029,
      "encoder_q-layer.1": 742.8884,
      "encoder_q-layer.10": 1163.4054,
      "encoder_q-layer.11": 2801.9749,
      "encoder_q-layer.2": 866.127,
      "encoder_q-layer.3": 898.3571,
      "encoder_q-layer.4": 982.1168,
      "encoder_q-layer.5": 1004.651,
      "encoder_q-layer.6": 1091.8055,
      "encoder_q-layer.7": 1276.4609,
      "encoder_q-layer.8": 1358.309,
      "encoder_q-layer.9": 1179.4246,
      "epoch": 0.74,
      "inbatch_neg_score": 0.6264,
      "inbatch_pos_score": 1.2725,
      "learning_rate": 1.3166666666666665e-05,
      "loss": 3.1664,
      "norm_diff": 0.0691,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1935.3609,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.626,
      "query_norm": 1.4888,
      "queue_k_norm": 1.5625,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0156,
      "sent_len_1": 66.5257,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2488,
      "stdk": 0.0488,
      "stdq": 0.0448,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76300
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1997,
      "doc_norm": 1.5667,
      "encoder_q-embeddings": 1561.6692,
      "encoder_q-layer.0": 987.5603,
      "encoder_q-layer.1": 1127.7711,
      "encoder_q-layer.10": 1187.0997,
      "encoder_q-layer.11": 2831.792,
      "encoder_q-layer.2": 1293.6871,
      "encoder_q-layer.3": 1468.5544,
      "encoder_q-layer.4": 1614.1093,
      "encoder_q-layer.5": 1710.3893,
      "encoder_q-layer.6": 1564.3428,
      "encoder_q-layer.7": 1715.9398,
      "encoder_q-layer.8": 1661.1283,
      "encoder_q-layer.9": 1246.7933,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6287,
      "inbatch_pos_score": 1.3086,
      "learning_rate": 1.3111111111111113e-05,
      "loss": 3.1997,
      "norm_diff": 0.0563,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2467.1987,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6289,
      "query_norm": 1.5104,
      "queue_k_norm": 1.5624,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1011,
      "sent_len_1": 66.8078,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8787,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76400
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1786,
      "doc_norm": 1.5662,
      "encoder_q-embeddings": 998.1996,
      "encoder_q-layer.0": 663.5818,
      "encoder_q-layer.1": 728.3683,
      "encoder_q-layer.10": 1279.9879,
      "encoder_q-layer.11": 2878.512,
      "encoder_q-layer.2": 815.5669,
      "encoder_q-layer.3": 850.9406,
      "encoder_q-layer.4": 921.0896,
      "encoder_q-layer.5": 968.4423,
      "encoder_q-layer.6": 1082.563,
      "encoder_q-layer.7": 1295.6964,
      "encoder_q-layer.8": 1483.65,
      "encoder_q-layer.9": 1258.7258,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6364,
      "inbatch_pos_score": 1.3271,
      "learning_rate": 1.3055555555555557e-05,
      "loss": 3.1786,
      "norm_diff": 0.0398,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1982.0136,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6362,
      "query_norm": 1.5264,
      "queue_k_norm": 1.5645,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0673,
      "sent_len_1": 66.8818,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9275,
      "stdk": 0.0491,
      "stdq": 0.0462,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76500
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1451,
      "doc_norm": 1.5626,
      "encoder_q-embeddings": 2521.4482,
      "encoder_q-layer.0": 1716.4009,
      "encoder_q-layer.1": 2008.2189,
      "encoder_q-layer.10": 1225.6606,
      "encoder_q-layer.11": 2662.0867,
      "encoder_q-layer.2": 2477.1814,
      "encoder_q-layer.3": 2657.448,
      "encoder_q-layer.4": 2895.3589,
      "encoder_q-layer.5": 2873.5422,
      "encoder_q-layer.6": 3298.5574,
      "encoder_q-layer.7": 3642.1726,
      "encoder_q-layer.8": 3657.6707,
      "encoder_q-layer.9": 2391.9124,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6358,
      "inbatch_pos_score": 1.3457,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 3.1451,
      "norm_diff": 0.0457,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4140.6427,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6353,
      "query_norm": 1.5169,
      "queue_k_norm": 1.5653,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.932,
      "sent_len_1": 66.8922,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.945,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76600
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.182,
      "doc_norm": 1.5658,
      "encoder_q-embeddings": 1079.988,
      "encoder_q-layer.0": 761.9317,
      "encoder_q-layer.1": 779.9326,
      "encoder_q-layer.10": 1307.7113,
      "encoder_q-layer.11": 2982.3381,
      "encoder_q-layer.2": 864.5114,
      "encoder_q-layer.3": 897.7054,
      "encoder_q-layer.4": 976.7544,
      "encoder_q-layer.5": 1018.8286,
      "encoder_q-layer.6": 1080.2323,
      "encoder_q-layer.7": 1252.153,
      "encoder_q-layer.8": 1494.0441,
      "encoder_q-layer.9": 1280.1099,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6409,
      "inbatch_pos_score": 1.292,
      "learning_rate": 1.2944444444444445e-05,
      "loss": 3.182,
      "norm_diff": 0.0659,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2040.1818,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6416,
      "query_norm": 1.4998,
      "queue_k_norm": 1.5668,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0921,
      "sent_len_1": 66.8219,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4462,
      "stdk": 0.049,
      "stdq": 0.045,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76700
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1544,
      "doc_norm": 1.568,
      "encoder_q-embeddings": 953.4173,
      "encoder_q-layer.0": 614.189,
      "encoder_q-layer.1": 637.1108,
      "encoder_q-layer.10": 1246.2124,
      "encoder_q-layer.11": 2731.4817,
      "encoder_q-layer.2": 697.0402,
      "encoder_q-layer.3": 702.1325,
      "encoder_q-layer.4": 774.4581,
      "encoder_q-layer.5": 809.3976,
      "encoder_q-layer.6": 959.8362,
      "encoder_q-layer.7": 1175.7621,
      "encoder_q-layer.8": 1374.2388,
      "encoder_q-layer.9": 1276.4059,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6431,
      "inbatch_pos_score": 1.332,
      "learning_rate": 1.2888888888888889e-05,
      "loss": 3.1544,
      "norm_diff": 0.0425,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1837.9876,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6426,
      "query_norm": 1.5255,
      "queue_k_norm": 1.5656,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7801,
      "sent_len_1": 66.5179,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.53,
      "stdk": 0.049,
      "stdq": 0.0462,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76800
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1514,
      "doc_norm": 1.5722,
      "encoder_q-embeddings": 920.1546,
      "encoder_q-layer.0": 636.2035,
      "encoder_q-layer.1": 671.1888,
      "encoder_q-layer.10": 1251.2987,
      "encoder_q-layer.11": 2801.0696,
      "encoder_q-layer.2": 750.7462,
      "encoder_q-layer.3": 800.4999,
      "encoder_q-layer.4": 864.861,
      "encoder_q-layer.5": 939.7134,
      "encoder_q-layer.6": 1051.2018,
      "encoder_q-layer.7": 1208.6002,
      "encoder_q-layer.8": 1386.0709,
      "encoder_q-layer.9": 1232.5308,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6418,
      "inbatch_pos_score": 1.3076,
      "learning_rate": 1.2833333333333333e-05,
      "loss": 3.1514,
      "norm_diff": 0.0731,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1926.1918,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6411,
      "query_norm": 1.4991,
      "queue_k_norm": 1.5683,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0248,
      "sent_len_1": 66.8433,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9737,
      "stdk": 0.0492,
      "stdq": 0.045,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 76900
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1544,
      "doc_norm": 1.5702,
      "encoder_q-embeddings": 978.2535,
      "encoder_q-layer.0": 665.0071,
      "encoder_q-layer.1": 728.5564,
      "encoder_q-layer.10": 1245.0553,
      "encoder_q-layer.11": 2656.2292,
      "encoder_q-layer.2": 824.3855,
      "encoder_q-layer.3": 902.9703,
      "encoder_q-layer.4": 990.199,
      "encoder_q-layer.5": 1053.0598,
      "encoder_q-layer.6": 1177.7483,
      "encoder_q-layer.7": 1301.2131,
      "encoder_q-layer.8": 1324.5337,
      "encoder_q-layer.9": 1125.8601,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6434,
      "inbatch_pos_score": 1.3467,
      "learning_rate": 1.2777777777777777e-05,
      "loss": 3.1544,
      "norm_diff": 0.0566,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1942.416,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6431,
      "query_norm": 1.5136,
      "queue_k_norm": 1.569,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8247,
      "sent_len_1": 66.7542,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4575,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77000
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.194,
      "doc_norm": 1.5681,
      "encoder_q-embeddings": 1028.9396,
      "encoder_q-layer.0": 659.2719,
      "encoder_q-layer.1": 710.2027,
      "encoder_q-layer.10": 1280.4664,
      "encoder_q-layer.11": 2707.8152,
      "encoder_q-layer.2": 815.1221,
      "encoder_q-layer.3": 912.9161,
      "encoder_q-layer.4": 974.3571,
      "encoder_q-layer.5": 1008.4022,
      "encoder_q-layer.6": 1139.8645,
      "encoder_q-layer.7": 1377.3804,
      "encoder_q-layer.8": 1621.3558,
      "encoder_q-layer.9": 1290.194,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6412,
      "inbatch_pos_score": 1.3584,
      "learning_rate": 1.2722222222222221e-05,
      "loss": 3.194,
      "norm_diff": 0.0501,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1999.3117,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6426,
      "query_norm": 1.518,
      "queue_k_norm": 1.57,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0393,
      "sent_len_1": 66.7324,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.13,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77100
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1651,
      "doc_norm": 1.5754,
      "encoder_q-embeddings": 1763.7333,
      "encoder_q-layer.0": 1290.5435,
      "encoder_q-layer.1": 1551.7937,
      "encoder_q-layer.10": 1191.5386,
      "encoder_q-layer.11": 2762.1438,
      "encoder_q-layer.2": 1937.2554,
      "encoder_q-layer.3": 1981.0013,
      "encoder_q-layer.4": 2084.343,
      "encoder_q-layer.5": 2057.5503,
      "encoder_q-layer.6": 1967.8538,
      "encoder_q-layer.7": 1950.6313,
      "encoder_q-layer.8": 1891.0977,
      "encoder_q-layer.9": 1305.879,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6443,
      "inbatch_pos_score": 1.3467,
      "learning_rate": 1.2666666666666668e-05,
      "loss": 3.1651,
      "norm_diff": 0.0566,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2853.2264,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.645,
      "query_norm": 1.5188,
      "queue_k_norm": 1.5683,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.968,
      "sent_len_1": 66.6408,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2887,
      "stdk": 0.0492,
      "stdq": 0.0458,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 77200
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1585,
      "doc_norm": 1.5727,
      "encoder_q-embeddings": 951.5279,
      "encoder_q-layer.0": 643.2775,
      "encoder_q-layer.1": 674.7125,
      "encoder_q-layer.10": 1274.2588,
      "encoder_q-layer.11": 2827.6821,
      "encoder_q-layer.2": 748.3366,
      "encoder_q-layer.3": 770.0674,
      "encoder_q-layer.4": 816.3086,
      "encoder_q-layer.5": 872.8497,
      "encoder_q-layer.6": 968.0383,
      "encoder_q-layer.7": 1153.6809,
      "encoder_q-layer.8": 1302.3416,
      "encoder_q-layer.9": 1193.8947,
      "epoch": 0.75,
      "inbatch_neg_score": 0.6454,
      "inbatch_pos_score": 1.3096,
      "learning_rate": 1.2611111111111113e-05,
      "loss": 3.1585,
      "norm_diff": 0.0675,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1901.5899,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.645,
      "query_norm": 1.5052,
      "queue_k_norm": 1.5714,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9241,
      "sent_len_1": 66.8445,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4875,
      "stdk": 0.0491,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 77300
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1616,
      "doc_norm": 1.564,
      "encoder_q-embeddings": 1947.1647,
      "encoder_q-layer.0": 1396.7216,
      "encoder_q-layer.1": 1495.7775,
      "encoder_q-layer.10": 1378.4695,
      "encoder_q-layer.11": 2915.2759,
      "encoder_q-layer.2": 1728.9567,
      "encoder_q-layer.3": 1682.5164,
      "encoder_q-layer.4": 1681.6162,
      "encoder_q-layer.5": 1637.8372,
      "encoder_q-layer.6": 1658.2618,
      "encoder_q-layer.7": 1751.0413,
      "encoder_q-layer.8": 1850.1465,
      "encoder_q-layer.9": 1401.1697,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6453,
      "inbatch_pos_score": 1.3232,
      "learning_rate": 1.2555555555555557e-05,
      "loss": 3.1616,
      "norm_diff": 0.0534,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2706.5116,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6465,
      "query_norm": 1.5107,
      "queue_k_norm": 1.5711,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9462,
      "sent_len_1": 66.7849,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2438,
      "stdk": 0.0487,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77400
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1303,
      "doc_norm": 1.5744,
      "encoder_q-embeddings": 992.5045,
      "encoder_q-layer.0": 648.1384,
      "encoder_q-layer.1": 705.2957,
      "encoder_q-layer.10": 1253.2083,
      "encoder_q-layer.11": 2636.2209,
      "encoder_q-layer.2": 803.4568,
      "encoder_q-layer.3": 829.6835,
      "encoder_q-layer.4": 897.5078,
      "encoder_q-layer.5": 966.9257,
      "encoder_q-layer.6": 1074.9622,
      "encoder_q-layer.7": 1197.9172,
      "encoder_q-layer.8": 1352.994,
      "encoder_q-layer.9": 1171.5554,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6474,
      "inbatch_pos_score": 1.333,
      "learning_rate": 1.25e-05,
      "loss": 3.1303,
      "norm_diff": 0.0556,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1861.8785,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6475,
      "query_norm": 1.5188,
      "queue_k_norm": 1.5702,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8754,
      "sent_len_1": 66.8516,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.355,
      "stdk": 0.0491,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77500
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1622,
      "doc_norm": 1.5727,
      "encoder_q-embeddings": 1128.3977,
      "encoder_q-layer.0": 743.1681,
      "encoder_q-layer.1": 820.5738,
      "encoder_q-layer.10": 1285.3599,
      "encoder_q-layer.11": 2791.366,
      "encoder_q-layer.2": 902.5099,
      "encoder_q-layer.3": 966.1047,
      "encoder_q-layer.4": 1054.7731,
      "encoder_q-layer.5": 1127.0139,
      "encoder_q-layer.6": 1197.5981,
      "encoder_q-layer.7": 1330.8566,
      "encoder_q-layer.8": 1467.4712,
      "encoder_q-layer.9": 1211.824,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6466,
      "inbatch_pos_score": 1.3359,
      "learning_rate": 1.2444444444444445e-05,
      "loss": 3.1622,
      "norm_diff": 0.0543,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2042.7284,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6475,
      "query_norm": 1.5184,
      "queue_k_norm": 1.573,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.86,
      "sent_len_1": 67.0601,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2537,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77600
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1993,
      "doc_norm": 1.5724,
      "encoder_q-embeddings": 2983.4451,
      "encoder_q-layer.0": 1969.2716,
      "encoder_q-layer.1": 2097.3706,
      "encoder_q-layer.10": 2454.803,
      "encoder_q-layer.11": 5525.9941,
      "encoder_q-layer.2": 2427.1655,
      "encoder_q-layer.3": 2479.3674,
      "encoder_q-layer.4": 2591.0503,
      "encoder_q-layer.5": 2572.616,
      "encoder_q-layer.6": 2590.6553,
      "encoder_q-layer.7": 2685.2559,
      "encoder_q-layer.8": 2825.4033,
      "encoder_q-layer.9": 2398.1396,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6481,
      "inbatch_pos_score": 1.3164,
      "learning_rate": 1.238888888888889e-05,
      "loss": 3.1993,
      "norm_diff": 0.0661,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4443.6066,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6489,
      "query_norm": 1.5064,
      "queue_k_norm": 1.571,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1373,
      "sent_len_1": 66.6674,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7463,
      "stdk": 0.049,
      "stdq": 0.0453,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 77700
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1545,
      "doc_norm": 1.5742,
      "encoder_q-embeddings": 2905.5227,
      "encoder_q-layer.0": 1858.0099,
      "encoder_q-layer.1": 2136.9873,
      "encoder_q-layer.10": 2727.5796,
      "encoder_q-layer.11": 5765.6577,
      "encoder_q-layer.2": 2453.8853,
      "encoder_q-layer.3": 2376.6465,
      "encoder_q-layer.4": 2489.5022,
      "encoder_q-layer.5": 2440.0747,
      "encoder_q-layer.6": 2643.6157,
      "encoder_q-layer.7": 3013.2292,
      "encoder_q-layer.8": 3129.5364,
      "encoder_q-layer.9": 2597.2,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6512,
      "inbatch_pos_score": 1.332,
      "learning_rate": 1.2333333333333334e-05,
      "loss": 3.1545,
      "norm_diff": 0.0483,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4546.5951,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6514,
      "query_norm": 1.5258,
      "queue_k_norm": 1.5733,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0055,
      "sent_len_1": 66.6709,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1525,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77800
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1658,
      "doc_norm": 1.5761,
      "encoder_q-embeddings": 2336.7141,
      "encoder_q-layer.0": 1639.288,
      "encoder_q-layer.1": 1768.2217,
      "encoder_q-layer.10": 2591.6689,
      "encoder_q-layer.11": 5784.9277,
      "encoder_q-layer.2": 2088.874,
      "encoder_q-layer.3": 2149.7478,
      "encoder_q-layer.4": 2287.0979,
      "encoder_q-layer.5": 2361.519,
      "encoder_q-layer.6": 2598.3044,
      "encoder_q-layer.7": 2906.2161,
      "encoder_q-layer.8": 3163.4297,
      "encoder_q-layer.9": 2445.865,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6536,
      "inbatch_pos_score": 1.3203,
      "learning_rate": 1.2277777777777778e-05,
      "loss": 3.1658,
      "norm_diff": 0.0626,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4278.7935,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6528,
      "query_norm": 1.5135,
      "queue_k_norm": 1.5744,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0081,
      "sent_len_1": 66.628,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7463,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 77900
    },
    {
      "accuracy": 53.7109,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1611,
      "doc_norm": 1.572,
      "encoder_q-embeddings": 2287.3555,
      "encoder_q-layer.0": 1506.9319,
      "encoder_q-layer.1": 1710.493,
      "encoder_q-layer.10": 2425.9717,
      "encoder_q-layer.11": 5566.875,
      "encoder_q-layer.2": 2039.054,
      "encoder_q-layer.3": 2100.9792,
      "encoder_q-layer.4": 2250.8562,
      "encoder_q-layer.5": 2275.2554,
      "encoder_q-layer.6": 2489.3774,
      "encoder_q-layer.7": 2685.1077,
      "encoder_q-layer.8": 2994.9541,
      "encoder_q-layer.9": 2495.335,
      "epoch": 0.76,
      "inbatch_neg_score": 0.653,
      "inbatch_pos_score": 1.3252,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 3.1611,
      "norm_diff": 0.0537,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4168.7797,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6528,
      "query_norm": 1.5184,
      "queue_k_norm": 1.5759,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0234,
      "sent_len_1": 66.8237,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4338,
      "stdk": 0.0489,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 78000
    },
    {
      "accuracy": 56.9336,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1382,
      "doc_norm": 1.5787,
      "encoder_q-embeddings": 2228.3474,
      "encoder_q-layer.0": 1515.8003,
      "encoder_q-layer.1": 1756.9882,
      "encoder_q-layer.10": 2447.3245,
      "encoder_q-layer.11": 5217.5049,
      "encoder_q-layer.2": 2045.1913,
      "encoder_q-layer.3": 2120.0552,
      "encoder_q-layer.4": 2320.0547,
      "encoder_q-layer.5": 2623.645,
      "encoder_q-layer.6": 3019.3906,
      "encoder_q-layer.7": 3560.823,
      "encoder_q-layer.8": 3676.324,
      "encoder_q-layer.9": 2557.2544,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6552,
      "inbatch_pos_score": 1.3516,
      "learning_rate": 1.2166666666666668e-05,
      "loss": 3.1382,
      "norm_diff": 0.0565,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4311.9979,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6538,
      "query_norm": 1.5222,
      "queue_k_norm": 1.5757,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9969,
      "sent_len_1": 66.7135,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6275,
      "stdk": 0.0491,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78100
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1851,
      "doc_norm": 1.5771,
      "encoder_q-embeddings": 2001.933,
      "encoder_q-layer.0": 1379.3608,
      "encoder_q-layer.1": 1509.239,
      "encoder_q-layer.10": 2569.6267,
      "encoder_q-layer.11": 5762.7783,
      "encoder_q-layer.2": 1702.9415,
      "encoder_q-layer.3": 1790.7684,
      "encoder_q-layer.4": 2021.3563,
      "encoder_q-layer.5": 2070.2349,
      "encoder_q-layer.6": 2319.6672,
      "encoder_q-layer.7": 2593.9941,
      "encoder_q-layer.8": 2971.6345,
      "encoder_q-layer.9": 2667.7646,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6507,
      "inbatch_pos_score": 1.3281,
      "learning_rate": 1.2111111111111112e-05,
      "loss": 3.1851,
      "norm_diff": 0.0628,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4009.6781,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6504,
      "query_norm": 1.5143,
      "queue_k_norm": 1.5753,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9702,
      "sent_len_1": 66.7414,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5325,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78200
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1756,
      "doc_norm": 1.5745,
      "encoder_q-embeddings": 2377.8452,
      "encoder_q-layer.0": 1617.1125,
      "encoder_q-layer.1": 1780.6348,
      "encoder_q-layer.10": 2662.5571,
      "encoder_q-layer.11": 6093.249,
      "encoder_q-layer.2": 2113.3198,
      "encoder_q-layer.3": 2149.0605,
      "encoder_q-layer.4": 2332.6001,
      "encoder_q-layer.5": 2528.7229,
      "encoder_q-layer.6": 2742.5955,
      "encoder_q-layer.7": 3055.9192,
      "encoder_q-layer.8": 3592.4653,
      "encoder_q-layer.9": 2853.9546,
      "epoch": 0.76,
      "inbatch_neg_score": 0.6531,
      "inbatch_pos_score": 1.3145,
      "learning_rate": 1.2055555555555556e-05,
      "loss": 3.1756,
      "norm_diff": 0.061,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4510.3225,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6523,
      "query_norm": 1.5134,
      "queue_k_norm": 1.5751,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2194,
      "sent_len_1": 66.7451,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5475,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78300
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1533,
      "doc_norm": 1.5779,
      "encoder_q-embeddings": 2259.5396,
      "encoder_q-layer.0": 1844.7709,
      "encoder_q-layer.1": 1958.3999,
      "encoder_q-layer.10": 1224.0638,
      "encoder_q-layer.11": 2880.1111,
      "encoder_q-layer.2": 2337.1819,
      "encoder_q-layer.3": 2105.0728,
      "encoder_q-layer.4": 2572.5725,
      "encoder_q-layer.5": 2348.8123,
      "encoder_q-layer.6": 1711.4771,
      "encoder_q-layer.7": 1675.2809,
      "encoder_q-layer.8": 1565.0474,
      "encoder_q-layer.9": 1215.0515,
      "epoch": 0.77,
      "inbatch_neg_score": 0.653,
      "inbatch_pos_score": 1.3203,
      "learning_rate": 1.2e-05,
      "loss": 3.1533,
      "norm_diff": 0.0733,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3107.1675,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6538,
      "query_norm": 1.5047,
      "queue_k_norm": 1.5767,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9691,
      "sent_len_1": 66.774,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6188,
      "stdk": 0.0491,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78400
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.153,
      "doc_norm": 1.5662,
      "encoder_q-embeddings": 2033.8506,
      "encoder_q-layer.0": 1383.6649,
      "encoder_q-layer.1": 1498.9988,
      "encoder_q-layer.10": 1268.4584,
      "encoder_q-layer.11": 2806.7173,
      "encoder_q-layer.2": 1806.3229,
      "encoder_q-layer.3": 1978.803,
      "encoder_q-layer.4": 2223.6487,
      "encoder_q-layer.5": 2357.6406,
      "encoder_q-layer.6": 2209.5381,
      "encoder_q-layer.7": 2322.3784,
      "encoder_q-layer.8": 1981.8619,
      "encoder_q-layer.9": 1349.6571,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6534,
      "inbatch_pos_score": 1.3242,
      "learning_rate": 1.1944444444444446e-05,
      "loss": 3.153,
      "norm_diff": 0.0571,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2993.7936,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6533,
      "query_norm": 1.5091,
      "queue_k_norm": 1.5771,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.962,
      "sent_len_1": 66.9877,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5925,
      "stdk": 0.0486,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78500
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1786,
      "doc_norm": 1.5793,
      "encoder_q-embeddings": 1125.6294,
      "encoder_q-layer.0": 758.6716,
      "encoder_q-layer.1": 776.98,
      "encoder_q-layer.10": 1252.7545,
      "encoder_q-layer.11": 2922.0474,
      "encoder_q-layer.2": 849.4283,
      "encoder_q-layer.3": 885.3347,
      "encoder_q-layer.4": 953.5569,
      "encoder_q-layer.5": 952.3735,
      "encoder_q-layer.6": 1081.7649,
      "encoder_q-layer.7": 1230.0795,
      "encoder_q-layer.8": 1405.9008,
      "encoder_q-layer.9": 1227.3896,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6531,
      "inbatch_pos_score": 1.2988,
      "learning_rate": 1.188888888888889e-05,
      "loss": 3.1786,
      "norm_diff": 0.0735,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2029.0651,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6533,
      "query_norm": 1.5058,
      "queue_k_norm": 1.5775,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9991,
      "sent_len_1": 66.9956,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.595,
      "stdk": 0.0491,
      "stdq": 0.0453,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78600
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1424,
      "doc_norm": 1.5825,
      "encoder_q-embeddings": 962.7748,
      "encoder_q-layer.0": 634.7428,
      "encoder_q-layer.1": 682.8058,
      "encoder_q-layer.10": 1264.923,
      "encoder_q-layer.11": 2713.3135,
      "encoder_q-layer.2": 794.4268,
      "encoder_q-layer.3": 802.4898,
      "encoder_q-layer.4": 876.6925,
      "encoder_q-layer.5": 974.6038,
      "encoder_q-layer.6": 1051.3634,
      "encoder_q-layer.7": 1261.2944,
      "encoder_q-layer.8": 1409.4144,
      "encoder_q-layer.9": 1226.7225,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6533,
      "inbatch_pos_score": 1.3506,
      "learning_rate": 1.1833333333333334e-05,
      "loss": 3.1424,
      "norm_diff": 0.0634,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1923.3459,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6543,
      "query_norm": 1.5192,
      "queue_k_norm": 1.5764,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2499,
      "sent_len_1": 66.8894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3738,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78700
    },
    {
      "accuracy": 57.9102,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1652,
      "doc_norm": 1.5769,
      "encoder_q-embeddings": 972.7644,
      "encoder_q-layer.0": 629.7687,
      "encoder_q-layer.1": 680.2618,
      "encoder_q-layer.10": 1366.9229,
      "encoder_q-layer.11": 2709.4756,
      "encoder_q-layer.2": 765.659,
      "encoder_q-layer.3": 798.9073,
      "encoder_q-layer.4": 843.3757,
      "encoder_q-layer.5": 895.2687,
      "encoder_q-layer.6": 971.7783,
      "encoder_q-layer.7": 1137.8065,
      "encoder_q-layer.8": 1402.8965,
      "encoder_q-layer.9": 1221.605,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6527,
      "inbatch_pos_score": 1.3555,
      "learning_rate": 1.1777777777777778e-05,
      "loss": 3.1652,
      "norm_diff": 0.0548,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1880.6348,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6533,
      "query_norm": 1.5221,
      "queue_k_norm": 1.5782,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9389,
      "sent_len_1": 66.7439,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.14,
      "stdk": 0.049,
      "stdq": 0.046,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78800
    },
    {
      "accuracy": 55.1758,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1705,
      "doc_norm": 1.5787,
      "encoder_q-embeddings": 1423.2039,
      "encoder_q-layer.0": 936.6588,
      "encoder_q-layer.1": 994.168,
      "encoder_q-layer.10": 1256.6857,
      "encoder_q-layer.11": 2751.1064,
      "encoder_q-layer.2": 1168.0264,
      "encoder_q-layer.3": 1271.1741,
      "encoder_q-layer.4": 1406.915,
      "encoder_q-layer.5": 1445.5547,
      "encoder_q-layer.6": 1475.8647,
      "encoder_q-layer.7": 1676.7598,
      "encoder_q-layer.8": 1656.8119,
      "encoder_q-layer.9": 1295.1346,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6532,
      "inbatch_pos_score": 1.3154,
      "learning_rate": 1.1722222222222224e-05,
      "loss": 3.1705,
      "norm_diff": 0.0744,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2291.6989,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6533,
      "query_norm": 1.5043,
      "queue_k_norm": 1.5777,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8341,
      "sent_len_1": 66.5852,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9787,
      "stdk": 0.0491,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 78900
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1668,
      "doc_norm": 1.5835,
      "encoder_q-embeddings": 952.2517,
      "encoder_q-layer.0": 617.5647,
      "encoder_q-layer.1": 671.2285,
      "encoder_q-layer.10": 1265.5879,
      "encoder_q-layer.11": 2834.821,
      "encoder_q-layer.2": 754.6312,
      "encoder_q-layer.3": 788.4008,
      "encoder_q-layer.4": 863.3146,
      "encoder_q-layer.5": 947.9061,
      "encoder_q-layer.6": 1006.546,
      "encoder_q-layer.7": 1192.5332,
      "encoder_q-layer.8": 1387.7178,
      "encoder_q-layer.9": 1192.4296,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6547,
      "inbatch_pos_score": 1.3516,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 3.1668,
      "norm_diff": 0.0652,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1902.4495,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6543,
      "query_norm": 1.5183,
      "queue_k_norm": 1.578,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.028,
      "sent_len_1": 66.5997,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.2125,
      "stdk": 0.0492,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 79000
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1574,
      "doc_norm": 1.576,
      "encoder_q-embeddings": 975.2648,
      "encoder_q-layer.0": 650.8516,
      "encoder_q-layer.1": 697.6703,
      "encoder_q-layer.10": 1216.4587,
      "encoder_q-layer.11": 2697.6394,
      "encoder_q-layer.2": 810.0733,
      "encoder_q-layer.3": 823.8498,
      "encoder_q-layer.4": 868.7519,
      "encoder_q-layer.5": 927.0549,
      "encoder_q-layer.6": 1078.176,
      "encoder_q-layer.7": 1149.4763,
      "encoder_q-layer.8": 1398.3998,
      "encoder_q-layer.9": 1173.3846,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6583,
      "inbatch_pos_score": 1.3496,
      "learning_rate": 1.1611111111111112e-05,
      "loss": 3.1574,
      "norm_diff": 0.0571,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1881.8266,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6577,
      "query_norm": 1.5189,
      "queue_k_norm": 1.5791,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7122,
      "sent_len_1": 66.8566,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.8938,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 79100
    },
    {
      "accuracy": 59.7656,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1579,
      "doc_norm": 1.5784,
      "encoder_q-embeddings": 1129.6061,
      "encoder_q-layer.0": 744.3054,
      "encoder_q-layer.1": 832.7578,
      "encoder_q-layer.10": 1200.6946,
      "encoder_q-layer.11": 2716.4832,
      "encoder_q-layer.2": 924.7032,
      "encoder_q-layer.3": 1016.14,
      "encoder_q-layer.4": 1055.583,
      "encoder_q-layer.5": 1099.9939,
      "encoder_q-layer.6": 1220.917,
      "encoder_q-layer.7": 1373.1429,
      "encoder_q-layer.8": 1482.0388,
      "encoder_q-layer.9": 1173.1982,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6608,
      "inbatch_pos_score": 1.3721,
      "learning_rate": 1.1555555555555556e-05,
      "loss": 3.1579,
      "norm_diff": 0.0549,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2008.2147,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6597,
      "query_norm": 1.5235,
      "queue_k_norm": 1.5801,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0631,
      "sent_len_1": 66.9734,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7862,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 79200
    },
    {
      "accuracy": 58.9844,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1681,
      "doc_norm": 1.583,
      "encoder_q-embeddings": 1155.6843,
      "encoder_q-layer.0": 798.1595,
      "encoder_q-layer.1": 932.6233,
      "encoder_q-layer.10": 1117.0315,
      "encoder_q-layer.11": 2581.0459,
      "encoder_q-layer.2": 1136.2098,
      "encoder_q-layer.3": 1279.2827,
      "encoder_q-layer.4": 1351.9166,
      "encoder_q-layer.5": 1292.7146,
      "encoder_q-layer.6": 1325.3082,
      "encoder_q-layer.7": 1462.5278,
      "encoder_q-layer.8": 1590.1484,
      "encoder_q-layer.9": 1191.401,
      "epoch": 0.77,
      "inbatch_neg_score": 0.6597,
      "inbatch_pos_score": 1.3838,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 3.1681,
      "norm_diff": 0.0473,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2113.711,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6597,
      "query_norm": 1.5357,
      "queue_k_norm": 1.5788,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9448,
      "sent_len_1": 66.834,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4025,
      "stdk": 0.0492,
      "stdq": 0.0465,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 79300
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1408,
      "doc_norm": 1.5853,
      "encoder_q-embeddings": 955.3694,
      "encoder_q-layer.0": 638.131,
      "encoder_q-layer.1": 680.5255,
      "encoder_q-layer.10": 1324.7545,
      "encoder_q-layer.11": 2815.7415,
      "encoder_q-layer.2": 744.665,
      "encoder_q-layer.3": 763.1681,
      "encoder_q-layer.4": 801.1353,
      "encoder_q-layer.5": 854.6644,
      "encoder_q-layer.6": 968.6303,
      "encoder_q-layer.7": 1143.1102,
      "encoder_q-layer.8": 1345.4923,
      "encoder_q-layer.9": 1256.8074,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6637,
      "inbatch_pos_score": 1.3809,
      "learning_rate": 1.1444444444444446e-05,
      "loss": 3.1408,
      "norm_diff": 0.0366,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1860.985,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6641,
      "query_norm": 1.5488,
      "queue_k_norm": 1.5795,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7901,
      "sent_len_1": 66.5234,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2887,
      "stdk": 0.0493,
      "stdq": 0.047,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 79400
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1659,
      "doc_norm": 1.579,
      "encoder_q-embeddings": 1047.1289,
      "encoder_q-layer.0": 713.1416,
      "encoder_q-layer.1": 779.677,
      "encoder_q-layer.10": 1286.0146,
      "encoder_q-layer.11": 2886.283,
      "encoder_q-layer.2": 852.8483,
      "encoder_q-layer.3": 903.6511,
      "encoder_q-layer.4": 987.2844,
      "encoder_q-layer.5": 1077.0369,
      "encoder_q-layer.6": 1181.5812,
      "encoder_q-layer.7": 1298.2281,
      "encoder_q-layer.8": 1417.9137,
      "encoder_q-layer.9": 1245.6595,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6644,
      "inbatch_pos_score": 1.3604,
      "learning_rate": 1.138888888888889e-05,
      "loss": 3.1659,
      "norm_diff": 0.0445,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2022.3239,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6646,
      "query_norm": 1.5345,
      "queue_k_norm": 1.5801,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7916,
      "sent_len_1": 66.9398,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5775,
      "stdk": 0.049,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 79500
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1724,
      "doc_norm": 1.5827,
      "encoder_q-embeddings": 2171.8479,
      "encoder_q-layer.0": 1513.0347,
      "encoder_q-layer.1": 1585.1149,
      "encoder_q-layer.10": 1322.5426,
      "encoder_q-layer.11": 2852.1719,
      "encoder_q-layer.2": 1895.6962,
      "encoder_q-layer.3": 2071.7053,
      "encoder_q-layer.4": 2267.1003,
      "encoder_q-layer.5": 2159.1406,
      "encoder_q-layer.6": 2504.1887,
      "encoder_q-layer.7": 2368.9309,
      "encoder_q-layer.8": 2252.1953,
      "encoder_q-layer.9": 1335.7159,
      "epoch": 0.78,
      "inbatch_neg_score": 0.665,
      "inbatch_pos_score": 1.3525,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 3.1724,
      "norm_diff": 0.0544,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3080.1219,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.665,
      "query_norm": 1.5283,
      "queue_k_norm": 1.5806,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.024,
      "sent_len_1": 66.8762,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2988,
      "stdk": 0.0491,
      "stdq": 0.0461,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 79600
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1303,
      "doc_norm": 1.5842,
      "encoder_q-embeddings": 2251.7649,
      "encoder_q-layer.0": 1578.4786,
      "encoder_q-layer.1": 1832.4518,
      "encoder_q-layer.10": 1334.3176,
      "encoder_q-layer.11": 2831.446,
      "encoder_q-layer.2": 2236.5737,
      "encoder_q-layer.3": 2564.0015,
      "encoder_q-layer.4": 2474.7388,
      "encoder_q-layer.5": 2923.1765,
      "encoder_q-layer.6": 2639.1714,
      "encoder_q-layer.7": 3012.0156,
      "encoder_q-layer.8": 2562.2314,
      "encoder_q-layer.9": 1386.5248,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6694,
      "inbatch_pos_score": 1.3682,
      "learning_rate": 1.127777777777778e-05,
      "loss": 3.1303,
      "norm_diff": 0.0616,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3572.2336,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6694,
      "query_norm": 1.5225,
      "queue_k_norm": 1.5802,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8234,
      "sent_len_1": 66.79,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4762,
      "stdk": 0.0492,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 79700
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1697,
      "doc_norm": 1.5786,
      "encoder_q-embeddings": 1025.8494,
      "encoder_q-layer.0": 679.7411,
      "encoder_q-layer.1": 726.0505,
      "encoder_q-layer.10": 1528.2931,
      "encoder_q-layer.11": 3024.332,
      "encoder_q-layer.2": 831.5852,
      "encoder_q-layer.3": 886.8042,
      "encoder_q-layer.4": 937.981,
      "encoder_q-layer.5": 1013.5902,
      "encoder_q-layer.6": 1133.1135,
      "encoder_q-layer.7": 1288.5684,
      "encoder_q-layer.8": 1514.683,
      "encoder_q-layer.9": 1359.4454,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6724,
      "inbatch_pos_score": 1.3408,
      "learning_rate": 1.1222222222222224e-05,
      "loss": 3.1697,
      "norm_diff": 0.062,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2060.977,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6719,
      "query_norm": 1.5166,
      "queue_k_norm": 1.5816,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.905,
      "sent_len_1": 66.9644,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1175,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 79800
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1798,
      "doc_norm": 1.5762,
      "encoder_q-embeddings": 1015.8392,
      "encoder_q-layer.0": 675.3879,
      "encoder_q-layer.1": 712.3515,
      "encoder_q-layer.10": 1249.215,
      "encoder_q-layer.11": 2937.3428,
      "encoder_q-layer.2": 807.3736,
      "encoder_q-layer.3": 838.2741,
      "encoder_q-layer.4": 902.4457,
      "encoder_q-layer.5": 944.3777,
      "encoder_q-layer.6": 1083.7146,
      "encoder_q-layer.7": 1245.6196,
      "encoder_q-layer.8": 1426.897,
      "encoder_q-layer.9": 1256.4763,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6717,
      "inbatch_pos_score": 1.3184,
      "learning_rate": 1.1166666666666668e-05,
      "loss": 3.1798,
      "norm_diff": 0.0754,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2018.2152,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6719,
      "query_norm": 1.5009,
      "queue_k_norm": 1.5791,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7898,
      "sent_len_1": 66.5531,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2925,
      "stdk": 0.0488,
      "stdq": 0.0449,
      "stdqueue_k": 0.049,
      "stdqueue_q": 0.0,
      "step": 79900
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1805,
      "doc_norm": 1.5786,
      "encoder_q-embeddings": 1144.4414,
      "encoder_q-layer.0": 808.533,
      "encoder_q-layer.1": 873.9615,
      "encoder_q-layer.10": 1305.3239,
      "encoder_q-layer.11": 2968.7556,
      "encoder_q-layer.2": 1039.2085,
      "encoder_q-layer.3": 1029.0969,
      "encoder_q-layer.4": 949.4698,
      "encoder_q-layer.5": 890.2075,
      "encoder_q-layer.6": 992.3794,
      "encoder_q-layer.7": 1150.4061,
      "encoder_q-layer.8": 1380.6515,
      "encoder_q-layer.9": 1270.5737,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6737,
      "inbatch_pos_score": 1.3691,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 3.1805,
      "norm_diff": 0.0613,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2031.3872,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6743,
      "query_norm": 1.5173,
      "queue_k_norm": 1.5815,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8406,
      "sent_len_1": 66.7115,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1788,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 80000
    },
    {
      "dev_runtime": 37.2903,
      "dev_samples_per_second": 1.716,
      "dev_steps_per_second": 0.027,
      "epoch": 0.78,
      "step": 80000,
      "test_accuracy": 94.88525390625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3015986382961273,
      "test_doc_norm": 1.5760226249694824,
      "test_inbatch_neg_score": 1.0465819835662842,
      "test_inbatch_pos_score": 2.0188961029052734,
      "test_loss": 0.3015986382961273,
      "test_loss_align": 1.0064592361450195,
      "test_loss_unif": 3.147366523742676,
      "test_loss_unif_q@queue": 3.147366523742676,
      "test_norm_diff": 0.016963131725788116,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.659213662147522,
      "test_query_norm": 1.5925878286361694,
      "test_queue_k_norm": 1.5812711715698242,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.0432203933596611,
      "test_stdq": 0.0432535745203495,
      "test_stdqueue_k": 0.04909998178482056,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 37.2903,
      "dev_samples_per_second": 1.716,
      "dev_steps_per_second": 0.027,
      "epoch": 0.78,
      "eval_beir-arguana_ndcg@10": 0.38533,
      "eval_beir-arguana_recall@10": 0.65789,
      "eval_beir-arguana_recall@100": 0.93954,
      "eval_beir-arguana_recall@20": 0.78805,
      "eval_beir-avg_ndcg@10": 0.37835558333333336,
      "eval_beir-avg_recall@10": 0.44511025000000004,
      "eval_beir-avg_recall@100": 0.6245525,
      "eval_beir-avg_recall@20": 0.5054496666666667,
      "eval_beir-cqadupstack_ndcg@10": 0.26795583333333334,
      "eval_beir-cqadupstack_recall@10": 0.3588925,
      "eval_beir-cqadupstack_recall@100": 0.5835049999999999,
      "eval_beir-cqadupstack_recall@20": 0.4238666666666666,
      "eval_beir-fiqa_ndcg@10": 0.22794,
      "eval_beir-fiqa_recall@10": 0.2852,
      "eval_beir-fiqa_recall@100": 0.52992,
      "eval_beir-fiqa_recall@20": 0.33986,
      "eval_beir-nfcorpus_ndcg@10": 0.29796,
      "eval_beir-nfcorpus_recall@10": 0.14283,
      "eval_beir-nfcorpus_recall@100": 0.2767,
      "eval_beir-nfcorpus_recall@20": 0.1788,
      "eval_beir-nq_ndcg@10": 0.2705,
      "eval_beir-nq_recall@10": 0.44535,
      "eval_beir-nq_recall@100": 0.78764,
      "eval_beir-nq_recall@20": 0.5612,
      "eval_beir-quora_ndcg@10": 0.82223,
      "eval_beir-quora_recall@10": 0.91454,
      "eval_beir-quora_recall@100": 0.9852,
      "eval_beir-quora_recall@20": 0.94913,
      "eval_beir-scidocs_ndcg@10": 0.14503,
      "eval_beir-scidocs_recall@10": 0.15228,
      "eval_beir-scidocs_recall@100": 0.35082,
      "eval_beir-scidocs_recall@20": 0.20648,
      "eval_beir-scifact_ndcg@10": 0.63083,
      "eval_beir-scifact_recall@10": 0.77,
      "eval_beir-scifact_recall@100": 0.92322,
      "eval_beir-scifact_recall@20": 0.83344,
      "eval_beir-trec-covid_ndcg@10": 0.54308,
      "eval_beir-trec-covid_recall@10": 0.582,
      "eval_beir-trec-covid_recall@100": 0.4304,
      "eval_beir-trec-covid_recall@20": 0.56,
      "eval_beir-webis-touche2020_ndcg@10": 0.1927,
      "eval_beir-webis-touche2020_recall@10": 0.14212,
      "eval_beir-webis-touche2020_recall@100": 0.43858,
      "eval_beir-webis-touche2020_recall@20": 0.21367,
      "eval_senteval-avg_sts": 0.7211243300477561,
      "eval_senteval-sickr_spearman": 0.6918598219594686,
      "eval_senteval-stsb_spearman": 0.7503888381360436,
      "step": 80000,
      "test_accuracy": 94.88525390625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3015986382961273,
      "test_doc_norm": 1.5760226249694824,
      "test_inbatch_neg_score": 1.0465819835662842,
      "test_inbatch_pos_score": 2.0188961029052734,
      "test_loss": 0.3015986382961273,
      "test_loss_align": 1.0064592361450195,
      "test_loss_unif": 3.147366523742676,
      "test_loss_unif_q@queue": 3.147366523742676,
      "test_norm_diff": 0.016963131725788116,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.659213662147522,
      "test_query_norm": 1.5925878286361694,
      "test_queue_k_norm": 1.5812711715698242,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.0432203933596611,
      "test_stdq": 0.0432535745203495,
      "test_stdqueue_k": 0.04909998178482056,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 52.6367,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1638,
      "doc_norm": 1.5895,
      "encoder_q-embeddings": 1400.7617,
      "encoder_q-layer.0": 965.9867,
      "encoder_q-layer.1": 1060.2559,
      "encoder_q-layer.10": 1268.0103,
      "encoder_q-layer.11": 3028.1399,
      "encoder_q-layer.2": 1186.0958,
      "encoder_q-layer.3": 1221.2538,
      "encoder_q-layer.4": 1319.1344,
      "encoder_q-layer.5": 1304.0043,
      "encoder_q-layer.6": 1450.1825,
      "encoder_q-layer.7": 1595.1479,
      "encoder_q-layer.8": 1694.5428,
      "encoder_q-layer.9": 1360.759,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6772,
      "inbatch_pos_score": 1.3379,
      "learning_rate": 1.1055555555555556e-05,
      "loss": 3.1638,
      "norm_diff": 0.0795,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2333.945,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6777,
      "query_norm": 1.51,
      "queue_k_norm": 1.5824,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0083,
      "sent_len_1": 66.8374,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.565,
      "stdk": 0.0493,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 80100
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1626,
      "doc_norm": 1.5862,
      "encoder_q-embeddings": 1579.9305,
      "encoder_q-layer.0": 1089.163,
      "encoder_q-layer.1": 1149.2412,
      "encoder_q-layer.10": 1365.926,
      "encoder_q-layer.11": 2868.8459,
      "encoder_q-layer.2": 1424.2704,
      "encoder_q-layer.3": 1536.2289,
      "encoder_q-layer.4": 1569.3265,
      "encoder_q-layer.5": 1670.3801,
      "encoder_q-layer.6": 1677.0023,
      "encoder_q-layer.7": 1742.2429,
      "encoder_q-layer.8": 1710.124,
      "encoder_q-layer.9": 1277.5892,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6764,
      "inbatch_pos_score": 1.3779,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 3.1626,
      "norm_diff": 0.0587,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2531.037,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6772,
      "query_norm": 1.5275,
      "queue_k_norm": 1.5844,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7941,
      "sent_len_1": 66.6855,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4275,
      "stdk": 0.0492,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 80200
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1625,
      "doc_norm": 1.5842,
      "encoder_q-embeddings": 1257.8727,
      "encoder_q-layer.0": 812.7773,
      "encoder_q-layer.1": 906.0044,
      "encoder_q-layer.10": 1325.7552,
      "encoder_q-layer.11": 2915.1455,
      "encoder_q-layer.2": 1067.7902,
      "encoder_q-layer.3": 1122.0974,
      "encoder_q-layer.4": 1229.8771,
      "encoder_q-layer.5": 1304.3699,
      "encoder_q-layer.6": 1313.2994,
      "encoder_q-layer.7": 1402.3756,
      "encoder_q-layer.8": 1438.7482,
      "encoder_q-layer.9": 1269.2555,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6785,
      "inbatch_pos_score": 1.3496,
      "learning_rate": 1.0944444444444445e-05,
      "loss": 3.1625,
      "norm_diff": 0.0648,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2190.6802,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6777,
      "query_norm": 1.5195,
      "queue_k_norm": 1.584,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7864,
      "sent_len_1": 66.8181,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7763,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 80300
    },
    {
      "accuracy": 58.1055,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1595,
      "doc_norm": 1.5873,
      "encoder_q-embeddings": 2538.7866,
      "encoder_q-layer.0": 1708.1127,
      "encoder_q-layer.1": 1870.1849,
      "encoder_q-layer.10": 2477.9517,
      "encoder_q-layer.11": 5465.9282,
      "encoder_q-layer.2": 2243.8794,
      "encoder_q-layer.3": 2302.5898,
      "encoder_q-layer.4": 2429.0613,
      "encoder_q-layer.5": 2450.9224,
      "encoder_q-layer.6": 2766.3201,
      "encoder_q-layer.7": 3468.6184,
      "encoder_q-layer.8": 3345.9448,
      "encoder_q-layer.9": 2658.407,
      "epoch": 0.78,
      "inbatch_neg_score": 0.6764,
      "inbatch_pos_score": 1.4004,
      "learning_rate": 1.088888888888889e-05,
      "loss": 3.1595,
      "norm_diff": 0.0581,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4426.4098,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6772,
      "query_norm": 1.5292,
      "queue_k_norm": 1.5831,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8657,
      "sent_len_1": 66.8246,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.865,
      "stdk": 0.0492,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 80400
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1359,
      "doc_norm": 1.5847,
      "encoder_q-embeddings": 2665.5244,
      "encoder_q-layer.0": 1916.9335,
      "encoder_q-layer.1": 2065.2722,
      "encoder_q-layer.10": 2590.1963,
      "encoder_q-layer.11": 5689.1021,
      "encoder_q-layer.2": 2379.0696,
      "encoder_q-layer.3": 2439.3516,
      "encoder_q-layer.4": 2764.4619,
      "encoder_q-layer.5": 2701.4604,
      "encoder_q-layer.6": 2945.0581,
      "encoder_q-layer.7": 3222.8892,
      "encoder_q-layer.8": 3276.0549,
      "encoder_q-layer.9": 2524.1277,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6804,
      "inbatch_pos_score": 1.3408,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 3.1359,
      "norm_diff": 0.0717,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4559.3355,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6792,
      "query_norm": 1.513,
      "queue_k_norm": 1.5833,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1389,
      "sent_len_1": 66.7879,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7413,
      "stdk": 0.049,
      "stdq": 0.0453,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 80500
    },
    {
      "accuracy": 57.2266,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1759,
      "doc_norm": 1.5859,
      "encoder_q-embeddings": 1959.6002,
      "encoder_q-layer.0": 1287.25,
      "encoder_q-layer.1": 1343.1055,
      "encoder_q-layer.10": 2281.1526,
      "encoder_q-layer.11": 5452.895,
      "encoder_q-layer.2": 1538.5356,
      "encoder_q-layer.3": 1572.1865,
      "encoder_q-layer.4": 1645.0712,
      "encoder_q-layer.5": 1781.1532,
      "encoder_q-layer.6": 1979.3684,
      "encoder_q-layer.7": 2277.436,
      "encoder_q-layer.8": 2620.5503,
      "encoder_q-layer.9": 2329.1213,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6807,
      "inbatch_pos_score": 1.3496,
      "learning_rate": 1.0777777777777778e-05,
      "loss": 3.1759,
      "norm_diff": 0.0759,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3711.6782,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6807,
      "query_norm": 1.51,
      "queue_k_norm": 1.5846,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9242,
      "sent_len_1": 66.8275,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.365,
      "stdk": 0.0491,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 80600
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1578,
      "doc_norm": 1.5806,
      "encoder_q-embeddings": 2157.3584,
      "encoder_q-layer.0": 1473.7998,
      "encoder_q-layer.1": 1619.2266,
      "encoder_q-layer.10": 2698.6846,
      "encoder_q-layer.11": 5936.2935,
      "encoder_q-layer.2": 1908.2031,
      "encoder_q-layer.3": 1942.3086,
      "encoder_q-layer.4": 2097.6775,
      "encoder_q-layer.5": 2202.9539,
      "encoder_q-layer.6": 2520.8799,
      "encoder_q-layer.7": 2714.8232,
      "encoder_q-layer.8": 2988.5583,
      "encoder_q-layer.9": 2462.5679,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6832,
      "inbatch_pos_score": 1.3457,
      "learning_rate": 1.0722222222222222e-05,
      "loss": 3.1578,
      "norm_diff": 0.0549,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4185.7014,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6821,
      "query_norm": 1.5257,
      "queue_k_norm": 1.5846,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8776,
      "sent_len_1": 66.6474,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8212,
      "stdk": 0.0488,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 80700
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1511,
      "doc_norm": 1.5886,
      "encoder_q-embeddings": 2006.6925,
      "encoder_q-layer.0": 1311.2059,
      "encoder_q-layer.1": 1428.442,
      "encoder_q-layer.10": 2353.3542,
      "encoder_q-layer.11": 5332.9766,
      "encoder_q-layer.2": 1646.4973,
      "encoder_q-layer.3": 1790.9276,
      "encoder_q-layer.4": 1922.8077,
      "encoder_q-layer.5": 2116.7239,
      "encoder_q-layer.6": 2365.6714,
      "encoder_q-layer.7": 2633.0508,
      "encoder_q-layer.8": 2760.7378,
      "encoder_q-layer.9": 2444.658,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6831,
      "inbatch_pos_score": 1.3887,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 3.1511,
      "norm_diff": 0.0556,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3825.693,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6831,
      "query_norm": 1.533,
      "queue_k_norm": 1.5863,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0671,
      "sent_len_1": 66.8788,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.7325,
      "stdk": 0.0492,
      "stdq": 0.0461,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 80800
    },
    {
      "accuracy": 59.4727,
      "active_queue_size": 16384.0,
      "cl_loss": 3.168,
      "doc_norm": 1.5853,
      "encoder_q-embeddings": 1126.8324,
      "encoder_q-layer.0": 738.218,
      "encoder_q-layer.1": 832.9293,
      "encoder_q-layer.10": 1143.1877,
      "encoder_q-layer.11": 2756.5835,
      "encoder_q-layer.2": 963.8879,
      "encoder_q-layer.3": 1089.4331,
      "encoder_q-layer.4": 1213.651,
      "encoder_q-layer.5": 1338.7495,
      "encoder_q-layer.6": 1418.1177,
      "encoder_q-layer.7": 1489.3083,
      "encoder_q-layer.8": 1536.4492,
      "encoder_q-layer.9": 1213.5121,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6835,
      "inbatch_pos_score": 1.3916,
      "learning_rate": 1.0611111111111111e-05,
      "loss": 3.168,
      "norm_diff": 0.0504,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2132.4717,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6836,
      "query_norm": 1.5349,
      "queue_k_norm": 1.5887,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9637,
      "sent_len_1": 66.9363,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2262,
      "stdk": 0.049,
      "stdq": 0.0463,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 80900
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1584,
      "doc_norm": 1.5895,
      "encoder_q-embeddings": 1695.5626,
      "encoder_q-layer.0": 1172.1503,
      "encoder_q-layer.1": 1283.7329,
      "encoder_q-layer.10": 1244.2246,
      "encoder_q-layer.11": 2868.2112,
      "encoder_q-layer.2": 1506.9012,
      "encoder_q-layer.3": 1554.045,
      "encoder_q-layer.4": 1634.9988,
      "encoder_q-layer.5": 1789.9016,
      "encoder_q-layer.6": 1928.594,
      "encoder_q-layer.7": 1865.0463,
      "encoder_q-layer.8": 1792.4965,
      "encoder_q-layer.9": 1297.8828,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6844,
      "inbatch_pos_score": 1.3594,
      "learning_rate": 1.0555555555555555e-05,
      "loss": 3.1584,
      "norm_diff": 0.0661,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2565.7209,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6846,
      "query_norm": 1.5234,
      "queue_k_norm": 1.5873,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7889,
      "sent_len_1": 66.6344,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4575,
      "stdk": 0.0492,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 81000
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1767,
      "doc_norm": 1.5892,
      "encoder_q-embeddings": 1221.5774,
      "encoder_q-layer.0": 798.8763,
      "encoder_q-layer.1": 867.3085,
      "encoder_q-layer.10": 1344.6053,
      "encoder_q-layer.11": 3007.4163,
      "encoder_q-layer.2": 1062.2803,
      "encoder_q-layer.3": 1071.825,
      "encoder_q-layer.4": 1109.8955,
      "encoder_q-layer.5": 1114.2322,
      "encoder_q-layer.6": 1293.1104,
      "encoder_q-layer.7": 1556.6875,
      "encoder_q-layer.8": 1646.8997,
      "encoder_q-layer.9": 1351.084,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6818,
      "inbatch_pos_score": 1.3643,
      "learning_rate": 1.05e-05,
      "loss": 3.1767,
      "norm_diff": 0.0774,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2214.0435,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6816,
      "query_norm": 1.5118,
      "queue_k_norm": 1.5871,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8857,
      "sent_len_1": 66.9106,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3475,
      "stdk": 0.0491,
      "stdq": 0.0453,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81100
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1633,
      "doc_norm": 1.5826,
      "encoder_q-embeddings": 960.2942,
      "encoder_q-layer.0": 646.8708,
      "encoder_q-layer.1": 708.2198,
      "encoder_q-layer.10": 1157.2889,
      "encoder_q-layer.11": 2682.0654,
      "encoder_q-layer.2": 805.8773,
      "encoder_q-layer.3": 853.3541,
      "encoder_q-layer.4": 923.1658,
      "encoder_q-layer.5": 971.7415,
      "encoder_q-layer.6": 1066.8857,
      "encoder_q-layer.7": 1204.6906,
      "encoder_q-layer.8": 1324.5514,
      "encoder_q-layer.9": 1154.5015,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6821,
      "inbatch_pos_score": 1.3535,
      "learning_rate": 1.0444444444444445e-05,
      "loss": 3.1633,
      "norm_diff": 0.0648,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1861.4283,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6826,
      "query_norm": 1.5178,
      "queue_k_norm": 1.5872,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.877,
      "sent_len_1": 66.6393,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3212,
      "stdk": 0.0489,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81200
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1684,
      "doc_norm": 1.5892,
      "encoder_q-embeddings": 2554.8455,
      "encoder_q-layer.0": 1642.9718,
      "encoder_q-layer.1": 1928.7162,
      "encoder_q-layer.10": 1342.5637,
      "encoder_q-layer.11": 2877.9834,
      "encoder_q-layer.2": 2391.6687,
      "encoder_q-layer.3": 2783.4021,
      "encoder_q-layer.4": 3140.7356,
      "encoder_q-layer.5": 3858.5183,
      "encoder_q-layer.6": 4359.4097,
      "encoder_q-layer.7": 4691.5229,
      "encoder_q-layer.8": 4806.1841,
      "encoder_q-layer.9": 2729.178,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6851,
      "inbatch_pos_score": 1.3623,
      "learning_rate": 1.038888888888889e-05,
      "loss": 3.1684,
      "norm_diff": 0.071,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4917.5387,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6846,
      "query_norm": 1.5182,
      "queue_k_norm": 1.5873,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8911,
      "sent_len_1": 66.6048,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7125,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81300
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1353,
      "doc_norm": 1.5877,
      "encoder_q-embeddings": 1023.3998,
      "encoder_q-layer.0": 677.4228,
      "encoder_q-layer.1": 674.5443,
      "encoder_q-layer.10": 1269.4103,
      "encoder_q-layer.11": 2769.9185,
      "encoder_q-layer.2": 802.902,
      "encoder_q-layer.3": 827.1428,
      "encoder_q-layer.4": 851.5632,
      "encoder_q-layer.5": 862.9307,
      "encoder_q-layer.6": 1009.8301,
      "encoder_q-layer.7": 1155.5106,
      "encoder_q-layer.8": 1336.1256,
      "encoder_q-layer.9": 1216.6882,
      "epoch": 0.79,
      "inbatch_neg_score": 0.6833,
      "inbatch_pos_score": 1.3711,
      "learning_rate": 1.0333333333333333e-05,
      "loss": 3.1353,
      "norm_diff": 0.0729,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1905.3619,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6831,
      "query_norm": 1.5148,
      "queue_k_norm": 1.5871,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0704,
      "sent_len_1": 66.7664,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1375,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81400
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1558,
      "doc_norm": 1.5859,
      "encoder_q-embeddings": 1153.7661,
      "encoder_q-layer.0": 749.3093,
      "encoder_q-layer.1": 833.0377,
      "encoder_q-layer.10": 1177.8032,
      "encoder_q-layer.11": 2727.425,
      "encoder_q-layer.2": 981.3732,
      "encoder_q-layer.3": 1038.7552,
      "encoder_q-layer.4": 1124.0024,
      "encoder_q-layer.5": 1220.3698,
      "encoder_q-layer.6": 1434.7595,
      "encoder_q-layer.7": 1539.281,
      "encoder_q-layer.8": 1606.5076,
      "encoder_q-layer.9": 1269.3333,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6844,
      "inbatch_pos_score": 1.3887,
      "learning_rate": 1.0277777777777777e-05,
      "loss": 3.1558,
      "norm_diff": 0.057,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2120.155,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6855,
      "query_norm": 1.5289,
      "queue_k_norm": 1.5875,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.5732,
      "sent_len_1": 66.6563,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7788,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81500
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1533,
      "doc_norm": 1.5827,
      "encoder_q-embeddings": 1037.3945,
      "encoder_q-layer.0": 647.414,
      "encoder_q-layer.1": 664.1012,
      "encoder_q-layer.10": 1189.2025,
      "encoder_q-layer.11": 2872.0857,
      "encoder_q-layer.2": 732.4588,
      "encoder_q-layer.3": 772.7777,
      "encoder_q-layer.4": 834.499,
      "encoder_q-layer.5": 864.7521,
      "encoder_q-layer.6": 984.3871,
      "encoder_q-layer.7": 1160.3185,
      "encoder_q-layer.8": 1472.6077,
      "encoder_q-layer.9": 1217.6105,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6889,
      "inbatch_pos_score": 1.374,
      "learning_rate": 1.0222222222222223e-05,
      "loss": 3.1533,
      "norm_diff": 0.0469,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1932.2574,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6885,
      "query_norm": 1.5358,
      "queue_k_norm": 1.5878,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8039,
      "sent_len_1": 66.8151,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.22,
      "stdk": 0.0488,
      "stdq": 0.0462,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81600
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1453,
      "doc_norm": 1.5811,
      "encoder_q-embeddings": 1234.7152,
      "encoder_q-layer.0": 868.3758,
      "encoder_q-layer.1": 1008.8305,
      "encoder_q-layer.10": 1280.7366,
      "encoder_q-layer.11": 2805.1125,
      "encoder_q-layer.2": 1146.34,
      "encoder_q-layer.3": 1275.0625,
      "encoder_q-layer.4": 1425.4331,
      "encoder_q-layer.5": 1474.6484,
      "encoder_q-layer.6": 1461.6285,
      "encoder_q-layer.7": 1477.2516,
      "encoder_q-layer.8": 1549.2378,
      "encoder_q-layer.9": 1278.2766,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6893,
      "inbatch_pos_score": 1.3604,
      "learning_rate": 1.0166666666666667e-05,
      "loss": 3.1453,
      "norm_diff": 0.0542,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2204.983,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.689,
      "query_norm": 1.5268,
      "queue_k_norm": 1.5879,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0632,
      "sent_len_1": 66.8188,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8475,
      "stdk": 0.0487,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81700
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1348,
      "doc_norm": 1.59,
      "encoder_q-embeddings": 1587.7563,
      "encoder_q-layer.0": 1099.8848,
      "encoder_q-layer.1": 1305.6074,
      "encoder_q-layer.10": 1239.2166,
      "encoder_q-layer.11": 2888.23,
      "encoder_q-layer.2": 1468.6234,
      "encoder_q-layer.3": 1648.178,
      "encoder_q-layer.4": 1768.4263,
      "encoder_q-layer.5": 1842.5452,
      "encoder_q-layer.6": 1965.9202,
      "encoder_q-layer.7": 2106.22,
      "encoder_q-layer.8": 1916.136,
      "encoder_q-layer.9": 1286.8361,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6911,
      "inbatch_pos_score": 1.3496,
      "learning_rate": 1.0111111111111111e-05,
      "loss": 3.1348,
      "norm_diff": 0.0771,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2666.0876,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6914,
      "query_norm": 1.5129,
      "queue_k_norm": 1.5878,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0924,
      "sent_len_1": 66.8254,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3413,
      "stdk": 0.0491,
      "stdq": 0.0451,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81800
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1647,
      "doc_norm": 1.5934,
      "encoder_q-embeddings": 2469.5005,
      "encoder_q-layer.0": 1852.484,
      "encoder_q-layer.1": 2165.3694,
      "encoder_q-layer.10": 1399.9178,
      "encoder_q-layer.11": 3009.7151,
      "encoder_q-layer.2": 2776.6545,
      "encoder_q-layer.3": 2972.8269,
      "encoder_q-layer.4": 3202.5774,
      "encoder_q-layer.5": 3408.9404,
      "encoder_q-layer.6": 3816.6567,
      "encoder_q-layer.7": 3641.8855,
      "encoder_q-layer.8": 2414.3347,
      "encoder_q-layer.9": 1399.2362,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6909,
      "inbatch_pos_score": 1.415,
      "learning_rate": 1.0055555555555555e-05,
      "loss": 3.1647,
      "norm_diff": 0.0509,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4157.7581,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6909,
      "query_norm": 1.5426,
      "queue_k_norm": 1.5889,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.208,
      "sent_len_1": 66.9368,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.055,
      "stdk": 0.0492,
      "stdq": 0.0465,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 81900
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1462,
      "doc_norm": 1.5813,
      "encoder_q-embeddings": 1050.3915,
      "encoder_q-layer.0": 700.4301,
      "encoder_q-layer.1": 762.6502,
      "encoder_q-layer.10": 1327.6494,
      "encoder_q-layer.11": 2786.0969,
      "encoder_q-layer.2": 835.535,
      "encoder_q-layer.3": 913.1284,
      "encoder_q-layer.4": 1008.1145,
      "encoder_q-layer.5": 1077.8026,
      "encoder_q-layer.6": 1172.1234,
      "encoder_q-layer.7": 1319.7034,
      "encoder_q-layer.8": 1420.2125,
      "encoder_q-layer.9": 1233.0776,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6929,
      "inbatch_pos_score": 1.3984,
      "learning_rate": 1e-05,
      "loss": 3.1462,
      "norm_diff": 0.0504,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1995.3183,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6914,
      "query_norm": 1.5308,
      "queue_k_norm": 1.5896,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.951,
      "sent_len_1": 66.9386,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7113,
      "stdk": 0.0487,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82000
    },
    {
      "accuracy": 59.2773,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1719,
      "doc_norm": 1.6012,
      "encoder_q-embeddings": 862.9777,
      "encoder_q-layer.0": 555.4623,
      "encoder_q-layer.1": 577.1544,
      "encoder_q-layer.10": 1191.1334,
      "encoder_q-layer.11": 2572.7344,
      "encoder_q-layer.2": 642.679,
      "encoder_q-layer.3": 703.2296,
      "encoder_q-layer.4": 733.4401,
      "encoder_q-layer.5": 763.1398,
      "encoder_q-layer.6": 891.8499,
      "encoder_q-layer.7": 1069.9352,
      "encoder_q-layer.8": 1234.7516,
      "encoder_q-layer.9": 1163.613,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6912,
      "inbatch_pos_score": 1.4053,
      "learning_rate": 9.944444444444445e-06,
      "loss": 3.1719,
      "norm_diff": 0.0687,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1735.5737,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6909,
      "query_norm": 1.5325,
      "queue_k_norm": 1.5888,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9605,
      "sent_len_1": 66.765,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.8562,
      "stdk": 0.0495,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82100
    },
    {
      "accuracy": 58.3008,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1378,
      "doc_norm": 1.5927,
      "encoder_q-embeddings": 1050.9352,
      "encoder_q-layer.0": 682.3998,
      "encoder_q-layer.1": 711.1276,
      "encoder_q-layer.10": 1229.5808,
      "encoder_q-layer.11": 2710.4517,
      "encoder_q-layer.2": 819.4929,
      "encoder_q-layer.3": 838.0463,
      "encoder_q-layer.4": 909.5482,
      "encoder_q-layer.5": 933.9225,
      "encoder_q-layer.6": 1048.6969,
      "encoder_q-layer.7": 1211.6991,
      "encoder_q-layer.8": 1345.8118,
      "encoder_q-layer.9": 1184.2097,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6934,
      "inbatch_pos_score": 1.4131,
      "learning_rate": 9.888888888888889e-06,
      "loss": 3.1378,
      "norm_diff": 0.0455,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1894.4208,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6929,
      "query_norm": 1.5472,
      "queue_k_norm": 1.5898,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1597,
      "sent_len_1": 66.8207,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4613,
      "stdk": 0.0492,
      "stdq": 0.0466,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82200
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1638,
      "doc_norm": 1.5916,
      "encoder_q-embeddings": 1091.7739,
      "encoder_q-layer.0": 734.6206,
      "encoder_q-layer.1": 842.9809,
      "encoder_q-layer.10": 1242.4427,
      "encoder_q-layer.11": 2683.3745,
      "encoder_q-layer.2": 1044.2399,
      "encoder_q-layer.3": 1158.0393,
      "encoder_q-layer.4": 1235.2672,
      "encoder_q-layer.5": 1329.2577,
      "encoder_q-layer.6": 1502.4653,
      "encoder_q-layer.7": 1620.3137,
      "encoder_q-layer.8": 1674.936,
      "encoder_q-layer.9": 1248.8359,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6965,
      "inbatch_pos_score": 1.3662,
      "learning_rate": 9.833333333333333e-06,
      "loss": 3.1638,
      "norm_diff": 0.0817,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2137.7964,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6963,
      "query_norm": 1.5099,
      "queue_k_norm": 1.5903,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.972,
      "sent_len_1": 66.7428,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.02,
      "stdk": 0.0491,
      "stdq": 0.0449,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82300
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1339,
      "doc_norm": 1.5938,
      "encoder_q-embeddings": 1192.9722,
      "encoder_q-layer.0": 774.4518,
      "encoder_q-layer.1": 818.5733,
      "encoder_q-layer.10": 1310.6936,
      "encoder_q-layer.11": 2798.0615,
      "encoder_q-layer.2": 969.2155,
      "encoder_q-layer.3": 1093.8545,
      "encoder_q-layer.4": 1194.8337,
      "encoder_q-layer.5": 1199.5719,
      "encoder_q-layer.6": 1360.5826,
      "encoder_q-layer.7": 1455.6652,
      "encoder_q-layer.8": 1515.9982,
      "encoder_q-layer.9": 1287.9614,
      "epoch": 0.8,
      "inbatch_neg_score": 0.6976,
      "inbatch_pos_score": 1.375,
      "learning_rate": 9.777777777777779e-06,
      "loss": 3.1339,
      "norm_diff": 0.0675,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2135.397,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6978,
      "query_norm": 1.5263,
      "queue_k_norm": 1.5904,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8079,
      "sent_len_1": 66.7187,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9888,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82400
    },
    {
      "accuracy": 59.4727,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1589,
      "doc_norm": 1.596,
      "encoder_q-embeddings": 1631.6688,
      "encoder_q-layer.0": 1020.1454,
      "encoder_q-layer.1": 1084.6709,
      "encoder_q-layer.10": 1208.3673,
      "encoder_q-layer.11": 2635.2107,
      "encoder_q-layer.2": 1232.4893,
      "encoder_q-layer.3": 1286.8735,
      "encoder_q-layer.4": 1337.0674,
      "encoder_q-layer.5": 1511.1357,
      "encoder_q-layer.6": 1400.3566,
      "encoder_q-layer.7": 1426.9537,
      "encoder_q-layer.8": 1483.6189,
      "encoder_q-layer.9": 1168.9686,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7,
      "inbatch_pos_score": 1.4072,
      "learning_rate": 9.722222222222223e-06,
      "loss": 3.1589,
      "norm_diff": 0.0669,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2267.2417,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.6997,
      "query_norm": 1.5291,
      "queue_k_norm": 1.5929,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9664,
      "sent_len_1": 66.7294,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7413,
      "stdk": 0.0492,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 82500
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1492,
      "doc_norm": 1.5896,
      "encoder_q-embeddings": 1208.4355,
      "encoder_q-layer.0": 824.8386,
      "encoder_q-layer.1": 932.1114,
      "encoder_q-layer.10": 1160.9698,
      "encoder_q-layer.11": 2679.335,
      "encoder_q-layer.2": 1134.3196,
      "encoder_q-layer.3": 1210.6642,
      "encoder_q-layer.4": 1343.5011,
      "encoder_q-layer.5": 1395.3096,
      "encoder_q-layer.6": 1492.6533,
      "encoder_q-layer.7": 1525.5842,
      "encoder_q-layer.8": 1557.5972,
      "encoder_q-layer.9": 1283.2218,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7009,
      "inbatch_pos_score": 1.3691,
      "learning_rate": 9.666666666666667e-06,
      "loss": 3.1492,
      "norm_diff": 0.0634,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2190.6605,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7017,
      "query_norm": 1.5262,
      "queue_k_norm": 1.5919,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7419,
      "sent_len_1": 66.7727,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5437,
      "stdk": 0.049,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82600
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1473,
      "doc_norm": 1.5889,
      "encoder_q-embeddings": 428.7136,
      "encoder_q-layer.0": 287.7683,
      "encoder_q-layer.1": 301.2601,
      "encoder_q-layer.10": 639.2629,
      "encoder_q-layer.11": 1430.0751,
      "encoder_q-layer.2": 329.322,
      "encoder_q-layer.3": 351.7398,
      "encoder_q-layer.4": 370.6867,
      "encoder_q-layer.5": 407.2479,
      "encoder_q-layer.6": 511.6207,
      "encoder_q-layer.7": 575.3125,
      "encoder_q-layer.8": 677.9387,
      "encoder_q-layer.9": 638.6132,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7018,
      "inbatch_pos_score": 1.3721,
      "learning_rate": 9.61111111111111e-06,
      "loss": 3.1473,
      "norm_diff": 0.0638,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 943.6993,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7026,
      "query_norm": 1.5251,
      "queue_k_norm": 1.5921,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9716,
      "sent_len_1": 66.6937,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8988,
      "stdk": 0.0489,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82700
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1459,
      "doc_norm": 1.5994,
      "encoder_q-embeddings": 496.5228,
      "encoder_q-layer.0": 328.4564,
      "encoder_q-layer.1": 350.1421,
      "encoder_q-layer.10": 569.9866,
      "encoder_q-layer.11": 1358.2792,
      "encoder_q-layer.2": 411.1279,
      "encoder_q-layer.3": 432.5877,
      "encoder_q-layer.4": 466.9399,
      "encoder_q-layer.5": 479.3528,
      "encoder_q-layer.6": 547.3011,
      "encoder_q-layer.7": 614.1255,
      "encoder_q-layer.8": 691.1078,
      "encoder_q-layer.9": 595.005,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7027,
      "inbatch_pos_score": 1.3906,
      "learning_rate": 9.555555555555556e-06,
      "loss": 3.1459,
      "norm_diff": 0.0577,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 963.147,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7036,
      "query_norm": 1.5418,
      "queue_k_norm": 1.5923,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9439,
      "sent_len_1": 67.0518,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2663,
      "stdk": 0.0494,
      "stdq": 0.0461,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 82800
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1808,
      "doc_norm": 1.5896,
      "encoder_q-embeddings": 861.5897,
      "encoder_q-layer.0": 585.9266,
      "encoder_q-layer.1": 650.5472,
      "encoder_q-layer.10": 618.6367,
      "encoder_q-layer.11": 1385.608,
      "encoder_q-layer.2": 751.0944,
      "encoder_q-layer.3": 757.1058,
      "encoder_q-layer.4": 762.2744,
      "encoder_q-layer.5": 804.0931,
      "encoder_q-layer.6": 769.2767,
      "encoder_q-layer.7": 799.6569,
      "encoder_q-layer.8": 860.8244,
      "encoder_q-layer.9": 649.0984,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7063,
      "inbatch_pos_score": 1.4121,
      "learning_rate": 9.5e-06,
      "loss": 3.1808,
      "norm_diff": 0.0358,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1255.712,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7051,
      "query_norm": 1.5537,
      "queue_k_norm": 1.5947,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9745,
      "sent_len_1": 66.8681,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8288,
      "stdk": 0.0489,
      "stdq": 0.0466,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 82900
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.144,
      "doc_norm": 1.6011,
      "encoder_q-embeddings": 1088.8429,
      "encoder_q-layer.0": 744.1373,
      "encoder_q-layer.1": 842.8552,
      "encoder_q-layer.10": 583.097,
      "encoder_q-layer.11": 1354.1475,
      "encoder_q-layer.2": 963.0492,
      "encoder_q-layer.3": 1052.5409,
      "encoder_q-layer.4": 1207.8446,
      "encoder_q-layer.5": 1134.4492,
      "encoder_q-layer.6": 1375.9495,
      "encoder_q-layer.7": 1636.5223,
      "encoder_q-layer.8": 1367.4836,
      "encoder_q-layer.9": 637.2008,
      "epoch": 0.81,
      "inbatch_neg_score": 0.709,
      "inbatch_pos_score": 1.4092,
      "learning_rate": 9.444444444444445e-06,
      "loss": 3.144,
      "norm_diff": 0.0567,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1677.174,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7085,
      "query_norm": 1.5445,
      "queue_k_norm": 1.5939,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0475,
      "sent_len_1": 66.9843,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1775,
      "stdk": 0.0494,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 83000
    },
    {
      "accuracy": 57.2266,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1238,
      "doc_norm": 1.5977,
      "encoder_q-embeddings": 513.5504,
      "encoder_q-layer.0": 329.8863,
      "encoder_q-layer.1": 354.5135,
      "encoder_q-layer.10": 630.1382,
      "encoder_q-layer.11": 1452.3663,
      "encoder_q-layer.2": 411.706,
      "encoder_q-layer.3": 438.4115,
      "encoder_q-layer.4": 476.0347,
      "encoder_q-layer.5": 513.4276,
      "encoder_q-layer.6": 551.3489,
      "encoder_q-layer.7": 644.8112,
      "encoder_q-layer.8": 732.5329,
      "encoder_q-layer.9": 626.4625,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7094,
      "inbatch_pos_score": 1.4131,
      "learning_rate": 9.388888888888889e-06,
      "loss": 3.1238,
      "norm_diff": 0.0579,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1003.3023,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.71,
      "query_norm": 1.5399,
      "queue_k_norm": 1.5956,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0884,
      "sent_len_1": 66.8956,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2775,
      "stdk": 0.0492,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 83100
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1764,
      "doc_norm": 1.5971,
      "encoder_q-embeddings": 500.8457,
      "encoder_q-layer.0": 324.3179,
      "encoder_q-layer.1": 350.8561,
      "encoder_q-layer.10": 592.8919,
      "encoder_q-layer.11": 1401.2787,
      "encoder_q-layer.2": 391.9362,
      "encoder_q-layer.3": 415.4826,
      "encoder_q-layer.4": 445.6059,
      "encoder_q-layer.5": 482.9255,
      "encoder_q-layer.6": 544.5093,
      "encoder_q-layer.7": 598.9939,
      "encoder_q-layer.8": 705.1648,
      "encoder_q-layer.9": 604.4468,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7132,
      "inbatch_pos_score": 1.4092,
      "learning_rate": 9.333333333333334e-06,
      "loss": 3.1764,
      "norm_diff": 0.057,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 964.8494,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7124,
      "query_norm": 1.54,
      "queue_k_norm": 1.5943,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0326,
      "sent_len_1": 66.8134,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.905,
      "stdk": 0.0491,
      "stdq": 0.046,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 83200
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1453,
      "doc_norm": 1.5964,
      "encoder_q-embeddings": 636.4559,
      "encoder_q-layer.0": 447.3067,
      "encoder_q-layer.1": 487.0229,
      "encoder_q-layer.10": 629.3448,
      "encoder_q-layer.11": 1431.16,
      "encoder_q-layer.2": 572.957,
      "encoder_q-layer.3": 568.6093,
      "encoder_q-layer.4": 625.3925,
      "encoder_q-layer.5": 674.2769,
      "encoder_q-layer.6": 724.1157,
      "encoder_q-layer.7": 784.3571,
      "encoder_q-layer.8": 822.6544,
      "encoder_q-layer.9": 642.6747,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7123,
      "inbatch_pos_score": 1.4082,
      "learning_rate": 9.277777777777778e-06,
      "loss": 3.1453,
      "norm_diff": 0.0508,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1124.2812,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7119,
      "query_norm": 1.5456,
      "queue_k_norm": 1.595,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.994,
      "sent_len_1": 66.822,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0825,
      "stdk": 0.0491,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 83300
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1295,
      "doc_norm": 1.5989,
      "encoder_q-embeddings": 842.6699,
      "encoder_q-layer.0": 589.929,
      "encoder_q-layer.1": 746.4144,
      "encoder_q-layer.10": 666.9658,
      "encoder_q-layer.11": 1443.7682,
      "encoder_q-layer.2": 754.1427,
      "encoder_q-layer.3": 784.9452,
      "encoder_q-layer.4": 750.2047,
      "encoder_q-layer.5": 871.0732,
      "encoder_q-layer.6": 792.4036,
      "encoder_q-layer.7": 841.3145,
      "encoder_q-layer.8": 897.6783,
      "encoder_q-layer.9": 746.673,
      "epoch": 0.81,
      "inbatch_neg_score": 0.7183,
      "inbatch_pos_score": 1.4033,
      "learning_rate": 9.222222222222222e-06,
      "loss": 3.1295,
      "norm_diff": 0.0533,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1306.0358,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7168,
      "query_norm": 1.5457,
      "queue_k_norm": 1.5977,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.103,
      "sent_len_1": 66.6731,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7525,
      "stdk": 0.0492,
      "stdq": 0.0461,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 83400
    },
    {
      "accuracy": 57.3242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.179,
      "doc_norm": 1.6005,
      "encoder_q-embeddings": 866.5565,
      "encoder_q-layer.0": 574.7142,
      "encoder_q-layer.1": 615.9963,
      "encoder_q-layer.10": 638.2462,
      "encoder_q-layer.11": 1377.2192,
      "encoder_q-layer.2": 745.267,
      "encoder_q-layer.3": 833.1776,
      "encoder_q-layer.4": 910.8803,
      "encoder_q-layer.5": 985.2453,
      "encoder_q-layer.6": 988.4469,
      "encoder_q-layer.7": 1027.1809,
      "encoder_q-layer.8": 862.3759,
      "encoder_q-layer.9": 627.1058,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7161,
      "inbatch_pos_score": 1.4297,
      "learning_rate": 9.166666666666666e-06,
      "loss": 3.179,
      "norm_diff": 0.0486,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1340.1969,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7158,
      "query_norm": 1.5519,
      "queue_k_norm": 1.5972,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9635,
      "sent_len_1": 66.8313,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9875,
      "stdk": 0.0492,
      "stdq": 0.0465,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 83500
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1366,
      "doc_norm": 1.5946,
      "encoder_q-embeddings": 1077.759,
      "encoder_q-layer.0": 768.4734,
      "encoder_q-layer.1": 832.3279,
      "encoder_q-layer.10": 644.4451,
      "encoder_q-layer.11": 1434.908,
      "encoder_q-layer.2": 1011.7556,
      "encoder_q-layer.3": 1125.4667,
      "encoder_q-layer.4": 1298.5347,
      "encoder_q-layer.5": 1429.8214,
      "encoder_q-layer.6": 1349.2075,
      "encoder_q-layer.7": 1464.7662,
      "encoder_q-layer.8": 1109.6293,
      "encoder_q-layer.9": 683.6385,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7167,
      "inbatch_pos_score": 1.418,
      "learning_rate": 9.111111111111112e-06,
      "loss": 3.1366,
      "norm_diff": 0.0523,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1727.4957,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7163,
      "query_norm": 1.5423,
      "queue_k_norm": 1.5985,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9223,
      "sent_len_1": 66.866,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6362,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 83600
    },
    {
      "accuracy": 59.1797,
      "active_queue_size": 16384.0,
      "cl_loss": 3.164,
      "doc_norm": 1.6012,
      "encoder_q-embeddings": 607.5798,
      "encoder_q-layer.0": 413.6207,
      "encoder_q-layer.1": 451.1064,
      "encoder_q-layer.10": 644.0535,
      "encoder_q-layer.11": 1360.5404,
      "encoder_q-layer.2": 538.9041,
      "encoder_q-layer.3": 606.4691,
      "encoder_q-layer.4": 646.8145,
      "encoder_q-layer.5": 677.4081,
      "encoder_q-layer.6": 703.9799,
      "encoder_q-layer.7": 769.2367,
      "encoder_q-layer.8": 838.3353,
      "encoder_q-layer.9": 634.0035,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7172,
      "inbatch_pos_score": 1.4385,
      "learning_rate": 9.055555555555556e-06,
      "loss": 3.164,
      "norm_diff": 0.0521,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1103.3993,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7168,
      "query_norm": 1.5492,
      "queue_k_norm": 1.598,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2937,
      "sent_len_1": 66.7254,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5387,
      "stdk": 0.0492,
      "stdq": 0.0464,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 83700
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.139,
      "doc_norm": 1.6002,
      "encoder_q-embeddings": 584.3364,
      "encoder_q-layer.0": 368.6953,
      "encoder_q-layer.1": 400.4036,
      "encoder_q-layer.10": 602.1673,
      "encoder_q-layer.11": 1444.5146,
      "encoder_q-layer.2": 452.7757,
      "encoder_q-layer.3": 483.8344,
      "encoder_q-layer.4": 507.1364,
      "encoder_q-layer.5": 544.5378,
      "encoder_q-layer.6": 565.6555,
      "encoder_q-layer.7": 653.4116,
      "encoder_q-layer.8": 708.0535,
      "encoder_q-layer.9": 598.1591,
      "epoch": 0.82,
      "inbatch_neg_score": 0.72,
      "inbatch_pos_score": 1.3887,
      "learning_rate": 9e-06,
      "loss": 3.139,
      "norm_diff": 0.0686,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1025.6236,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7197,
      "query_norm": 1.5317,
      "queue_k_norm": 1.5976,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9398,
      "sent_len_1": 66.7185,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6337,
      "stdk": 0.0492,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 83800
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1456,
      "doc_norm": 1.5985,
      "encoder_q-embeddings": 529.8552,
      "encoder_q-layer.0": 362.4557,
      "encoder_q-layer.1": 405.6829,
      "encoder_q-layer.10": 603.1805,
      "encoder_q-layer.11": 1366.8771,
      "encoder_q-layer.2": 454.6916,
      "encoder_q-layer.3": 537.0372,
      "encoder_q-layer.4": 539.2597,
      "encoder_q-layer.5": 541.2477,
      "encoder_q-layer.6": 562.0154,
      "encoder_q-layer.7": 632.7623,
      "encoder_q-layer.8": 672.1741,
      "encoder_q-layer.9": 612.2368,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7186,
      "inbatch_pos_score": 1.3838,
      "learning_rate": 8.944444444444444e-06,
      "loss": 3.1456,
      "norm_diff": 0.071,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1003.5439,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7197,
      "query_norm": 1.5275,
      "queue_k_norm": 1.597,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9981,
      "sent_len_1": 66.4758,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4112,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 83900
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1624,
      "doc_norm": 1.5966,
      "encoder_q-embeddings": 687.4369,
      "encoder_q-layer.0": 452.4724,
      "encoder_q-layer.1": 501.2323,
      "encoder_q-layer.10": 615.1482,
      "encoder_q-layer.11": 1521.6288,
      "encoder_q-layer.2": 564.5645,
      "encoder_q-layer.3": 584.6169,
      "encoder_q-layer.4": 659.0112,
      "encoder_q-layer.5": 674.6292,
      "encoder_q-layer.6": 687.08,
      "encoder_q-layer.7": 742.1185,
      "encoder_q-layer.8": 775.5295,
      "encoder_q-layer.9": 649.0464,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7211,
      "inbatch_pos_score": 1.3955,
      "learning_rate": 8.88888888888889e-06,
      "loss": 3.1624,
      "norm_diff": 0.0658,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1164.1368,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7217,
      "query_norm": 1.5308,
      "queue_k_norm": 1.5979,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1195,
      "sent_len_1": 66.7199,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.96,
      "stdk": 0.049,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84000
    },
    {
      "accuracy": 57.8125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1419,
      "doc_norm": 1.6016,
      "encoder_q-embeddings": 3731.8335,
      "encoder_q-layer.0": 2702.2905,
      "encoder_q-layer.1": 3414.4497,
      "encoder_q-layer.10": 603.6252,
      "encoder_q-layer.11": 1463.6638,
      "encoder_q-layer.2": 4198.4043,
      "encoder_q-layer.3": 4348.207,
      "encoder_q-layer.4": 5152.7202,
      "encoder_q-layer.5": 4850.0732,
      "encoder_q-layer.6": 4826.2427,
      "encoder_q-layer.7": 4564.3564,
      "encoder_q-layer.8": 2613.7126,
      "encoder_q-layer.9": 927.2697,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7209,
      "inbatch_pos_score": 1.4141,
      "learning_rate": 8.833333333333334e-06,
      "loss": 3.1419,
      "norm_diff": 0.0713,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5495.3062,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7217,
      "query_norm": 1.5304,
      "queue_k_norm": 1.5972,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9666,
      "sent_len_1": 66.9023,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4225,
      "stdk": 0.0492,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 84100
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1489,
      "doc_norm": 1.5979,
      "encoder_q-embeddings": 681.487,
      "encoder_q-layer.0": 453.9416,
      "encoder_q-layer.1": 523.4345,
      "encoder_q-layer.10": 658.0157,
      "encoder_q-layer.11": 1447.5847,
      "encoder_q-layer.2": 637.6132,
      "encoder_q-layer.3": 705.1503,
      "encoder_q-layer.4": 783.4854,
      "encoder_q-layer.5": 846.2351,
      "encoder_q-layer.6": 843.8357,
      "encoder_q-layer.7": 870.6875,
      "encoder_q-layer.8": 841.5008,
      "encoder_q-layer.9": 659.7213,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7238,
      "inbatch_pos_score": 1.4111,
      "learning_rate": 8.777777777777778e-06,
      "loss": 3.1489,
      "norm_diff": 0.0535,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1218.3007,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7231,
      "query_norm": 1.5444,
      "queue_k_norm": 1.5986,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0287,
      "sent_len_1": 66.7097,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7413,
      "stdk": 0.049,
      "stdq": 0.0461,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84200
    },
    {
      "accuracy": 51.9531,
      "active_queue_size": 16384.0,
      "cl_loss": 3.153,
      "doc_norm": 1.5956,
      "encoder_q-embeddings": 490.7633,
      "encoder_q-layer.0": 317.4752,
      "encoder_q-layer.1": 340.8247,
      "encoder_q-layer.10": 719.5267,
      "encoder_q-layer.11": 1579.9561,
      "encoder_q-layer.2": 378.8933,
      "encoder_q-layer.3": 403.4693,
      "encoder_q-layer.4": 424.0768,
      "encoder_q-layer.5": 440.6971,
      "encoder_q-layer.6": 502.9122,
      "encoder_q-layer.7": 589.1382,
      "encoder_q-layer.8": 721.7319,
      "encoder_q-layer.9": 685.2681,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7229,
      "inbatch_pos_score": 1.3877,
      "learning_rate": 8.722222222222224e-06,
      "loss": 3.153,
      "norm_diff": 0.0588,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1029.8083,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7222,
      "query_norm": 1.5367,
      "queue_k_norm": 1.5989,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9428,
      "sent_len_1": 66.6355,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4412,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84300
    },
    {
      "accuracy": 58.4961,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1415,
      "doc_norm": 1.6035,
      "encoder_q-embeddings": 691.9155,
      "encoder_q-layer.0": 480.5439,
      "encoder_q-layer.1": 559.4642,
      "encoder_q-layer.10": 621.3391,
      "encoder_q-layer.11": 1385.2007,
      "encoder_q-layer.2": 662.6675,
      "encoder_q-layer.3": 768.387,
      "encoder_q-layer.4": 809.3804,
      "encoder_q-layer.5": 795.728,
      "encoder_q-layer.6": 797.606,
      "encoder_q-layer.7": 792.3213,
      "encoder_q-layer.8": 722.0623,
      "encoder_q-layer.9": 605.1194,
      "epoch": 0.82,
      "inbatch_neg_score": 0.724,
      "inbatch_pos_score": 1.4521,
      "learning_rate": 8.666666666666668e-06,
      "loss": 3.1415,
      "norm_diff": 0.054,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1189.4502,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7246,
      "query_norm": 1.5495,
      "queue_k_norm": 1.5982,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1075,
      "sent_len_1": 66.9742,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.21,
      "stdk": 0.0492,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 84400
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1544,
      "doc_norm": 1.6024,
      "encoder_q-embeddings": 3253.1201,
      "encoder_q-layer.0": 2198.7556,
      "encoder_q-layer.1": 2634.019,
      "encoder_q-layer.10": 633.363,
      "encoder_q-layer.11": 1461.712,
      "encoder_q-layer.2": 2952.6055,
      "encoder_q-layer.3": 3175.4258,
      "encoder_q-layer.4": 3372.7039,
      "encoder_q-layer.5": 3328.1389,
      "encoder_q-layer.6": 3411.1968,
      "encoder_q-layer.7": 3061.262,
      "encoder_q-layer.8": 2372.1958,
      "encoder_q-layer.9": 800.1552,
      "epoch": 0.82,
      "inbatch_neg_score": 0.7264,
      "inbatch_pos_score": 1.3789,
      "learning_rate": 8.611111111111112e-06,
      "loss": 3.1544,
      "norm_diff": 0.0755,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4076.0032,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7266,
      "query_norm": 1.5269,
      "queue_k_norm": 1.6005,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0041,
      "sent_len_1": 67.0007,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7887,
      "stdk": 0.0492,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84500
    },
    {
      "accuracy": 58.6914,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1551,
      "doc_norm": 1.5968,
      "encoder_q-embeddings": 556.0569,
      "encoder_q-layer.0": 384.8043,
      "encoder_q-layer.1": 430.8293,
      "encoder_q-layer.10": 593.0539,
      "encoder_q-layer.11": 1371.9709,
      "encoder_q-layer.2": 482.5587,
      "encoder_q-layer.3": 532.9375,
      "encoder_q-layer.4": 522.3454,
      "encoder_q-layer.5": 552.2042,
      "encoder_q-layer.6": 644.8157,
      "encoder_q-layer.7": 706.3272,
      "encoder_q-layer.8": 737.3195,
      "encoder_q-layer.9": 608.5025,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7253,
      "inbatch_pos_score": 1.4307,
      "learning_rate": 8.555555555555556e-06,
      "loss": 3.1551,
      "norm_diff": 0.0589,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1025.2035,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7251,
      "query_norm": 1.5379,
      "queue_k_norm": 1.6015,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.046,
      "sent_len_1": 66.6881,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4825,
      "stdk": 0.0489,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84600
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1563,
      "doc_norm": 1.6039,
      "encoder_q-embeddings": 1543.8729,
      "encoder_q-layer.0": 1028.4294,
      "encoder_q-layer.1": 1231.637,
      "encoder_q-layer.10": 1314.9498,
      "encoder_q-layer.11": 2858.948,
      "encoder_q-layer.2": 1227.1608,
      "encoder_q-layer.3": 1206.3489,
      "encoder_q-layer.4": 1231.4808,
      "encoder_q-layer.5": 1246.3386,
      "encoder_q-layer.6": 1305.6313,
      "encoder_q-layer.7": 1459.3333,
      "encoder_q-layer.8": 1533.0945,
      "encoder_q-layer.9": 1288.6835,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7256,
      "inbatch_pos_score": 1.4141,
      "learning_rate": 8.500000000000002e-06,
      "loss": 3.1563,
      "norm_diff": 0.0542,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2304.7666,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7251,
      "query_norm": 1.5497,
      "queue_k_norm": 1.6011,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0792,
      "sent_len_1": 66.8078,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2475,
      "stdk": 0.0492,
      "stdq": 0.0464,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84700
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.131,
      "doc_norm": 1.6018,
      "encoder_q-embeddings": 4574.9707,
      "encoder_q-layer.0": 3546.7019,
      "encoder_q-layer.1": 3964.5254,
      "encoder_q-layer.10": 1361.9261,
      "encoder_q-layer.11": 2838.9646,
      "encoder_q-layer.2": 5080.2549,
      "encoder_q-layer.3": 5959.5107,
      "encoder_q-layer.4": 6148.957,
      "encoder_q-layer.5": 6843.6494,
      "encoder_q-layer.6": 7309.6299,
      "encoder_q-layer.7": 6801.7412,
      "encoder_q-layer.8": 4453.23,
      "encoder_q-layer.9": 2061.4341,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7253,
      "inbatch_pos_score": 1.3672,
      "learning_rate": 8.444444444444446e-06,
      "loss": 3.131,
      "norm_diff": 0.0791,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 7492.1397,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7251,
      "query_norm": 1.5227,
      "queue_k_norm": 1.6017,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8694,
      "sent_len_1": 66.5674,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.5925,
      "stdk": 0.0491,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 84800
    },
    {
      "accuracy": 54.3945,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1282,
      "doc_norm": 1.5967,
      "encoder_q-embeddings": 1048.3694,
      "encoder_q-layer.0": 695.0166,
      "encoder_q-layer.1": 764.4494,
      "encoder_q-layer.10": 1357.9594,
      "encoder_q-layer.11": 2991.2676,
      "encoder_q-layer.2": 842.9973,
      "encoder_q-layer.3": 911.1052,
      "encoder_q-layer.4": 943.0338,
      "encoder_q-layer.5": 1012.1803,
      "encoder_q-layer.6": 1116.3276,
      "encoder_q-layer.7": 1312.1914,
      "encoder_q-layer.8": 1422.7733,
      "encoder_q-layer.9": 1260.2322,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7246,
      "inbatch_pos_score": 1.3965,
      "learning_rate": 8.38888888888889e-06,
      "loss": 3.1282,
      "norm_diff": 0.0598,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2056.5306,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7241,
      "query_norm": 1.5369,
      "queue_k_norm": 1.5989,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9999,
      "sent_len_1": 66.7752,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3738,
      "stdk": 0.0489,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 84900
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1382,
      "doc_norm": 1.602,
      "encoder_q-embeddings": 1000.2576,
      "encoder_q-layer.0": 679.7639,
      "encoder_q-layer.1": 776.4047,
      "encoder_q-layer.10": 1191.5157,
      "encoder_q-layer.11": 2829.4346,
      "encoder_q-layer.2": 953.0549,
      "encoder_q-layer.3": 976.9055,
      "encoder_q-layer.4": 1086.0452,
      "encoder_q-layer.5": 1072.3759,
      "encoder_q-layer.6": 1189.3647,
      "encoder_q-layer.7": 1365.1505,
      "encoder_q-layer.8": 1545.2611,
      "encoder_q-layer.9": 1232.9398,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7248,
      "inbatch_pos_score": 1.4258,
      "learning_rate": 8.333333333333334e-06,
      "loss": 3.1382,
      "norm_diff": 0.0686,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2046.3802,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7251,
      "query_norm": 1.5334,
      "queue_k_norm": 1.6017,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.002,
      "sent_len_1": 66.784,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9725,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 85000
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1567,
      "doc_norm": 1.5988,
      "encoder_q-embeddings": 3162.4412,
      "encoder_q-layer.0": 2152.3279,
      "encoder_q-layer.1": 2349.9121,
      "encoder_q-layer.10": 1247.0021,
      "encoder_q-layer.11": 2890.3755,
      "encoder_q-layer.2": 2775.7966,
      "encoder_q-layer.3": 2781.542,
      "encoder_q-layer.4": 2976.0833,
      "encoder_q-layer.5": 3188.4819,
      "encoder_q-layer.6": 2511.6277,
      "encoder_q-layer.7": 2752.2085,
      "encoder_q-layer.8": 2360.4963,
      "encoder_q-layer.9": 1362.9579,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7265,
      "inbatch_pos_score": 1.4121,
      "learning_rate": 8.27777777777778e-06,
      "loss": 3.1567,
      "norm_diff": 0.0699,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3988.1197,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7271,
      "query_norm": 1.5289,
      "queue_k_norm": 1.601,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9367,
      "sent_len_1": 66.8346,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5137,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 85100
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1397,
      "doc_norm": 1.6028,
      "encoder_q-embeddings": 889.1048,
      "encoder_q-layer.0": 582.3555,
      "encoder_q-layer.1": 620.5801,
      "encoder_q-layer.10": 1275.09,
      "encoder_q-layer.11": 2905.2439,
      "encoder_q-layer.2": 726.3112,
      "encoder_q-layer.3": 767.9339,
      "encoder_q-layer.4": 802.9529,
      "encoder_q-layer.5": 845.6888,
      "encoder_q-layer.6": 1012.9293,
      "encoder_q-layer.7": 1177.4303,
      "encoder_q-layer.8": 1408.6107,
      "encoder_q-layer.9": 1269.1923,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7273,
      "inbatch_pos_score": 1.4287,
      "learning_rate": 8.222222222222223e-06,
      "loss": 3.1397,
      "norm_diff": 0.056,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1918.7955,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7275,
      "query_norm": 1.5468,
      "queue_k_norm": 1.6014,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9942,
      "sent_len_1": 66.8168,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6438,
      "stdk": 0.0491,
      "stdq": 0.0462,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 85200
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1592,
      "doc_norm": 1.5969,
      "encoder_q-embeddings": 1285.7178,
      "encoder_q-layer.0": 818.681,
      "encoder_q-layer.1": 865.8196,
      "encoder_q-layer.10": 1246.8296,
      "encoder_q-layer.11": 2745.4724,
      "encoder_q-layer.2": 977.0714,
      "encoder_q-layer.3": 1034.3497,
      "encoder_q-layer.4": 1130.7074,
      "encoder_q-layer.5": 1208.6165,
      "encoder_q-layer.6": 1169.6394,
      "encoder_q-layer.7": 1234.1489,
      "encoder_q-layer.8": 1359.6877,
      "encoder_q-layer.9": 1200.1471,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7276,
      "inbatch_pos_score": 1.417,
      "learning_rate": 8.166666666666668e-06,
      "loss": 3.1592,
      "norm_diff": 0.0584,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2067.1677,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.728,
      "query_norm": 1.5385,
      "queue_k_norm": 1.6024,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9571,
      "sent_len_1": 66.6823,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1413,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 85300
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1585,
      "doc_norm": 1.5986,
      "encoder_q-embeddings": 1364.6509,
      "encoder_q-layer.0": 898.8739,
      "encoder_q-layer.1": 970.1165,
      "encoder_q-layer.10": 1196.5603,
      "encoder_q-layer.11": 2751.717,
      "encoder_q-layer.2": 1025.3386,
      "encoder_q-layer.3": 1137.9042,
      "encoder_q-layer.4": 1291.9949,
      "encoder_q-layer.5": 1297.7759,
      "encoder_q-layer.6": 1276.6163,
      "encoder_q-layer.7": 1374.432,
      "encoder_q-layer.8": 1469.2733,
      "encoder_q-layer.9": 1227.4169,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7313,
      "inbatch_pos_score": 1.4072,
      "learning_rate": 8.111111111111112e-06,
      "loss": 3.1585,
      "norm_diff": 0.0726,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2166.1527,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7305,
      "query_norm": 1.526,
      "queue_k_norm": 1.6017,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9035,
      "sent_len_1": 66.7035,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.795,
      "stdk": 0.0489,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 85400
    },
    {
      "accuracy": 54.6875,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1515,
      "doc_norm": 1.6036,
      "encoder_q-embeddings": 2079.8562,
      "encoder_q-layer.0": 1334.6339,
      "encoder_q-layer.1": 1551.1418,
      "encoder_q-layer.10": 1339.9177,
      "encoder_q-layer.11": 3042.4883,
      "encoder_q-layer.2": 1814.6854,
      "encoder_q-layer.3": 2048.2815,
      "encoder_q-layer.4": 2345.9646,
      "encoder_q-layer.5": 2693.9785,
      "encoder_q-layer.6": 3010.7546,
      "encoder_q-layer.7": 3302.9888,
      "encoder_q-layer.8": 2729.5837,
      "encoder_q-layer.9": 1526.1007,
      "epoch": 0.83,
      "inbatch_neg_score": 0.7309,
      "inbatch_pos_score": 1.4014,
      "learning_rate": 8.055555555555557e-06,
      "loss": 3.1515,
      "norm_diff": 0.0741,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3521.0177,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7319,
      "query_norm": 1.5295,
      "queue_k_norm": 1.6022,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9643,
      "sent_len_1": 66.799,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4338,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 85500
    },
    {
      "accuracy": 57.8125,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1468,
      "doc_norm": 1.6094,
      "encoder_q-embeddings": 4573.3838,
      "encoder_q-layer.0": 3278.594,
      "encoder_q-layer.1": 3942.1924,
      "encoder_q-layer.10": 1353.2748,
      "encoder_q-layer.11": 2823.8647,
      "encoder_q-layer.2": 5481.6953,
      "encoder_q-layer.3": 6413.5415,
      "encoder_q-layer.4": 6130.5049,
      "encoder_q-layer.5": 5167.4219,
      "encoder_q-layer.6": 3894.2197,
      "encoder_q-layer.7": 3471.3511,
      "encoder_q-layer.8": 2912.1863,
      "encoder_q-layer.9": 1394.115,
      "epoch": 0.84,
      "inbatch_neg_score": 0.732,
      "inbatch_pos_score": 1.4248,
      "learning_rate": 8.000000000000001e-06,
      "loss": 3.1468,
      "norm_diff": 0.0694,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6411.4491,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7319,
      "query_norm": 1.54,
      "queue_k_norm": 1.6033,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8493,
      "sent_len_1": 66.9353,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4975,
      "stdk": 0.0493,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 85600
    },
    {
      "accuracy": 58.0078,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1372,
      "doc_norm": 1.6069,
      "encoder_q-embeddings": 904.0558,
      "encoder_q-layer.0": 568.121,
      "encoder_q-layer.1": 593.3531,
      "encoder_q-layer.10": 1178.7076,
      "encoder_q-layer.11": 2774.7124,
      "encoder_q-layer.2": 688.4644,
      "encoder_q-layer.3": 736.2881,
      "encoder_q-layer.4": 772.3284,
      "encoder_q-layer.5": 815.9851,
      "encoder_q-layer.6": 946.6337,
      "encoder_q-layer.7": 1118.9174,
      "encoder_q-layer.8": 1313.9186,
      "encoder_q-layer.9": 1167.4999,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7341,
      "inbatch_pos_score": 1.4141,
      "learning_rate": 7.944444444444445e-06,
      "loss": 3.1372,
      "norm_diff": 0.0697,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1871.3746,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7339,
      "query_norm": 1.5372,
      "queue_k_norm": 1.6034,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9562,
      "sent_len_1": 66.7299,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.66,
      "stdk": 0.0492,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 85700
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1416,
      "doc_norm": 1.5957,
      "encoder_q-embeddings": 990.4097,
      "encoder_q-layer.0": 649.7525,
      "encoder_q-layer.1": 677.3521,
      "encoder_q-layer.10": 1219.6522,
      "encoder_q-layer.11": 2874.073,
      "encoder_q-layer.2": 785.6522,
      "encoder_q-layer.3": 807.7136,
      "encoder_q-layer.4": 927.0027,
      "encoder_q-layer.5": 971.7631,
      "encoder_q-layer.6": 1128.814,
      "encoder_q-layer.7": 1328.9222,
      "encoder_q-layer.8": 1391.2168,
      "encoder_q-layer.9": 1248.129,
      "epoch": 0.84,
      "inbatch_neg_score": 0.737,
      "inbatch_pos_score": 1.3955,
      "learning_rate": 7.88888888888889e-06,
      "loss": 3.1416,
      "norm_diff": 0.0705,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1976.9629,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7349,
      "query_norm": 1.5253,
      "queue_k_norm": 1.6039,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2476,
      "sent_len_1": 66.8576,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0125,
      "stdk": 0.0487,
      "stdq": 0.0451,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 85800
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.146,
      "doc_norm": 1.6049,
      "encoder_q-embeddings": 1252.8145,
      "encoder_q-layer.0": 799.9335,
      "encoder_q-layer.1": 857.5286,
      "encoder_q-layer.10": 1288.0626,
      "encoder_q-layer.11": 2983.6035,
      "encoder_q-layer.2": 937.4642,
      "encoder_q-layer.3": 985.9155,
      "encoder_q-layer.4": 1121.1606,
      "encoder_q-layer.5": 1193.4377,
      "encoder_q-layer.6": 1313.0557,
      "encoder_q-layer.7": 1419.0109,
      "encoder_q-layer.8": 1521.5708,
      "encoder_q-layer.9": 1278.7117,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7318,
      "inbatch_pos_score": 1.3916,
      "learning_rate": 7.833333333333333e-06,
      "loss": 3.146,
      "norm_diff": 0.0758,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2191.1509,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7324,
      "query_norm": 1.5291,
      "queue_k_norm": 1.6017,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7924,
      "sent_len_1": 66.5727,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8088,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 85900
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1046,
      "doc_norm": 1.6007,
      "encoder_q-embeddings": 981.351,
      "encoder_q-layer.0": 633.0244,
      "encoder_q-layer.1": 674.9125,
      "encoder_q-layer.10": 1354.5758,
      "encoder_q-layer.11": 2913.9653,
      "encoder_q-layer.2": 765.5729,
      "encoder_q-layer.3": 834.6436,
      "encoder_q-layer.4": 882.183,
      "encoder_q-layer.5": 897.3538,
      "encoder_q-layer.6": 1051.8455,
      "encoder_q-layer.7": 1161.2206,
      "encoder_q-layer.8": 1411.8285,
      "encoder_q-layer.9": 1236.9182,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7329,
      "inbatch_pos_score": 1.4258,
      "learning_rate": 7.777777777777777e-06,
      "loss": 3.1046,
      "norm_diff": 0.0615,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1992.8727,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7334,
      "query_norm": 1.5392,
      "queue_k_norm": 1.6056,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9925,
      "sent_len_1": 66.7666,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6287,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 86000
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1393,
      "doc_norm": 1.6073,
      "encoder_q-embeddings": 1935.1195,
      "encoder_q-layer.0": 1400.1554,
      "encoder_q-layer.1": 1413.1868,
      "encoder_q-layer.10": 1166.6813,
      "encoder_q-layer.11": 2772.5769,
      "encoder_q-layer.2": 1591.1808,
      "encoder_q-layer.3": 1695.1853,
      "encoder_q-layer.4": 1889.0079,
      "encoder_q-layer.5": 1934.7917,
      "encoder_q-layer.6": 2065.9685,
      "encoder_q-layer.7": 2448.9919,
      "encoder_q-layer.8": 2395.1282,
      "encoder_q-layer.9": 1683.3589,
      "epoch": 0.84,
      "inbatch_neg_score": 0.733,
      "inbatch_pos_score": 1.4219,
      "learning_rate": 7.722222222222223e-06,
      "loss": 3.1393,
      "norm_diff": 0.0665,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3035.562,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7329,
      "query_norm": 1.5408,
      "queue_k_norm": 1.6028,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1061,
      "sent_len_1": 66.7861,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4375,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 86100
    },
    {
      "accuracy": 51.1719,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1442,
      "doc_norm": 1.6042,
      "encoder_q-embeddings": 7605.2085,
      "encoder_q-layer.0": 4997.3813,
      "encoder_q-layer.1": 5407.3901,
      "encoder_q-layer.10": 1403.3583,
      "encoder_q-layer.11": 3044.7278,
      "encoder_q-layer.2": 6623.8223,
      "encoder_q-layer.3": 6839.3921,
      "encoder_q-layer.4": 7684.2559,
      "encoder_q-layer.5": 7646.2261,
      "encoder_q-layer.6": 7084.8569,
      "encoder_q-layer.7": 6477.8804,
      "encoder_q-layer.8": 5013.4902,
      "encoder_q-layer.9": 1746.9332,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7316,
      "inbatch_pos_score": 1.375,
      "learning_rate": 7.666666666666667e-06,
      "loss": 3.1442,
      "norm_diff": 0.0734,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9196.6719,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7314,
      "query_norm": 1.5309,
      "queue_k_norm": 1.6035,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0736,
      "sent_len_1": 66.6819,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.41,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 86200
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1354,
      "doc_norm": 1.601,
      "encoder_q-embeddings": 590.8639,
      "encoder_q-layer.0": 387.4693,
      "encoder_q-layer.1": 433.4784,
      "encoder_q-layer.10": 585.4716,
      "encoder_q-layer.11": 1379.1672,
      "encoder_q-layer.2": 498.3334,
      "encoder_q-layer.3": 533.4316,
      "encoder_q-layer.4": 577.3383,
      "encoder_q-layer.5": 623.726,
      "encoder_q-layer.6": 655.6407,
      "encoder_q-layer.7": 753.723,
      "encoder_q-layer.8": 750.6927,
      "encoder_q-layer.9": 616.2831,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7336,
      "inbatch_pos_score": 1.4336,
      "learning_rate": 7.611111111111112e-06,
      "loss": 3.1354,
      "norm_diff": 0.0469,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1060.9955,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7334,
      "query_norm": 1.5541,
      "queue_k_norm": 1.6046,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0191,
      "sent_len_1": 66.6135,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5613,
      "stdk": 0.0489,
      "stdq": 0.0465,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 86300
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1295,
      "doc_norm": 1.5977,
      "encoder_q-embeddings": 509.8725,
      "encoder_q-layer.0": 338.8928,
      "encoder_q-layer.1": 361.049,
      "encoder_q-layer.10": 612.2377,
      "encoder_q-layer.11": 1426.6755,
      "encoder_q-layer.2": 400.5798,
      "encoder_q-layer.3": 412.2474,
      "encoder_q-layer.4": 463.5526,
      "encoder_q-layer.5": 468.6347,
      "encoder_q-layer.6": 569.4904,
      "encoder_q-layer.7": 629.0486,
      "encoder_q-layer.8": 709.0264,
      "encoder_q-layer.9": 633.4814,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7355,
      "inbatch_pos_score": 1.4053,
      "learning_rate": 7.555555555555556e-06,
      "loss": 3.1295,
      "norm_diff": 0.0578,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 986.0093,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7339,
      "query_norm": 1.5399,
      "queue_k_norm": 1.6045,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1531,
      "sent_len_1": 66.6727,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.49,
      "stdk": 0.0487,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 86400
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1357,
      "doc_norm": 1.6043,
      "encoder_q-embeddings": 805.4828,
      "encoder_q-layer.0": 555.5412,
      "encoder_q-layer.1": 631.5233,
      "encoder_q-layer.10": 676.4863,
      "encoder_q-layer.11": 1441.1118,
      "encoder_q-layer.2": 717.6765,
      "encoder_q-layer.3": 735.3716,
      "encoder_q-layer.4": 758.8245,
      "encoder_q-layer.5": 664.8735,
      "encoder_q-layer.6": 661.0291,
      "encoder_q-layer.7": 714.8104,
      "encoder_q-layer.8": 815.4595,
      "encoder_q-layer.9": 667.0645,
      "epoch": 0.84,
      "inbatch_neg_score": 0.7334,
      "inbatch_pos_score": 1.4062,
      "learning_rate": 7.5e-06,
      "loss": 3.1357,
      "norm_diff": 0.071,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1214.6327,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7329,
      "query_norm": 1.5334,
      "queue_k_norm": 1.6068,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0352,
      "sent_len_1": 66.7954,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7788,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 86500
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1535,
      "doc_norm": 1.6114,
      "encoder_q-embeddings": 489.2008,
      "encoder_q-layer.0": 311.397,
      "encoder_q-layer.1": 331.0533,
      "encoder_q-layer.10": 718.8423,
      "encoder_q-layer.11": 1579.0376,
      "encoder_q-layer.2": 381.2522,
      "encoder_q-layer.3": 399.7148,
      "encoder_q-layer.4": 425.2156,
      "encoder_q-layer.5": 457.174,
      "encoder_q-layer.6": 517.2087,
      "encoder_q-layer.7": 639.0128,
      "encoder_q-layer.8": 724.2213,
      "encoder_q-layer.9": 674.0844,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7349,
      "inbatch_pos_score": 1.418,
      "learning_rate": 7.444444444444444e-06,
      "loss": 3.1535,
      "norm_diff": 0.0728,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1040.1821,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7349,
      "query_norm": 1.5385,
      "queue_k_norm": 1.6045,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.822,
      "sent_len_1": 66.6401,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2262,
      "stdk": 0.0493,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 86600
    },
    {
      "accuracy": 53.2227,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1599,
      "doc_norm": 1.6019,
      "encoder_q-embeddings": 1172.1667,
      "encoder_q-layer.0": 845.6879,
      "encoder_q-layer.1": 898.5078,
      "encoder_q-layer.10": 701.1465,
      "encoder_q-layer.11": 1480.2014,
      "encoder_q-layer.2": 1000.3802,
      "encoder_q-layer.3": 957.5507,
      "encoder_q-layer.4": 903.1422,
      "encoder_q-layer.5": 717.8761,
      "encoder_q-layer.6": 784.0015,
      "encoder_q-layer.7": 842.14,
      "encoder_q-layer.8": 1002.8482,
      "encoder_q-layer.9": 753.6838,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7361,
      "inbatch_pos_score": 1.3945,
      "learning_rate": 7.38888888888889e-06,
      "loss": 3.1599,
      "norm_diff": 0.063,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1481.346,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7354,
      "query_norm": 1.5388,
      "queue_k_norm": 1.6055,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.05,
      "sent_len_1": 66.9407,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3963,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 86700
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1255,
      "doc_norm": 1.6102,
      "encoder_q-embeddings": 1150.7521,
      "encoder_q-layer.0": 829.0579,
      "encoder_q-layer.1": 1123.0554,
      "encoder_q-layer.10": 648.4327,
      "encoder_q-layer.11": 1456.2189,
      "encoder_q-layer.2": 1261.7789,
      "encoder_q-layer.3": 1400.9304,
      "encoder_q-layer.4": 1429.0868,
      "encoder_q-layer.5": 1311.3599,
      "encoder_q-layer.6": 1349.9885,
      "encoder_q-layer.7": 1128.6432,
      "encoder_q-layer.8": 981.3893,
      "encoder_q-layer.9": 665.5588,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7377,
      "inbatch_pos_score": 1.4277,
      "learning_rate": 7.333333333333334e-06,
      "loss": 3.1255,
      "norm_diff": 0.0767,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1768.8606,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7373,
      "query_norm": 1.5335,
      "queue_k_norm": 1.6065,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9824,
      "sent_len_1": 66.7072,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6175,
      "stdk": 0.0493,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 86800
    },
    {
      "accuracy": 58.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1291,
      "doc_norm": 1.6133,
      "encoder_q-embeddings": 508.1518,
      "encoder_q-layer.0": 334.6613,
      "encoder_q-layer.1": 351.3994,
      "encoder_q-layer.10": 616.761,
      "encoder_q-layer.11": 1414.5791,
      "encoder_q-layer.2": 396.5141,
      "encoder_q-layer.3": 407.5815,
      "encoder_q-layer.4": 440.9627,
      "encoder_q-layer.5": 445.119,
      "encoder_q-layer.6": 502.4234,
      "encoder_q-layer.7": 579.3381,
      "encoder_q-layer.8": 671.99,
      "encoder_q-layer.9": 611.6899,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7347,
      "inbatch_pos_score": 1.459,
      "learning_rate": 7.277777777777778e-06,
      "loss": 3.1291,
      "norm_diff": 0.0686,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 972.9452,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7358,
      "query_norm": 1.5447,
      "queue_k_norm": 1.6072,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2136,
      "sent_len_1": 66.9535,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3013,
      "stdk": 0.0494,
      "stdq": 0.0461,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 86900
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1552,
      "doc_norm": 1.6033,
      "encoder_q-embeddings": 680.1686,
      "encoder_q-layer.0": 447.6671,
      "encoder_q-layer.1": 484.7368,
      "encoder_q-layer.10": 675.102,
      "encoder_q-layer.11": 1481.6161,
      "encoder_q-layer.2": 564.5352,
      "encoder_q-layer.3": 604.0179,
      "encoder_q-layer.4": 696.072,
      "encoder_q-layer.5": 695.9638,
      "encoder_q-layer.6": 782.225,
      "encoder_q-layer.7": 838.2415,
      "encoder_q-layer.8": 849.0966,
      "encoder_q-layer.9": 633.645,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7381,
      "inbatch_pos_score": 1.3896,
      "learning_rate": 7.222222222222222e-06,
      "loss": 3.1552,
      "norm_diff": 0.0845,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1176.6001,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7368,
      "query_norm": 1.5189,
      "queue_k_norm": 1.606,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1587,
      "sent_len_1": 67.0213,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.47,
      "stdk": 0.0489,
      "stdq": 0.0449,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87000
    },
    {
      "accuracy": 54.5898,
      "active_queue_size": 16384.0,
      "cl_loss": 3.152,
      "doc_norm": 1.6075,
      "encoder_q-embeddings": 801.3821,
      "encoder_q-layer.0": 546.8835,
      "encoder_q-layer.1": 674.1807,
      "encoder_q-layer.10": 619.283,
      "encoder_q-layer.11": 1441.53,
      "encoder_q-layer.2": 742.942,
      "encoder_q-layer.3": 864.0368,
      "encoder_q-layer.4": 1026.2726,
      "encoder_q-layer.5": 1014.359,
      "encoder_q-layer.6": 1019.268,
      "encoder_q-layer.7": 897.3403,
      "encoder_q-layer.8": 804.9567,
      "encoder_q-layer.9": 641.6282,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7352,
      "inbatch_pos_score": 1.4141,
      "learning_rate": 7.166666666666667e-06,
      "loss": 3.152,
      "norm_diff": 0.0731,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1336.9772,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7358,
      "query_norm": 1.5344,
      "queue_k_norm": 1.6068,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9336,
      "sent_len_1": 66.7137,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.68,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87100
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1303,
      "doc_norm": 1.608,
      "encoder_q-embeddings": 1520.2295,
      "encoder_q-layer.0": 1055.1656,
      "encoder_q-layer.1": 1116.0465,
      "encoder_q-layer.10": 637.6906,
      "encoder_q-layer.11": 1426.7794,
      "encoder_q-layer.2": 1360.5637,
      "encoder_q-layer.3": 1366.1183,
      "encoder_q-layer.4": 1508.4449,
      "encoder_q-layer.5": 1618.8055,
      "encoder_q-layer.6": 1601.4653,
      "encoder_q-layer.7": 1480.2795,
      "encoder_q-layer.8": 1054.395,
      "encoder_q-layer.9": 633.6291,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7382,
      "inbatch_pos_score": 1.4512,
      "learning_rate": 7.111111111111112e-06,
      "loss": 3.1303,
      "norm_diff": 0.0655,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2016.3317,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7383,
      "query_norm": 1.5425,
      "queue_k_norm": 1.6043,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9163,
      "sent_len_1": 66.7894,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6188,
      "stdk": 0.0491,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 87200
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1396,
      "doc_norm": 1.6032,
      "encoder_q-embeddings": 495.5591,
      "encoder_q-layer.0": 339.437,
      "encoder_q-layer.1": 366.5645,
      "encoder_q-layer.10": 651.4262,
      "encoder_q-layer.11": 1530.5125,
      "encoder_q-layer.2": 403.4842,
      "encoder_q-layer.3": 398.283,
      "encoder_q-layer.4": 413.7554,
      "encoder_q-layer.5": 420.3734,
      "encoder_q-layer.6": 486.7055,
      "encoder_q-layer.7": 592.6597,
      "encoder_q-layer.8": 716.5886,
      "encoder_q-layer.9": 630.8024,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7401,
      "inbatch_pos_score": 1.4033,
      "learning_rate": 7.055555555555556e-06,
      "loss": 3.1396,
      "norm_diff": 0.0649,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1019.4036,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7397,
      "query_norm": 1.5383,
      "queue_k_norm": 1.6076,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.945,
      "sent_len_1": 66.6747,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.31,
      "stdk": 0.0489,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87300
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1409,
      "doc_norm": 1.6047,
      "encoder_q-embeddings": 580.7345,
      "encoder_q-layer.0": 394.1264,
      "encoder_q-layer.1": 448.5361,
      "encoder_q-layer.10": 672.0183,
      "encoder_q-layer.11": 1507.646,
      "encoder_q-layer.2": 604.5476,
      "encoder_q-layer.3": 666.7718,
      "encoder_q-layer.4": 740.227,
      "encoder_q-layer.5": 715.7291,
      "encoder_q-layer.6": 750.6434,
      "encoder_q-layer.7": 801.6014,
      "encoder_q-layer.8": 858.2729,
      "encoder_q-layer.9": 703.2618,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7397,
      "inbatch_pos_score": 1.4238,
      "learning_rate": 7.000000000000001e-06,
      "loss": 3.1409,
      "norm_diff": 0.0595,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1189.7611,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7397,
      "query_norm": 1.5451,
      "queue_k_norm": 1.6081,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8833,
      "sent_len_1": 66.7482,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3775,
      "stdk": 0.049,
      "stdq": 0.046,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 87400
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1435,
      "doc_norm": 1.6147,
      "encoder_q-embeddings": 605.9847,
      "encoder_q-layer.0": 397.5941,
      "encoder_q-layer.1": 456.6804,
      "encoder_q-layer.10": 678.9271,
      "encoder_q-layer.11": 1561.4218,
      "encoder_q-layer.2": 530.9473,
      "encoder_q-layer.3": 552.7935,
      "encoder_q-layer.4": 587.2938,
      "encoder_q-layer.5": 606.896,
      "encoder_q-layer.6": 665.7611,
      "encoder_q-layer.7": 759.6003,
      "encoder_q-layer.8": 783.7995,
      "encoder_q-layer.9": 641.7421,
      "epoch": 0.85,
      "inbatch_neg_score": 0.7389,
      "inbatch_pos_score": 1.3906,
      "learning_rate": 6.944444444444445e-06,
      "loss": 3.1435,
      "norm_diff": 0.097,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1146.372,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7397,
      "query_norm": 1.5177,
      "queue_k_norm": 1.6091,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8881,
      "sent_len_1": 66.6261,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.28,
      "stdk": 0.0495,
      "stdq": 0.0449,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 87500
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1658,
      "doc_norm": 1.6033,
      "encoder_q-embeddings": 3105.9534,
      "encoder_q-layer.0": 2035.3473,
      "encoder_q-layer.1": 2554.6204,
      "encoder_q-layer.10": 630.8448,
      "encoder_q-layer.11": 1465.2303,
      "encoder_q-layer.2": 3193.5984,
      "encoder_q-layer.3": 3526.5605,
      "encoder_q-layer.4": 4581.519,
      "encoder_q-layer.5": 5148.3018,
      "encoder_q-layer.6": 5012.873,
      "encoder_q-layer.7": 4627.4595,
      "encoder_q-layer.8": 3371.6694,
      "encoder_q-layer.9": 1178.3876,
      "epoch": 0.86,
      "inbatch_neg_score": 0.741,
      "inbatch_pos_score": 1.4297,
      "learning_rate": 6.888888888888889e-06,
      "loss": 3.1658,
      "norm_diff": 0.0586,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5119.7419,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7402,
      "query_norm": 1.5446,
      "queue_k_norm": 1.6072,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0547,
      "sent_len_1": 66.6269,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.26,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87600
    },
    {
      "accuracy": 59.9609,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1437,
      "doc_norm": 1.608,
      "encoder_q-embeddings": 471.7063,
      "encoder_q-layer.0": 328.3268,
      "encoder_q-layer.1": 351.9655,
      "encoder_q-layer.10": 596.0638,
      "encoder_q-layer.11": 1397.3508,
      "encoder_q-layer.2": 404.6376,
      "encoder_q-layer.3": 437.3637,
      "encoder_q-layer.4": 463.6908,
      "encoder_q-layer.5": 485.0525,
      "encoder_q-layer.6": 530.1406,
      "encoder_q-layer.7": 610.8171,
      "encoder_q-layer.8": 666.4631,
      "encoder_q-layer.9": 580.4933,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7429,
      "inbatch_pos_score": 1.4512,
      "learning_rate": 6.833333333333333e-06,
      "loss": 3.1437,
      "norm_diff": 0.072,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 969.7448,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7417,
      "query_norm": 1.536,
      "queue_k_norm": 1.6081,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7977,
      "sent_len_1": 66.6602,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2375,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87700
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1277,
      "doc_norm": 1.6077,
      "encoder_q-embeddings": 579.3748,
      "encoder_q-layer.0": 404.4576,
      "encoder_q-layer.1": 427.475,
      "encoder_q-layer.10": 643.4861,
      "encoder_q-layer.11": 1508.744,
      "encoder_q-layer.2": 481.0397,
      "encoder_q-layer.3": 499.3064,
      "encoder_q-layer.4": 550.4728,
      "encoder_q-layer.5": 557.6143,
      "encoder_q-layer.6": 635.332,
      "encoder_q-layer.7": 699.4645,
      "encoder_q-layer.8": 757.9754,
      "encoder_q-layer.9": 629.8373,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7463,
      "inbatch_pos_score": 1.4189,
      "learning_rate": 6.777777777777779e-06,
      "loss": 3.1277,
      "norm_diff": 0.0812,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1091.4426,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7456,
      "query_norm": 1.5265,
      "queue_k_norm": 1.6084,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.065,
      "sent_len_1": 66.7875,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0575,
      "stdk": 0.0491,
      "stdq": 0.0451,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87800
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1199,
      "doc_norm": 1.6046,
      "encoder_q-embeddings": 472.5318,
      "encoder_q-layer.0": 325.5838,
      "encoder_q-layer.1": 353.7538,
      "encoder_q-layer.10": 621.599,
      "encoder_q-layer.11": 1456.1063,
      "encoder_q-layer.2": 402.6558,
      "encoder_q-layer.3": 432.1302,
      "encoder_q-layer.4": 470.6508,
      "encoder_q-layer.5": 470.5816,
      "encoder_q-layer.6": 564.9692,
      "encoder_q-layer.7": 650.0301,
      "encoder_q-layer.8": 718.6084,
      "encoder_q-layer.9": 595.4985,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7441,
      "inbatch_pos_score": 1.4092,
      "learning_rate": 6.722222222222223e-06,
      "loss": 3.1199,
      "norm_diff": 0.069,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 988.6872,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7441,
      "query_norm": 1.5357,
      "queue_k_norm": 1.6081,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.975,
      "sent_len_1": 66.7711,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1113,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 87900
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1257,
      "doc_norm": 1.6078,
      "encoder_q-embeddings": 878.1038,
      "encoder_q-layer.0": 630.6515,
      "encoder_q-layer.1": 695.8009,
      "encoder_q-layer.10": 683.9221,
      "encoder_q-layer.11": 1508.5696,
      "encoder_q-layer.2": 836.981,
      "encoder_q-layer.3": 930.632,
      "encoder_q-layer.4": 1109.8821,
      "encoder_q-layer.5": 1006.9022,
      "encoder_q-layer.6": 1138.4294,
      "encoder_q-layer.7": 1249.5524,
      "encoder_q-layer.8": 1090.1233,
      "encoder_q-layer.9": 656.6712,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7462,
      "inbatch_pos_score": 1.4141,
      "learning_rate": 6.666666666666667e-06,
      "loss": 3.1257,
      "norm_diff": 0.0773,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1528.3578,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7466,
      "query_norm": 1.5305,
      "queue_k_norm": 1.6073,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6635,
      "sent_len_1": 66.6499,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.995,
      "stdk": 0.0491,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88000
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1602,
      "doc_norm": 1.608,
      "encoder_q-embeddings": 470.7149,
      "encoder_q-layer.0": 311.1802,
      "encoder_q-layer.1": 340.7619,
      "encoder_q-layer.10": 608.9062,
      "encoder_q-layer.11": 1399.2073,
      "encoder_q-layer.2": 384.3034,
      "encoder_q-layer.3": 398.1575,
      "encoder_q-layer.4": 435.1395,
      "encoder_q-layer.5": 434.7696,
      "encoder_q-layer.6": 510.9488,
      "encoder_q-layer.7": 582.2281,
      "encoder_q-layer.8": 681.0874,
      "encoder_q-layer.9": 594.2287,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7455,
      "inbatch_pos_score": 1.4443,
      "learning_rate": 6.611111111111111e-06,
      "loss": 3.1602,
      "norm_diff": 0.0657,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 960.214,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7451,
      "query_norm": 1.5422,
      "queue_k_norm": 1.6076,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8163,
      "sent_len_1": 66.5858,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.485,
      "stdk": 0.0491,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88100
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1618,
      "doc_norm": 1.6065,
      "encoder_q-embeddings": 474.5019,
      "encoder_q-layer.0": 311.8459,
      "encoder_q-layer.1": 327.2066,
      "encoder_q-layer.10": 718.2875,
      "encoder_q-layer.11": 1500.6125,
      "encoder_q-layer.2": 368.0543,
      "encoder_q-layer.3": 388.2578,
      "encoder_q-layer.4": 414.6167,
      "encoder_q-layer.5": 445.9731,
      "encoder_q-layer.6": 518.7112,
      "encoder_q-layer.7": 593.9631,
      "encoder_q-layer.8": 685.4482,
      "encoder_q-layer.9": 629.9381,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7456,
      "inbatch_pos_score": 1.4277,
      "learning_rate": 6.555555555555556e-06,
      "loss": 3.1618,
      "norm_diff": 0.063,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 993.1077,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7456,
      "query_norm": 1.5434,
      "queue_k_norm": 1.6089,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7974,
      "sent_len_1": 66.7207,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7575,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88200
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1107,
      "doc_norm": 1.6066,
      "encoder_q-embeddings": 1079.6151,
      "encoder_q-layer.0": 732.6586,
      "encoder_q-layer.1": 814.6108,
      "encoder_q-layer.10": 1432.5499,
      "encoder_q-layer.11": 3075.9556,
      "encoder_q-layer.2": 948.8271,
      "encoder_q-layer.3": 1054.2454,
      "encoder_q-layer.4": 1193.5258,
      "encoder_q-layer.5": 1226.4131,
      "encoder_q-layer.6": 1396.0271,
      "encoder_q-layer.7": 1684.2476,
      "encoder_q-layer.8": 1575.0068,
      "encoder_q-layer.9": 1300.2169,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7465,
      "inbatch_pos_score": 1.4033,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 3.1107,
      "norm_diff": 0.0669,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2244.2018,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7471,
      "query_norm": 1.5396,
      "queue_k_norm": 1.6078,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6599,
      "sent_len_1": 66.6981,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.695,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88300
    },
    {
      "accuracy": 59.5703,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1499,
      "doc_norm": 1.6073,
      "encoder_q-embeddings": 2044.0859,
      "encoder_q-layer.0": 1411.9917,
      "encoder_q-layer.1": 1934.3342,
      "encoder_q-layer.10": 1219.4608,
      "encoder_q-layer.11": 2774.4194,
      "encoder_q-layer.2": 2247.5806,
      "encoder_q-layer.3": 2541.4116,
      "encoder_q-layer.4": 3261.8398,
      "encoder_q-layer.5": 3284.9001,
      "encoder_q-layer.6": 3668.0396,
      "encoder_q-layer.7": 3634.0305,
      "encoder_q-layer.8": 2882.9429,
      "encoder_q-layer.9": 1416.0897,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7494,
      "inbatch_pos_score": 1.4785,
      "learning_rate": 6.4444444444444445e-06,
      "loss": 3.1499,
      "norm_diff": 0.0598,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3835.1516,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.748,
      "query_norm": 1.5475,
      "queue_k_norm": 1.6096,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9113,
      "sent_len_1": 66.7798,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2425,
      "stdk": 0.049,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88400
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1258,
      "doc_norm": 1.6076,
      "encoder_q-embeddings": 1540.4893,
      "encoder_q-layer.0": 1118.3387,
      "encoder_q-layer.1": 1160.9492,
      "encoder_q-layer.10": 1252.0814,
      "encoder_q-layer.11": 3020.0864,
      "encoder_q-layer.2": 1372.4899,
      "encoder_q-layer.3": 1386.7609,
      "encoder_q-layer.4": 1390.2485,
      "encoder_q-layer.5": 1463.2375,
      "encoder_q-layer.6": 1560.5322,
      "encoder_q-layer.7": 1552.5322,
      "encoder_q-layer.8": 1507.5658,
      "encoder_q-layer.9": 1283.0413,
      "epoch": 0.86,
      "inbatch_neg_score": 0.7501,
      "inbatch_pos_score": 1.418,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 3.1258,
      "norm_diff": 0.0644,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2444.2945,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7505,
      "query_norm": 1.5432,
      "queue_k_norm": 1.6103,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9589,
      "sent_len_1": 66.9719,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4412,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88500
    },
    {
      "accuracy": 53.6133,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1597,
      "doc_norm": 1.6049,
      "encoder_q-embeddings": 1081.8311,
      "encoder_q-layer.0": 730.4937,
      "encoder_q-layer.1": 784.7132,
      "encoder_q-layer.10": 1295.6649,
      "encoder_q-layer.11": 3039.2781,
      "encoder_q-layer.2": 906.4891,
      "encoder_q-layer.3": 970.6371,
      "encoder_q-layer.4": 1040.3593,
      "encoder_q-layer.5": 1106.0332,
      "encoder_q-layer.6": 1177.3824,
      "encoder_q-layer.7": 1361.3596,
      "encoder_q-layer.8": 1433.6631,
      "encoder_q-layer.9": 1261.8667,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7504,
      "inbatch_pos_score": 1.4102,
      "learning_rate": 6.333333333333334e-06,
      "loss": 3.1597,
      "norm_diff": 0.0781,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2119.8818,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.75,
      "query_norm": 1.5268,
      "queue_k_norm": 1.609,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2243,
      "sent_len_1": 66.8779,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0362,
      "stdk": 0.0489,
      "stdq": 0.0451,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88600
    },
    {
      "accuracy": 55.1758,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1371,
      "doc_norm": 1.6082,
      "encoder_q-embeddings": 1365.8535,
      "encoder_q-layer.0": 905.0873,
      "encoder_q-layer.1": 1073.5514,
      "encoder_q-layer.10": 1305.0287,
      "encoder_q-layer.11": 3105.4036,
      "encoder_q-layer.2": 1208.705,
      "encoder_q-layer.3": 1373.917,
      "encoder_q-layer.4": 1468.8809,
      "encoder_q-layer.5": 1501.7742,
      "encoder_q-layer.6": 1550.1819,
      "encoder_q-layer.7": 1593.5337,
      "encoder_q-layer.8": 1471.7058,
      "encoder_q-layer.9": 1283.6974,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7521,
      "inbatch_pos_score": 1.4111,
      "learning_rate": 6.277777777777778e-06,
      "loss": 3.1371,
      "norm_diff": 0.0786,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2380.9035,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.751,
      "query_norm": 1.5296,
      "queue_k_norm": 1.6089,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0195,
      "sent_len_1": 66.9049,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.7688,
      "stdk": 0.049,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88700
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1441,
      "doc_norm": 1.61,
      "encoder_q-embeddings": 1104.2734,
      "encoder_q-layer.0": 744.5689,
      "encoder_q-layer.1": 822.3639,
      "encoder_q-layer.10": 1298.8982,
      "encoder_q-layer.11": 2921.2544,
      "encoder_q-layer.2": 957.4181,
      "encoder_q-layer.3": 997.5686,
      "encoder_q-layer.4": 1065.5387,
      "encoder_q-layer.5": 1131.7975,
      "encoder_q-layer.6": 1227.0737,
      "encoder_q-layer.7": 1348.6681,
      "encoder_q-layer.8": 1389.7007,
      "encoder_q-layer.9": 1210.75,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7521,
      "inbatch_pos_score": 1.4609,
      "learning_rate": 6.222222222222222e-06,
      "loss": 3.1441,
      "norm_diff": 0.0683,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2116.7333,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7515,
      "query_norm": 1.5417,
      "queue_k_norm": 1.6098,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8401,
      "sent_len_1": 66.818,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.5387,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 88800
    },
    {
      "accuracy": 55.0781,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1265,
      "doc_norm": 1.6107,
      "encoder_q-embeddings": 41690.2227,
      "encoder_q-layer.0": 27967.707,
      "encoder_q-layer.1": 34433.5703,
      "encoder_q-layer.10": 1592.0311,
      "encoder_q-layer.11": 3070.9932,
      "encoder_q-layer.2": 39765.957,
      "encoder_q-layer.3": 43374.375,
      "encoder_q-layer.4": 41726.3672,
      "encoder_q-layer.5": 40090.0664,
      "encoder_q-layer.6": 44450.375,
      "encoder_q-layer.7": 40365.793,
      "encoder_q-layer.8": 18821.3906,
      "encoder_q-layer.9": 4449.2349,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7533,
      "inbatch_pos_score": 1.4385,
      "learning_rate": 6.166666666666667e-06,
      "loss": 3.1265,
      "norm_diff": 0.0533,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 51842.0938,
      "preclip_grad_norm_avg": 0.0005,
      "q@queue_neg_score": 0.7524,
      "query_norm": 1.5574,
      "queue_k_norm": 1.6091,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8371,
      "sent_len_1": 66.6373,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8613,
      "stdk": 0.0491,
      "stdq": 0.0464,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 88900
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1511,
      "doc_norm": 1.6167,
      "encoder_q-embeddings": 1102.7654,
      "encoder_q-layer.0": 729.3633,
      "encoder_q-layer.1": 824.5349,
      "encoder_q-layer.10": 1234.9189,
      "encoder_q-layer.11": 3014.4922,
      "encoder_q-layer.2": 911.8994,
      "encoder_q-layer.3": 1024.7397,
      "encoder_q-layer.4": 1108.9088,
      "encoder_q-layer.5": 1212.2922,
      "encoder_q-layer.6": 1246.8088,
      "encoder_q-layer.7": 1322.6542,
      "encoder_q-layer.8": 1428.0056,
      "encoder_q-layer.9": 1241.5682,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7547,
      "inbatch_pos_score": 1.4355,
      "learning_rate": 6.111111111111111e-06,
      "loss": 3.1511,
      "norm_diff": 0.0707,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2131.6035,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7549,
      "query_norm": 1.5459,
      "queue_k_norm": 1.6092,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8164,
      "sent_len_1": 66.6215,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5412,
      "stdk": 0.0494,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 89000
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1411,
      "doc_norm": 1.6141,
      "encoder_q-embeddings": 1219.2277,
      "encoder_q-layer.0": 828.1174,
      "encoder_q-layer.1": 946.3163,
      "encoder_q-layer.10": 1293.3447,
      "encoder_q-layer.11": 2780.7686,
      "encoder_q-layer.2": 1163.0309,
      "encoder_q-layer.3": 1257.6085,
      "encoder_q-layer.4": 1324.3906,
      "encoder_q-layer.5": 1387.0599,
      "encoder_q-layer.6": 1467.6108,
      "encoder_q-layer.7": 1640.7971,
      "encoder_q-layer.8": 1623.3533,
      "encoder_q-layer.9": 1269.2028,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7555,
      "inbatch_pos_score": 1.4453,
      "learning_rate": 6.055555555555556e-06,
      "loss": 3.1411,
      "norm_diff": 0.0637,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2258.2792,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7549,
      "query_norm": 1.5504,
      "queue_k_norm": 1.6107,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9696,
      "sent_len_1": 66.8102,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.865,
      "stdk": 0.0492,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89100
    },
    {
      "accuracy": 58.0078,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1549,
      "doc_norm": 1.6074,
      "encoder_q-embeddings": 829.7554,
      "encoder_q-layer.0": 558.7236,
      "encoder_q-layer.1": 595.0117,
      "encoder_q-layer.10": 1174.3693,
      "encoder_q-layer.11": 2799.1196,
      "encoder_q-layer.2": 663.5145,
      "encoder_q-layer.3": 700.9609,
      "encoder_q-layer.4": 717.1572,
      "encoder_q-layer.5": 756.9156,
      "encoder_q-layer.6": 881.7628,
      "encoder_q-layer.7": 1041.6198,
      "encoder_q-layer.8": 1250.2991,
      "encoder_q-layer.9": 1149.0341,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7574,
      "inbatch_pos_score": 1.4541,
      "learning_rate": 6e-06,
      "loss": 3.1549,
      "norm_diff": 0.0654,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1807.1657,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7573,
      "query_norm": 1.542,
      "queue_k_norm": 1.6105,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8131,
      "sent_len_1": 66.7669,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6975,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89200
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1517,
      "doc_norm": 1.6176,
      "encoder_q-embeddings": 1408.931,
      "encoder_q-layer.0": 965.8429,
      "encoder_q-layer.1": 1117.1442,
      "encoder_q-layer.10": 1356.0736,
      "encoder_q-layer.11": 3056.4814,
      "encoder_q-layer.2": 1300.9963,
      "encoder_q-layer.3": 1398.3779,
      "encoder_q-layer.4": 1382.3214,
      "encoder_q-layer.5": 1509.976,
      "encoder_q-layer.6": 1713.9302,
      "encoder_q-layer.7": 1826.7628,
      "encoder_q-layer.8": 1861.7556,
      "encoder_q-layer.9": 1518.3682,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7578,
      "inbatch_pos_score": 1.459,
      "learning_rate": 5.944444444444445e-06,
      "loss": 3.1517,
      "norm_diff": 0.058,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2504.2066,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7578,
      "query_norm": 1.5596,
      "queue_k_norm": 1.6122,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7234,
      "sent_len_1": 66.9343,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1662,
      "stdk": 0.0493,
      "stdq": 0.0464,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89300
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1381,
      "doc_norm": 1.6202,
      "encoder_q-embeddings": 1034.99,
      "encoder_q-layer.0": 662.1221,
      "encoder_q-layer.1": 682.8,
      "encoder_q-layer.10": 1494.1919,
      "encoder_q-layer.11": 3231.0796,
      "encoder_q-layer.2": 750.0168,
      "encoder_q-layer.3": 802.4109,
      "encoder_q-layer.4": 865.1697,
      "encoder_q-layer.5": 914.944,
      "encoder_q-layer.6": 1043.3044,
      "encoder_q-layer.7": 1260.8402,
      "encoder_q-layer.8": 1526.3927,
      "encoder_q-layer.9": 1363.0908,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7579,
      "inbatch_pos_score": 1.4414,
      "learning_rate": 5.888888888888889e-06,
      "loss": 3.1381,
      "norm_diff": 0.0759,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2094.7234,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7598,
      "query_norm": 1.5443,
      "queue_k_norm": 1.6114,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0439,
      "sent_len_1": 66.705,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9525,
      "stdk": 0.0495,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89400
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.144,
      "doc_norm": 1.6152,
      "encoder_q-embeddings": 2905.2395,
      "encoder_q-layer.0": 2018.2566,
      "encoder_q-layer.1": 1936.736,
      "encoder_q-layer.10": 1434.5579,
      "encoder_q-layer.11": 3027.7832,
      "encoder_q-layer.2": 2318.3616,
      "encoder_q-layer.3": 2566.6243,
      "encoder_q-layer.4": 2812.4004,
      "encoder_q-layer.5": 2921.6021,
      "encoder_q-layer.6": 3395.3992,
      "encoder_q-layer.7": 3420.8196,
      "encoder_q-layer.8": 3588.8948,
      "encoder_q-layer.9": 2103.3892,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7613,
      "inbatch_pos_score": 1.4346,
      "learning_rate": 5.833333333333334e-06,
      "loss": 3.144,
      "norm_diff": 0.0771,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4235.2395,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7607,
      "query_norm": 1.5381,
      "queue_k_norm": 1.6115,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0105,
      "sent_len_1": 66.7953,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3137,
      "stdk": 0.0492,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89500
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1475,
      "doc_norm": 1.6084,
      "encoder_q-embeddings": 971.1769,
      "encoder_q-layer.0": 648.593,
      "encoder_q-layer.1": 692.6425,
      "encoder_q-layer.10": 1310.0708,
      "encoder_q-layer.11": 2946.8091,
      "encoder_q-layer.2": 793.0953,
      "encoder_q-layer.3": 832.8814,
      "encoder_q-layer.4": 896.5701,
      "encoder_q-layer.5": 943.4767,
      "encoder_q-layer.6": 1028.0768,
      "encoder_q-layer.7": 1224.8424,
      "encoder_q-layer.8": 1372.3284,
      "encoder_q-layer.9": 1242.8688,
      "epoch": 0.87,
      "inbatch_neg_score": 0.7606,
      "inbatch_pos_score": 1.4219,
      "learning_rate": 5.777777777777778e-06,
      "loss": 3.1475,
      "norm_diff": 0.0698,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2013.1303,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7607,
      "query_norm": 1.5387,
      "queue_k_norm": 1.6118,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0147,
      "sent_len_1": 66.7429,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.7175,
      "stdk": 0.0489,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89600
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1512,
      "doc_norm": 1.6169,
      "encoder_q-embeddings": 1269.8798,
      "encoder_q-layer.0": 883.5316,
      "encoder_q-layer.1": 1050.9283,
      "encoder_q-layer.10": 1274.8411,
      "encoder_q-layer.11": 2981.7571,
      "encoder_q-layer.2": 1219.8357,
      "encoder_q-layer.3": 1316.923,
      "encoder_q-layer.4": 1429.4182,
      "encoder_q-layer.5": 1447.7523,
      "encoder_q-layer.6": 1619.2318,
      "encoder_q-layer.7": 1729.8859,
      "encoder_q-layer.8": 1819.0326,
      "encoder_q-layer.9": 1446.3291,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7615,
      "inbatch_pos_score": 1.4629,
      "learning_rate": 5.722222222222223e-06,
      "loss": 3.1512,
      "norm_diff": 0.0633,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2445.703,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7622,
      "query_norm": 1.5537,
      "queue_k_norm": 1.6123,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8664,
      "sent_len_1": 66.6214,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4512,
      "stdk": 0.0493,
      "stdq": 0.0461,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89700
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1159,
      "doc_norm": 1.6139,
      "encoder_q-embeddings": 1041.223,
      "encoder_q-layer.0": 697.3228,
      "encoder_q-layer.1": 764.0959,
      "encoder_q-layer.10": 1532.272,
      "encoder_q-layer.11": 2986.0205,
      "encoder_q-layer.2": 882.8315,
      "encoder_q-layer.3": 935.0059,
      "encoder_q-layer.4": 1034.9604,
      "encoder_q-layer.5": 1083.231,
      "encoder_q-layer.6": 1280.6273,
      "encoder_q-layer.7": 1442.9371,
      "encoder_q-layer.8": 1523.8799,
      "encoder_q-layer.9": 1345.8608,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7637,
      "inbatch_pos_score": 1.457,
      "learning_rate": 5.666666666666667e-06,
      "loss": 3.1159,
      "norm_diff": 0.0558,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2120.3707,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7632,
      "query_norm": 1.5581,
      "queue_k_norm": 1.6139,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0539,
      "sent_len_1": 66.7329,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9375,
      "stdk": 0.0491,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89800
    },
    {
      "accuracy": 52.0508,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1381,
      "doc_norm": 1.6144,
      "encoder_q-embeddings": 1250.0187,
      "encoder_q-layer.0": 905.8939,
      "encoder_q-layer.1": 1032.6523,
      "encoder_q-layer.10": 1426.951,
      "encoder_q-layer.11": 3044.3823,
      "encoder_q-layer.2": 1178.998,
      "encoder_q-layer.3": 1192.6482,
      "encoder_q-layer.4": 1226.8818,
      "encoder_q-layer.5": 1214.5807,
      "encoder_q-layer.6": 1195.2158,
      "encoder_q-layer.7": 1344.2297,
      "encoder_q-layer.8": 1584.1763,
      "encoder_q-layer.9": 1369.239,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7653,
      "inbatch_pos_score": 1.4199,
      "learning_rate": 5.611111111111112e-06,
      "loss": 3.1381,
      "norm_diff": 0.0674,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2275.8723,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7651,
      "query_norm": 1.547,
      "queue_k_norm": 1.6128,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7422,
      "sent_len_1": 66.6328,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.165,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 89900
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1456,
      "doc_norm": 1.6171,
      "encoder_q-embeddings": 1654.5474,
      "encoder_q-layer.0": 1218.7119,
      "encoder_q-layer.1": 1426.957,
      "encoder_q-layer.10": 1262.2678,
      "encoder_q-layer.11": 2977.5002,
      "encoder_q-layer.2": 1691.9941,
      "encoder_q-layer.3": 1745.3152,
      "encoder_q-layer.4": 1796.0414,
      "encoder_q-layer.5": 1819.9469,
      "encoder_q-layer.6": 1683.3077,
      "encoder_q-layer.7": 1578.0001,
      "encoder_q-layer.8": 1561.6638,
      "encoder_q-layer.9": 1280.0563,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7653,
      "inbatch_pos_score": 1.4707,
      "learning_rate": 5.555555555555556e-06,
      "loss": 3.1456,
      "norm_diff": 0.06,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2647.0022,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7661,
      "query_norm": 1.5571,
      "queue_k_norm": 1.6163,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9086,
      "sent_len_1": 66.8065,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.9675,
      "stdk": 0.0493,
      "stdq": 0.0461,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 90000
    },
    {
      "dev_runtime": 37.6562,
      "dev_samples_per_second": 1.7,
      "dev_steps_per_second": 0.027,
      "epoch": 0.88,
      "step": 90000,
      "test_accuracy": 94.54345703125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3078896701335907,
      "test_doc_norm": 1.607322335243225,
      "test_inbatch_neg_score": 1.1371493339538574,
      "test_inbatch_pos_score": 2.1192989349365234,
      "test_loss": 0.3078896701335907,
      "test_loss_align": 0.9989704489707947,
      "test_loss_unif": 2.9055066108703613,
      "test_loss_unif_q@queue": 2.9055068492889404,
      "test_norm_diff": 0.015200823545455933,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.7489210367202759,
      "test_query_norm": 1.6220368146896362,
      "test_queue_k_norm": 1.6164119243621826,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04331135004758835,
      "test_stdq": 0.043481796979904175,
      "test_stdqueue_k": 0.049305226653814316,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 37.6562,
      "dev_samples_per_second": 1.7,
      "dev_steps_per_second": 0.027,
      "epoch": 0.88,
      "eval_beir-arguana_ndcg@10": 0.38415,
      "eval_beir-arguana_recall@10": 0.6522,
      "eval_beir-arguana_recall@100": 0.93457,
      "eval_beir-arguana_recall@20": 0.77881,
      "eval_beir-avg_ndcg@10": 0.37948008333333333,
      "eval_beir-avg_recall@10": 0.4468521666666666,
      "eval_beir-avg_recall@100": 0.6270814166666667,
      "eval_beir-avg_recall@20": 0.50763825,
      "eval_beir-cqadupstack_ndcg@10": 0.2709708333333333,
      "eval_beir-cqadupstack_recall@10": 0.3618116666666667,
      "eval_beir-cqadupstack_recall@100": 0.5872541666666666,
      "eval_beir-cqadupstack_recall@20": 0.4275625,
      "eval_beir-fiqa_ndcg@10": 0.22238,
      "eval_beir-fiqa_recall@10": 0.276,
      "eval_beir-fiqa_recall@100": 0.53444,
      "eval_beir-fiqa_recall@20": 0.34488,
      "eval_beir-nfcorpus_ndcg@10": 0.30283,
      "eval_beir-nfcorpus_recall@10": 0.1499,
      "eval_beir-nfcorpus_recall@100": 0.27555,
      "eval_beir-nfcorpus_recall@20": 0.1801,
      "eval_beir-nq_ndcg@10": 0.27722,
      "eval_beir-nq_recall@10": 0.45614,
      "eval_beir-nq_recall@100": 0.79473,
      "eval_beir-nq_recall@20": 0.57546,
      "eval_beir-quora_ndcg@10": 0.82672,
      "eval_beir-quora_recall@10": 0.91859,
      "eval_beir-quora_recall@100": 0.98615,
      "eval_beir-quora_recall@20": 0.95147,
      "eval_beir-scidocs_ndcg@10": 0.14694,
      "eval_beir-scidocs_recall@10": 0.15473,
      "eval_beir-scidocs_recall@100": 0.35287,
      "eval_beir-scidocs_recall@20": 0.21193,
      "eval_beir-scifact_ndcg@10": 0.63432,
      "eval_beir-scifact_recall@10": 0.78733,
      "eval_beir-scifact_recall@100": 0.92822,
      "eval_beir-scifact_recall@20": 0.83511,
      "eval_beir-trec-covid_ndcg@10": 0.53785,
      "eval_beir-trec-covid_recall@10": 0.57,
      "eval_beir-trec-covid_recall@100": 0.4336,
      "eval_beir-trec-covid_recall@20": 0.551,
      "eval_beir-webis-touche2020_ndcg@10": 0.19142,
      "eval_beir-webis-touche2020_recall@10": 0.14182,
      "eval_beir-webis-touche2020_recall@100": 0.44343,
      "eval_beir-webis-touche2020_recall@20": 0.22006,
      "eval_senteval-avg_sts": 0.7251337697300739,
      "eval_senteval-sickr_spearman": 0.6928772647812322,
      "eval_senteval-stsb_spearman": 0.7573902746789156,
      "step": 90000,
      "test_accuracy": 94.54345703125,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.3078896701335907,
      "test_doc_norm": 1.607322335243225,
      "test_inbatch_neg_score": 1.1371493339538574,
      "test_inbatch_pos_score": 2.1192989349365234,
      "test_loss": 0.3078896701335907,
      "test_loss_align": 0.9989704489707947,
      "test_loss_unif": 2.9055066108703613,
      "test_loss_unif_q@queue": 2.9055068492889404,
      "test_norm_diff": 0.015200823545455933,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.7489210367202759,
      "test_query_norm": 1.6220368146896362,
      "test_queue_k_norm": 1.6164119243621826,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.04331135004758835,
      "test_stdq": 0.043481796979904175,
      "test_stdqueue_k": 0.049305226653814316,
      "test_stdqueue_q": 0.0
    },
    {
      "accuracy": 51.4648,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1306,
      "doc_norm": 1.6161,
      "encoder_q-embeddings": 2474.6719,
      "encoder_q-layer.0": 1676.6241,
      "encoder_q-layer.1": 1931.1167,
      "encoder_q-layer.10": 1321.4515,
      "encoder_q-layer.11": 3125.4355,
      "encoder_q-layer.2": 2330.282,
      "encoder_q-layer.3": 2577.4187,
      "encoder_q-layer.4": 2735.4917,
      "encoder_q-layer.5": 2743.9937,
      "encoder_q-layer.6": 2655.9907,
      "encoder_q-layer.7": 2615.9817,
      "encoder_q-layer.8": 2792.179,
      "encoder_q-layer.9": 1873.6127,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7686,
      "inbatch_pos_score": 1.4199,
      "learning_rate": 5.500000000000001e-06,
      "loss": 3.1306,
      "norm_diff": 0.0731,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3700.5862,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7676,
      "query_norm": 1.543,
      "queue_k_norm": 1.6145,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8351,
      "sent_len_1": 66.5841,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6163,
      "stdk": 0.0492,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90100
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1597,
      "doc_norm": 1.6189,
      "encoder_q-embeddings": 1085.6447,
      "encoder_q-layer.0": 706.3303,
      "encoder_q-layer.1": 797.4725,
      "encoder_q-layer.10": 1307.3092,
      "encoder_q-layer.11": 2850.9851,
      "encoder_q-layer.2": 905.741,
      "encoder_q-layer.3": 976.8034,
      "encoder_q-layer.4": 1014.8611,
      "encoder_q-layer.5": 1124.212,
      "encoder_q-layer.6": 1260.9174,
      "encoder_q-layer.7": 1407.193,
      "encoder_q-layer.8": 1458.8956,
      "encoder_q-layer.9": 1211.7335,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7667,
      "inbatch_pos_score": 1.46,
      "learning_rate": 5.444444444444445e-06,
      "loss": 3.1597,
      "norm_diff": 0.0704,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2052.8781,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7666,
      "query_norm": 1.5485,
      "queue_k_norm": 1.614,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0948,
      "sent_len_1": 66.8016,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4625,
      "stdk": 0.0493,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90200
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1257,
      "doc_norm": 1.6135,
      "encoder_q-embeddings": 1832.2013,
      "encoder_q-layer.0": 1234.4771,
      "encoder_q-layer.1": 1271.0392,
      "encoder_q-layer.10": 2511.8618,
      "encoder_q-layer.11": 5842.2686,
      "encoder_q-layer.2": 1448.0002,
      "encoder_q-layer.3": 1480.9048,
      "encoder_q-layer.4": 1609.2397,
      "encoder_q-layer.5": 1714.084,
      "encoder_q-layer.6": 1934.8002,
      "encoder_q-layer.7": 2212.3584,
      "encoder_q-layer.8": 2807.4512,
      "encoder_q-layer.9": 2451.7629,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7699,
      "inbatch_pos_score": 1.4434,
      "learning_rate": 5.388888888888889e-06,
      "loss": 3.1257,
      "norm_diff": 0.0632,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3893.3861,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7695,
      "query_norm": 1.5503,
      "queue_k_norm": 1.6151,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8803,
      "sent_len_1": 66.7331,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5875,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90300
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1583,
      "doc_norm": 1.6141,
      "encoder_q-embeddings": 2155.6863,
      "encoder_q-layer.0": 1436.1663,
      "encoder_q-layer.1": 1538.2712,
      "encoder_q-layer.10": 2552.8826,
      "encoder_q-layer.11": 5752.5479,
      "encoder_q-layer.2": 1743.8815,
      "encoder_q-layer.3": 1851.6606,
      "encoder_q-layer.4": 1979.2593,
      "encoder_q-layer.5": 2121.4082,
      "encoder_q-layer.6": 2284.9827,
      "encoder_q-layer.7": 2525.093,
      "encoder_q-layer.8": 2734.0586,
      "encoder_q-layer.9": 2416.3416,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7701,
      "inbatch_pos_score": 1.4854,
      "learning_rate": 5.333333333333334e-06,
      "loss": 3.1583,
      "norm_diff": 0.0545,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4035.8426,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7695,
      "query_norm": 1.5596,
      "queue_k_norm": 1.6153,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1055,
      "sent_len_1": 66.8693,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0538,
      "stdk": 0.049,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90400
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1466,
      "doc_norm": 1.6223,
      "encoder_q-embeddings": 5723.5454,
      "encoder_q-layer.0": 4043.6824,
      "encoder_q-layer.1": 4312.6855,
      "encoder_q-layer.10": 2478.1914,
      "encoder_q-layer.11": 5860.7354,
      "encoder_q-layer.2": 4849.5391,
      "encoder_q-layer.3": 4708.5063,
      "encoder_q-layer.4": 4896.5771,
      "encoder_q-layer.5": 5028.4321,
      "encoder_q-layer.6": 4276.5718,
      "encoder_q-layer.7": 4094.7104,
      "encoder_q-layer.8": 3546.2283,
      "encoder_q-layer.9": 2412.8381,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7713,
      "inbatch_pos_score": 1.4639,
      "learning_rate": 5.277777777777778e-06,
      "loss": 3.1466,
      "norm_diff": 0.0687,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 6884.9869,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.771,
      "query_norm": 1.5536,
      "queue_k_norm": 1.6159,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0198,
      "sent_len_1": 66.7664,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2237,
      "stdk": 0.0494,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90500
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1704,
      "doc_norm": 1.6199,
      "encoder_q-embeddings": 2908.2751,
      "encoder_q-layer.0": 1914.6709,
      "encoder_q-layer.1": 2212.1458,
      "encoder_q-layer.10": 2462.7739,
      "encoder_q-layer.11": 5710.9854,
      "encoder_q-layer.2": 2544.5913,
      "encoder_q-layer.3": 2751.8418,
      "encoder_q-layer.4": 3025.1614,
      "encoder_q-layer.5": 2956.1902,
      "encoder_q-layer.6": 3206.3508,
      "encoder_q-layer.7": 3290.9204,
      "encoder_q-layer.8": 3284.6006,
      "encoder_q-layer.9": 2439.3374,
      "epoch": 0.88,
      "inbatch_neg_score": 0.7733,
      "inbatch_pos_score": 1.4609,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 3.1704,
      "norm_diff": 0.0812,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4826.6592,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7729,
      "query_norm": 1.5386,
      "queue_k_norm": 1.6179,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1467,
      "sent_len_1": 66.7441,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0575,
      "stdk": 0.0493,
      "stdq": 0.0453,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 90600
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1327,
      "doc_norm": 1.6223,
      "encoder_q-embeddings": 2065.4966,
      "encoder_q-layer.0": 1409.8958,
      "encoder_q-layer.1": 1499.3433,
      "encoder_q-layer.10": 2461.9434,
      "encoder_q-layer.11": 5820.4233,
      "encoder_q-layer.2": 1780.1064,
      "encoder_q-layer.3": 1902.5603,
      "encoder_q-layer.4": 1966.05,
      "encoder_q-layer.5": 2055.5439,
      "encoder_q-layer.6": 2282.2839,
      "encoder_q-layer.7": 2589.7732,
      "encoder_q-layer.8": 2837.0588,
      "encoder_q-layer.9": 2446.5437,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7729,
      "inbatch_pos_score": 1.4619,
      "learning_rate": 5.166666666666667e-06,
      "loss": 3.1327,
      "norm_diff": 0.0753,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4060.6259,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.772,
      "query_norm": 1.547,
      "queue_k_norm": 1.6159,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0577,
      "sent_len_1": 66.9125,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1175,
      "stdk": 0.0494,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90700
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1228,
      "doc_norm": 1.6168,
      "encoder_q-embeddings": 1995.9663,
      "encoder_q-layer.0": 1323.5575,
      "encoder_q-layer.1": 1371.2231,
      "encoder_q-layer.10": 2680.8953,
      "encoder_q-layer.11": 5721.3535,
      "encoder_q-layer.2": 1539.8492,
      "encoder_q-layer.3": 1635.3392,
      "encoder_q-layer.4": 1751.0859,
      "encoder_q-layer.5": 1828.8677,
      "encoder_q-layer.6": 2086.1147,
      "encoder_q-layer.7": 2521.9558,
      "encoder_q-layer.8": 2657.1045,
      "encoder_q-layer.9": 2470.7744,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7751,
      "inbatch_pos_score": 1.4521,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 3.1228,
      "norm_diff": 0.07,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3939.1895,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7744,
      "query_norm": 1.5468,
      "queue_k_norm": 1.6177,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1393,
      "sent_len_1": 66.8646,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.6238,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 90800
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1337,
      "doc_norm": 1.61,
      "encoder_q-embeddings": 2218.2764,
      "encoder_q-layer.0": 1465.0325,
      "encoder_q-layer.1": 1630.7064,
      "encoder_q-layer.10": 2639.9194,
      "encoder_q-layer.11": 5888.957,
      "encoder_q-layer.2": 1819.0635,
      "encoder_q-layer.3": 2038.9138,
      "encoder_q-layer.4": 2335.5818,
      "encoder_q-layer.5": 2287.9797,
      "encoder_q-layer.6": 2552.4084,
      "encoder_q-layer.7": 2892.3948,
      "encoder_q-layer.8": 2842.8503,
      "encoder_q-layer.9": 2525.8157,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7775,
      "inbatch_pos_score": 1.4336,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 3.1337,
      "norm_diff": 0.0619,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4297.0332,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7764,
      "query_norm": 1.5481,
      "queue_k_norm": 1.6178,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.903,
      "sent_len_1": 67.021,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.8587,
      "stdk": 0.0488,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 90900
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1528,
      "doc_norm": 1.6167,
      "encoder_q-embeddings": 2010.4141,
      "encoder_q-layer.0": 1406.6013,
      "encoder_q-layer.1": 1504.3971,
      "encoder_q-layer.10": 2618.2209,
      "encoder_q-layer.11": 6197.2827,
      "encoder_q-layer.2": 1804.118,
      "encoder_q-layer.3": 1907.9352,
      "encoder_q-layer.4": 2100.4429,
      "encoder_q-layer.5": 2039.9465,
      "encoder_q-layer.6": 2303.9175,
      "encoder_q-layer.7": 2590.9717,
      "encoder_q-layer.8": 2779.9929,
      "encoder_q-layer.9": 2594.2358,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7762,
      "inbatch_pos_score": 1.4453,
      "learning_rate": 5e-06,
      "loss": 3.1528,
      "norm_diff": 0.0656,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4272.8183,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7769,
      "query_norm": 1.5511,
      "queue_k_norm": 1.6168,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0071,
      "sent_len_1": 66.4738,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.7975,
      "stdk": 0.0491,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91000
    },
    {
      "accuracy": 56.543,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1259,
      "doc_norm": 1.6149,
      "encoder_q-embeddings": 1082.7233,
      "encoder_q-layer.0": 740.4307,
      "encoder_q-layer.1": 833.2512,
      "encoder_q-layer.10": 1460.5959,
      "encoder_q-layer.11": 3040.0374,
      "encoder_q-layer.2": 949.7389,
      "encoder_q-layer.3": 1001.2498,
      "encoder_q-layer.4": 1095.3893,
      "encoder_q-layer.5": 1217.3209,
      "encoder_q-layer.6": 1394.2822,
      "encoder_q-layer.7": 1687.49,
      "encoder_q-layer.8": 1770.4647,
      "encoder_q-layer.9": 1327.8701,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7785,
      "inbatch_pos_score": 1.4512,
      "learning_rate": 4.9444444444444444e-06,
      "loss": 3.1259,
      "norm_diff": 0.0675,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2264.5312,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7773,
      "query_norm": 1.5474,
      "queue_k_norm": 1.6166,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1216,
      "sent_len_1": 66.8488,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1875,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91100
    },
    {
      "accuracy": 53.418,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1257,
      "doc_norm": 1.6166,
      "encoder_q-embeddings": 918.7981,
      "encoder_q-layer.0": 612.0817,
      "encoder_q-layer.1": 664.6818,
      "encoder_q-layer.10": 1259.2914,
      "encoder_q-layer.11": 3092.177,
      "encoder_q-layer.2": 723.8133,
      "encoder_q-layer.3": 755.3939,
      "encoder_q-layer.4": 805.8848,
      "encoder_q-layer.5": 844.566,
      "encoder_q-layer.6": 994.2491,
      "encoder_q-layer.7": 1127.2104,
      "encoder_q-layer.8": 1353.5579,
      "encoder_q-layer.9": 1237.8169,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7807,
      "inbatch_pos_score": 1.4199,
      "learning_rate": 4.888888888888889e-06,
      "loss": 3.1257,
      "norm_diff": 0.0812,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2014.5535,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7803,
      "query_norm": 1.5353,
      "queue_k_norm": 1.6175,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1308,
      "sent_len_1": 66.7059,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.0175,
      "stdk": 0.049,
      "stdq": 0.045,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91200
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1518,
      "doc_norm": 1.6161,
      "encoder_q-embeddings": 1018.4268,
      "encoder_q-layer.0": 658.7505,
      "encoder_q-layer.1": 689.6532,
      "encoder_q-layer.10": 1246.1846,
      "encoder_q-layer.11": 3056.9487,
      "encoder_q-layer.2": 746.248,
      "encoder_q-layer.3": 809.51,
      "encoder_q-layer.4": 855.9822,
      "encoder_q-layer.5": 892.1663,
      "encoder_q-layer.6": 1029.4948,
      "encoder_q-layer.7": 1222.3174,
      "encoder_q-layer.8": 1453.366,
      "encoder_q-layer.9": 1242.4265,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7812,
      "inbatch_pos_score": 1.4434,
      "learning_rate": 4.833333333333333e-06,
      "loss": 3.1518,
      "norm_diff": 0.0699,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2061.9206,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7808,
      "query_norm": 1.5463,
      "queue_k_norm": 1.6178,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0983,
      "sent_len_1": 66.8235,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.6887,
      "stdk": 0.049,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91300
    },
    {
      "accuracy": 51.6602,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1361,
      "doc_norm": 1.6185,
      "encoder_q-embeddings": 2391.1272,
      "encoder_q-layer.0": 1565.078,
      "encoder_q-layer.1": 1730.4836,
      "encoder_q-layer.10": 1333.6841,
      "encoder_q-layer.11": 3319.9734,
      "encoder_q-layer.2": 2025.8087,
      "encoder_q-layer.3": 2230.9026,
      "encoder_q-layer.4": 2473.8528,
      "encoder_q-layer.5": 2564.929,
      "encoder_q-layer.6": 2238.1052,
      "encoder_q-layer.7": 2312.2578,
      "encoder_q-layer.8": 1946.502,
      "encoder_q-layer.9": 1344.7091,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7799,
      "inbatch_pos_score": 1.4443,
      "learning_rate": 4.777777777777778e-06,
      "loss": 3.1361,
      "norm_diff": 0.0771,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3361.4168,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7798,
      "query_norm": 1.5414,
      "queue_k_norm": 1.6194,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9493,
      "sent_len_1": 66.9286,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.55,
      "stdk": 0.0491,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91400
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1167,
      "doc_norm": 1.623,
      "encoder_q-embeddings": 1370.5104,
      "encoder_q-layer.0": 886.6631,
      "encoder_q-layer.1": 1018.9932,
      "encoder_q-layer.10": 1567.5519,
      "encoder_q-layer.11": 3255.9453,
      "encoder_q-layer.2": 1113.9497,
      "encoder_q-layer.3": 1290.3982,
      "encoder_q-layer.4": 1396.6667,
      "encoder_q-layer.5": 1506.4825,
      "encoder_q-layer.6": 1727.2065,
      "encoder_q-layer.7": 1852.2405,
      "encoder_q-layer.8": 1671.7837,
      "encoder_q-layer.9": 1421.0969,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7787,
      "inbatch_pos_score": 1.4414,
      "learning_rate": 4.722222222222222e-06,
      "loss": 3.1167,
      "norm_diff": 0.0722,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2504.6794,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7798,
      "query_norm": 1.5507,
      "queue_k_norm": 1.6173,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0967,
      "sent_len_1": 66.5291,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.4787,
      "stdk": 0.0493,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 91500
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.151,
      "doc_norm": 1.6209,
      "encoder_q-embeddings": 1070.6423,
      "encoder_q-layer.0": 702.5678,
      "encoder_q-layer.1": 742.8641,
      "encoder_q-layer.10": 1329.2046,
      "encoder_q-layer.11": 3083.3459,
      "encoder_q-layer.2": 840.8929,
      "encoder_q-layer.3": 877.1118,
      "encoder_q-layer.4": 971.7667,
      "encoder_q-layer.5": 1029.7902,
      "encoder_q-layer.6": 1179.8107,
      "encoder_q-layer.7": 1237.5272,
      "encoder_q-layer.8": 1353.3536,
      "encoder_q-layer.9": 1233.2717,
      "epoch": 0.89,
      "inbatch_neg_score": 0.7816,
      "inbatch_pos_score": 1.4443,
      "learning_rate": 4.666666666666667e-06,
      "loss": 3.151,
      "norm_diff": 0.0816,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2084.523,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7808,
      "query_norm": 1.5393,
      "queue_k_norm": 1.6173,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1091,
      "sent_len_1": 66.9521,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.585,
      "stdk": 0.0492,
      "stdq": 0.0452,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 91600
    },
    {
      "accuracy": 58.7891,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1446,
      "doc_norm": 1.6141,
      "encoder_q-embeddings": 1142.4985,
      "encoder_q-layer.0": 776.1598,
      "encoder_q-layer.1": 816.4971,
      "encoder_q-layer.10": 1229.9894,
      "encoder_q-layer.11": 2754.127,
      "encoder_q-layer.2": 908.1188,
      "encoder_q-layer.3": 1014.8363,
      "encoder_q-layer.4": 1095.8186,
      "encoder_q-layer.5": 1127.4734,
      "encoder_q-layer.6": 1150.1523,
      "encoder_q-layer.7": 1345.2483,
      "encoder_q-layer.8": 1410.446,
      "encoder_q-layer.9": 1173.7563,
      "epoch": 0.9,
      "inbatch_neg_score": 0.783,
      "inbatch_pos_score": 1.4932,
      "learning_rate": 4.611111111111111e-06,
      "loss": 3.1446,
      "norm_diff": 0.0524,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2046.6213,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7822,
      "query_norm": 1.5617,
      "queue_k_norm": 1.6182,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1134,
      "sent_len_1": 66.9602,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5037,
      "stdk": 0.0489,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91700
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1605,
      "doc_norm": 1.6215,
      "encoder_q-embeddings": 1655.2108,
      "encoder_q-layer.0": 1197.9436,
      "encoder_q-layer.1": 1450.7717,
      "encoder_q-layer.10": 1248.8894,
      "encoder_q-layer.11": 2929.0046,
      "encoder_q-layer.2": 1808.9399,
      "encoder_q-layer.3": 1907.6146,
      "encoder_q-layer.4": 2213.0815,
      "encoder_q-layer.5": 2161.1528,
      "encoder_q-layer.6": 2232.353,
      "encoder_q-layer.7": 2225.4822,
      "encoder_q-layer.8": 1742.3076,
      "encoder_q-layer.9": 1319.7421,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7823,
      "inbatch_pos_score": 1.4648,
      "learning_rate": 4.555555555555556e-06,
      "loss": 3.1605,
      "norm_diff": 0.0764,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2913.273,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7817,
      "query_norm": 1.5451,
      "queue_k_norm": 1.6193,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0574,
      "sent_len_1": 66.7293,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.4175,
      "stdk": 0.0492,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91800
    },
    {
      "accuracy": 56.6406,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1332,
      "doc_norm": 1.6142,
      "encoder_q-embeddings": 1411.7457,
      "encoder_q-layer.0": 979.8738,
      "encoder_q-layer.1": 1054.0824,
      "encoder_q-layer.10": 1240.9027,
      "encoder_q-layer.11": 2940.8445,
      "encoder_q-layer.2": 1254.0383,
      "encoder_q-layer.3": 1327.2642,
      "encoder_q-layer.4": 1388.3171,
      "encoder_q-layer.5": 1389.9305,
      "encoder_q-layer.6": 1470.7432,
      "encoder_q-layer.7": 1474.6541,
      "encoder_q-layer.8": 1514.4585,
      "encoder_q-layer.9": 1228.8815,
      "epoch": 0.9,
      "inbatch_neg_score": 0.783,
      "inbatch_pos_score": 1.459,
      "learning_rate": 4.5e-06,
      "loss": 3.1332,
      "norm_diff": 0.0736,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2326.5752,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7832,
      "query_norm": 1.5406,
      "queue_k_norm": 1.6185,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0363,
      "sent_len_1": 66.6423,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.9087,
      "stdk": 0.0489,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 91900
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1285,
      "doc_norm": 1.6181,
      "encoder_q-embeddings": 1182.4786,
      "encoder_q-layer.0": 842.1204,
      "encoder_q-layer.1": 915.4605,
      "encoder_q-layer.10": 1272.6244,
      "encoder_q-layer.11": 3007.4805,
      "encoder_q-layer.2": 1107.6012,
      "encoder_q-layer.3": 1141.6085,
      "encoder_q-layer.4": 1159.5773,
      "encoder_q-layer.5": 1267.5977,
      "encoder_q-layer.6": 1348.1115,
      "encoder_q-layer.7": 1507.0249,
      "encoder_q-layer.8": 1563.3118,
      "encoder_q-layer.9": 1276.2053,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7806,
      "inbatch_pos_score": 1.4678,
      "learning_rate": 4.444444444444445e-06,
      "loss": 3.1285,
      "norm_diff": 0.0661,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2247.0159,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7817,
      "query_norm": 1.552,
      "queue_k_norm": 1.6184,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1792,
      "sent_len_1": 66.8932,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.9837,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 92000
    },
    {
      "accuracy": 55.5664,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1483,
      "doc_norm": 1.6081,
      "encoder_q-embeddings": 1163.5306,
      "encoder_q-layer.0": 763.8153,
      "encoder_q-layer.1": 830.0422,
      "encoder_q-layer.10": 1298.9288,
      "encoder_q-layer.11": 2997.0725,
      "encoder_q-layer.2": 910.5572,
      "encoder_q-layer.3": 1000.2888,
      "encoder_q-layer.4": 1121.353,
      "encoder_q-layer.5": 1151.4661,
      "encoder_q-layer.6": 1219.0945,
      "encoder_q-layer.7": 1429.4966,
      "encoder_q-layer.8": 1614.1975,
      "encoder_q-layer.9": 1277.6166,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7842,
      "inbatch_pos_score": 1.4639,
      "learning_rate": 4.388888888888889e-06,
      "loss": 3.1483,
      "norm_diff": 0.0671,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2203.8285,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7842,
      "query_norm": 1.5411,
      "queue_k_norm": 1.6205,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8205,
      "sent_len_1": 66.8201,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 193.2725,
      "stdk": 0.0486,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 92100
    },
    {
      "accuracy": 58.2031,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1321,
      "doc_norm": 1.6248,
      "encoder_q-embeddings": 1353.2745,
      "encoder_q-layer.0": 921.7433,
      "encoder_q-layer.1": 1030.1763,
      "encoder_q-layer.10": 1208.8793,
      "encoder_q-layer.11": 2962.8721,
      "encoder_q-layer.2": 1224.4006,
      "encoder_q-layer.3": 1408.1144,
      "encoder_q-layer.4": 1713.1532,
      "encoder_q-layer.5": 1796.2423,
      "encoder_q-layer.6": 1718.1333,
      "encoder_q-layer.7": 2033.1458,
      "encoder_q-layer.8": 1959.2148,
      "encoder_q-layer.9": 1343.3875,
      "epoch": 0.9,
      "inbatch_neg_score": 0.786,
      "inbatch_pos_score": 1.499,
      "learning_rate": 4.333333333333334e-06,
      "loss": 3.1321,
      "norm_diff": 0.0679,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2565.6656,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7856,
      "query_norm": 1.5569,
      "queue_k_norm": 1.6185,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8371,
      "sent_len_1": 66.651,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3887,
      "stdk": 0.0493,
      "stdq": 0.046,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 92200
    },
    {
      "accuracy": 58.3008,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1277,
      "doc_norm": 1.6227,
      "encoder_q-embeddings": 1291.1295,
      "encoder_q-layer.0": 932.9499,
      "encoder_q-layer.1": 1061.1166,
      "encoder_q-layer.10": 1210.5081,
      "encoder_q-layer.11": 2831.8218,
      "encoder_q-layer.2": 1232.515,
      "encoder_q-layer.3": 1282.8245,
      "encoder_q-layer.4": 1412.1436,
      "encoder_q-layer.5": 1556.5281,
      "encoder_q-layer.6": 1493.1489,
      "encoder_q-layer.7": 1570.2642,
      "encoder_q-layer.8": 1568.998,
      "encoder_q-layer.9": 1270.8473,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7836,
      "inbatch_pos_score": 1.4844,
      "learning_rate": 4.277777777777778e-06,
      "loss": 3.1277,
      "norm_diff": 0.0676,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2298.4656,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7837,
      "query_norm": 1.5551,
      "queue_k_norm": 1.6209,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7492,
      "sent_len_1": 66.5407,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1513,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 92300
    },
    {
      "accuracy": 58.5938,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1266,
      "doc_norm": 1.6236,
      "encoder_q-embeddings": 1036.2245,
      "encoder_q-layer.0": 674.2542,
      "encoder_q-layer.1": 745.5438,
      "encoder_q-layer.10": 1273.1465,
      "encoder_q-layer.11": 2817.665,
      "encoder_q-layer.2": 878.1381,
      "encoder_q-layer.3": 998.0298,
      "encoder_q-layer.4": 1087.4553,
      "encoder_q-layer.5": 1127.1315,
      "encoder_q-layer.6": 1162.8188,
      "encoder_q-layer.7": 1291.4846,
      "encoder_q-layer.8": 1404.3505,
      "encoder_q-layer.9": 1235.2609,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7844,
      "inbatch_pos_score": 1.4746,
      "learning_rate": 4.222222222222223e-06,
      "loss": 3.1266,
      "norm_diff": 0.0677,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2005.9712,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7852,
      "query_norm": 1.5559,
      "queue_k_norm": 1.6199,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9223,
      "sent_len_1": 66.785,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6788,
      "stdk": 0.0493,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 92400
    },
    {
      "accuracy": 54.9805,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1265,
      "doc_norm": 1.6164,
      "encoder_q-embeddings": 1007.8522,
      "encoder_q-layer.0": 662.0234,
      "encoder_q-layer.1": 686.5673,
      "encoder_q-layer.10": 1345.323,
      "encoder_q-layer.11": 3035.3542,
      "encoder_q-layer.2": 770.7716,
      "encoder_q-layer.3": 788.0311,
      "encoder_q-layer.4": 839.1525,
      "encoder_q-layer.5": 886.7028,
      "encoder_q-layer.6": 1038.8591,
      "encoder_q-layer.7": 1283.1396,
      "encoder_q-layer.8": 1399.2843,
      "encoder_q-layer.9": 1259.689,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7844,
      "inbatch_pos_score": 1.459,
      "learning_rate": 4.166666666666667e-06,
      "loss": 3.1265,
      "norm_diff": 0.0768,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2039.2101,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7852,
      "query_norm": 1.5397,
      "queue_k_norm": 1.6222,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9921,
      "sent_len_1": 66.724,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3887,
      "stdk": 0.0489,
      "stdq": 0.0452,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 92500
    },
    {
      "accuracy": 58.1055,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1215,
      "doc_norm": 1.6264,
      "encoder_q-embeddings": 1747.989,
      "encoder_q-layer.0": 1167.719,
      "encoder_q-layer.1": 1405.3662,
      "encoder_q-layer.10": 1240.0929,
      "encoder_q-layer.11": 2863.686,
      "encoder_q-layer.2": 1857.3412,
      "encoder_q-layer.3": 1988.991,
      "encoder_q-layer.4": 2085.2405,
      "encoder_q-layer.5": 2142.1973,
      "encoder_q-layer.6": 2079.8997,
      "encoder_q-layer.7": 2195.4072,
      "encoder_q-layer.8": 1794.6545,
      "encoder_q-layer.9": 1317.7954,
      "epoch": 0.9,
      "inbatch_neg_score": 0.7868,
      "inbatch_pos_score": 1.4961,
      "learning_rate": 4.111111111111112e-06,
      "loss": 3.1215,
      "norm_diff": 0.0761,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2876.8931,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.5503,
      "queue_k_norm": 1.6198,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0212,
      "sent_len_1": 66.9801,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.895,
      "stdk": 0.0494,
      "stdq": 0.0457,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 92600
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1462,
      "doc_norm": 1.6248,
      "encoder_q-embeddings": 1318.671,
      "encoder_q-layer.0": 872.1504,
      "encoder_q-layer.1": 971.0966,
      "encoder_q-layer.10": 1242.6879,
      "encoder_q-layer.11": 3018.552,
      "encoder_q-layer.2": 1175.1174,
      "encoder_q-layer.3": 1248.729,
      "encoder_q-layer.4": 1312.5579,
      "encoder_q-layer.5": 1351.3217,
      "encoder_q-layer.6": 1461.5479,
      "encoder_q-layer.7": 1625.6824,
      "encoder_q-layer.8": 1676.9071,
      "encoder_q-layer.9": 1248.5825,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7884,
      "inbatch_pos_score": 1.4648,
      "learning_rate": 4.055555555555556e-06,
      "loss": 3.1462,
      "norm_diff": 0.079,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2312.7251,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.5458,
      "queue_k_norm": 1.6221,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1184,
      "sent_len_1": 67.0419,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6413,
      "stdk": 0.0492,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 92700
    },
    {
      "accuracy": 59.2773,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1192,
      "doc_norm": 1.6251,
      "encoder_q-embeddings": 2013.3914,
      "encoder_q-layer.0": 1395.4536,
      "encoder_q-layer.1": 1515.3904,
      "encoder_q-layer.10": 1191.2639,
      "encoder_q-layer.11": 2794.2681,
      "encoder_q-layer.2": 1699.7468,
      "encoder_q-layer.3": 1949.6741,
      "encoder_q-layer.4": 2087.3528,
      "encoder_q-layer.5": 2174.0317,
      "encoder_q-layer.6": 2191.5125,
      "encoder_q-layer.7": 1912.5828,
      "encoder_q-layer.8": 1716.593,
      "encoder_q-layer.9": 1276.7861,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7867,
      "inbatch_pos_score": 1.5107,
      "learning_rate": 4.000000000000001e-06,
      "loss": 3.1192,
      "norm_diff": 0.0635,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2874.6399,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.5616,
      "queue_k_norm": 1.6204,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9872,
      "sent_len_1": 66.7734,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.64,
      "stdk": 0.0493,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 92800
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1336,
      "doc_norm": 1.6191,
      "encoder_q-embeddings": 910.9048,
      "encoder_q-layer.0": 595.2731,
      "encoder_q-layer.1": 616.0833,
      "encoder_q-layer.10": 1217.7056,
      "encoder_q-layer.11": 2762.5862,
      "encoder_q-layer.2": 695.5132,
      "encoder_q-layer.3": 730.767,
      "encoder_q-layer.4": 754.5601,
      "encoder_q-layer.5": 789.1526,
      "encoder_q-layer.6": 896.5803,
      "encoder_q-layer.7": 1036.9031,
      "encoder_q-layer.8": 1301.9276,
      "encoder_q-layer.9": 1207.2344,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7882,
      "inbatch_pos_score": 1.4736,
      "learning_rate": 3.944444444444445e-06,
      "loss": 3.1336,
      "norm_diff": 0.0671,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1822.1877,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.552,
      "queue_k_norm": 1.6209,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8469,
      "sent_len_1": 66.8491,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5938,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 92900
    },
    {
      "accuracy": 52.2461,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1424,
      "doc_norm": 1.6197,
      "encoder_q-embeddings": 1376.5316,
      "encoder_q-layer.0": 923.1931,
      "encoder_q-layer.1": 1103.8146,
      "encoder_q-layer.10": 1274.6606,
      "encoder_q-layer.11": 3058.8328,
      "encoder_q-layer.2": 1345.6653,
      "encoder_q-layer.3": 1461.6534,
      "encoder_q-layer.4": 1624.6903,
      "encoder_q-layer.5": 1571.7216,
      "encoder_q-layer.6": 1539.7407,
      "encoder_q-layer.7": 1691.1571,
      "encoder_q-layer.8": 1782.5784,
      "encoder_q-layer.9": 1348.8658,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7884,
      "inbatch_pos_score": 1.4512,
      "learning_rate": 3.888888888888889e-06,
      "loss": 3.1424,
      "norm_diff": 0.0759,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2500.4172,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.5437,
      "queue_k_norm": 1.6221,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9405,
      "sent_len_1": 66.8225,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.04,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93000
    },
    {
      "accuracy": 55.7617,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1435,
      "doc_norm": 1.616,
      "encoder_q-embeddings": 1168.6893,
      "encoder_q-layer.0": 779.5818,
      "encoder_q-layer.1": 861.9111,
      "encoder_q-layer.10": 1376.9758,
      "encoder_q-layer.11": 2956.4043,
      "encoder_q-layer.2": 988.1011,
      "encoder_q-layer.3": 1145.705,
      "encoder_q-layer.4": 1181.3917,
      "encoder_q-layer.5": 1230.3069,
      "encoder_q-layer.6": 1397.4772,
      "encoder_q-layer.7": 1564.8424,
      "encoder_q-layer.8": 1536.5613,
      "encoder_q-layer.9": 1281.6992,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7868,
      "inbatch_pos_score": 1.4678,
      "learning_rate": 3.833333333333334e-06,
      "loss": 3.1435,
      "norm_diff": 0.062,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2194.7191,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7866,
      "query_norm": 1.554,
      "queue_k_norm": 1.6222,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9483,
      "sent_len_1": 66.5102,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7525,
      "stdk": 0.0488,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93100
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1245,
      "doc_norm": 1.6182,
      "encoder_q-embeddings": 1474.9506,
      "encoder_q-layer.0": 1065.7274,
      "encoder_q-layer.1": 1218.2291,
      "encoder_q-layer.10": 1260.8616,
      "encoder_q-layer.11": 2789.8259,
      "encoder_q-layer.2": 1363.5448,
      "encoder_q-layer.3": 1459.1699,
      "encoder_q-layer.4": 1571.5029,
      "encoder_q-layer.5": 1707.3986,
      "encoder_q-layer.6": 1638.3802,
      "encoder_q-layer.7": 1509.33,
      "encoder_q-layer.8": 1443.3059,
      "encoder_q-layer.9": 1229.1401,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7883,
      "inbatch_pos_score": 1.501,
      "learning_rate": 3.777777777777778e-06,
      "loss": 3.1245,
      "norm_diff": 0.0612,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2377.2028,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.557,
      "queue_k_norm": 1.6226,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1361,
      "sent_len_1": 66.8775,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.535,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93200
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1424,
      "doc_norm": 1.6219,
      "encoder_q-embeddings": 981.7151,
      "encoder_q-layer.0": 616.7434,
      "encoder_q-layer.1": 654.4413,
      "encoder_q-layer.10": 1228.7557,
      "encoder_q-layer.11": 2896.7676,
      "encoder_q-layer.2": 735.9194,
      "encoder_q-layer.3": 790.5229,
      "encoder_q-layer.4": 863.2541,
      "encoder_q-layer.5": 876.7355,
      "encoder_q-layer.6": 991.8033,
      "encoder_q-layer.7": 1214.2354,
      "encoder_q-layer.8": 1356.38,
      "encoder_q-layer.9": 1222.1074,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7887,
      "inbatch_pos_score": 1.4727,
      "learning_rate": 3.722222222222222e-06,
      "loss": 3.1424,
      "norm_diff": 0.077,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1943.1084,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7881,
      "query_norm": 1.5449,
      "queue_k_norm": 1.6212,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8259,
      "sent_len_1": 66.6539,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.5163,
      "stdk": 0.0491,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93300
    },
    {
      "accuracy": 52.2461,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1026,
      "doc_norm": 1.6172,
      "encoder_q-embeddings": 1387.9846,
      "encoder_q-layer.0": 903.8665,
      "encoder_q-layer.1": 1019.2722,
      "encoder_q-layer.10": 1342.0372,
      "encoder_q-layer.11": 3175.6724,
      "encoder_q-layer.2": 1217.8506,
      "encoder_q-layer.3": 1426.8381,
      "encoder_q-layer.4": 1432.9938,
      "encoder_q-layer.5": 1538.9025,
      "encoder_q-layer.6": 1458.677,
      "encoder_q-layer.7": 1517.0863,
      "encoder_q-layer.8": 1545.7881,
      "encoder_q-layer.9": 1273.6444,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7906,
      "inbatch_pos_score": 1.4424,
      "learning_rate": 3.666666666666667e-06,
      "loss": 3.1026,
      "norm_diff": 0.0672,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2420.8622,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.79,
      "query_norm": 1.55,
      "queue_k_norm": 1.6237,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2333,
      "sent_len_1": 66.8661,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.585,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93400
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1497,
      "doc_norm": 1.6255,
      "encoder_q-embeddings": 1986.0638,
      "encoder_q-layer.0": 1289.3617,
      "encoder_q-layer.1": 1459.9329,
      "encoder_q-layer.10": 1303.1978,
      "encoder_q-layer.11": 2970.7102,
      "encoder_q-layer.2": 1670.0494,
      "encoder_q-layer.3": 1815.0896,
      "encoder_q-layer.4": 2029.5868,
      "encoder_q-layer.5": 2280.761,
      "encoder_q-layer.6": 2263.363,
      "encoder_q-layer.7": 2203.9207,
      "encoder_q-layer.8": 2171.1279,
      "encoder_q-layer.9": 1447.3342,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7879,
      "inbatch_pos_score": 1.4688,
      "learning_rate": 3.611111111111111e-06,
      "loss": 3.1497,
      "norm_diff": 0.0701,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3023.7614,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7876,
      "query_norm": 1.5553,
      "queue_k_norm": 1.6221,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6368,
      "sent_len_1": 66.598,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.6587,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93500
    },
    {
      "accuracy": 58.1055,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1376,
      "doc_norm": 1.6222,
      "encoder_q-embeddings": 1922.505,
      "encoder_q-layer.0": 1306.0651,
      "encoder_q-layer.1": 1470.2117,
      "encoder_q-layer.10": 1225.4946,
      "encoder_q-layer.11": 2835.5305,
      "encoder_q-layer.2": 1681.8344,
      "encoder_q-layer.3": 1673.7668,
      "encoder_q-layer.4": 1867.2692,
      "encoder_q-layer.5": 1863.4064,
      "encoder_q-layer.6": 2171.1902,
      "encoder_q-layer.7": 2084.6697,
      "encoder_q-layer.8": 1915.8844,
      "encoder_q-layer.9": 1261.3029,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7893,
      "inbatch_pos_score": 1.5,
      "learning_rate": 3.555555555555556e-06,
      "loss": 3.1376,
      "norm_diff": 0.0583,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2821.4225,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7891,
      "query_norm": 1.5638,
      "queue_k_norm": 1.6232,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2176,
      "sent_len_1": 66.991,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.545,
      "stdk": 0.0491,
      "stdq": 0.0463,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93600
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1332,
      "doc_norm": 1.6217,
      "encoder_q-embeddings": 1013.9414,
      "encoder_q-layer.0": 646.6998,
      "encoder_q-layer.1": 693.2679,
      "encoder_q-layer.10": 1250.9517,
      "encoder_q-layer.11": 2863.7568,
      "encoder_q-layer.2": 766.4033,
      "encoder_q-layer.3": 806.4169,
      "encoder_q-layer.4": 883.8264,
      "encoder_q-layer.5": 897.0276,
      "encoder_q-layer.6": 1064.9108,
      "encoder_q-layer.7": 1242.9199,
      "encoder_q-layer.8": 1436.54,
      "encoder_q-layer.9": 1272.1085,
      "epoch": 0.91,
      "inbatch_neg_score": 0.7902,
      "inbatch_pos_score": 1.4902,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 3.1332,
      "norm_diff": 0.0686,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1977.2107,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7896,
      "query_norm": 1.5531,
      "queue_k_norm": 1.6243,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.975,
      "sent_len_1": 66.7922,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.2625,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 93700
    },
    {
      "accuracy": 50.3906,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1447,
      "doc_norm": 1.6247,
      "encoder_q-embeddings": 3184.3501,
      "encoder_q-layer.0": 2469.2822,
      "encoder_q-layer.1": 3071.7944,
      "encoder_q-layer.10": 1375.694,
      "encoder_q-layer.11": 3220.1296,
      "encoder_q-layer.2": 4136.791,
      "encoder_q-layer.3": 4343.2471,
      "encoder_q-layer.4": 4890.0605,
      "encoder_q-layer.5": 4905.6553,
      "encoder_q-layer.6": 4528.6499,
      "encoder_q-layer.7": 5106.0322,
      "encoder_q-layer.8": 4480.1699,
      "encoder_q-layer.9": 2559.3923,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7906,
      "inbatch_pos_score": 1.4346,
      "learning_rate": 3.4444444444444444e-06,
      "loss": 3.1447,
      "norm_diff": 0.0846,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 5826.9178,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.791,
      "query_norm": 1.54,
      "queue_k_norm": 1.6233,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9229,
      "sent_len_1": 66.7941,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1637,
      "stdk": 0.0492,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93800
    },
    {
      "accuracy": 52.4414,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1254,
      "doc_norm": 1.6212,
      "encoder_q-embeddings": 1012.6208,
      "encoder_q-layer.0": 649.0029,
      "encoder_q-layer.1": 689.2939,
      "encoder_q-layer.10": 1343.6239,
      "encoder_q-layer.11": 3266.2822,
      "encoder_q-layer.2": 796.5593,
      "encoder_q-layer.3": 825.7484,
      "encoder_q-layer.4": 906.986,
      "encoder_q-layer.5": 931.8576,
      "encoder_q-layer.6": 1088.5962,
      "encoder_q-layer.7": 1296.3296,
      "encoder_q-layer.8": 1511.5347,
      "encoder_q-layer.9": 1313.5409,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7916,
      "inbatch_pos_score": 1.4316,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 3.1254,
      "norm_diff": 0.0815,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2173.0458,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7915,
      "query_norm": 1.5398,
      "queue_k_norm": 1.6237,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0266,
      "sent_len_1": 66.8344,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1775,
      "stdk": 0.049,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 93900
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1498,
      "doc_norm": 1.6214,
      "encoder_q-embeddings": 15242.0293,
      "encoder_q-layer.0": 10864.2871,
      "encoder_q-layer.1": 12540.6006,
      "encoder_q-layer.10": 1371.4846,
      "encoder_q-layer.11": 3052.8528,
      "encoder_q-layer.2": 13971.8086,
      "encoder_q-layer.3": 16330.3389,
      "encoder_q-layer.4": 17514.1523,
      "encoder_q-layer.5": 22599.0742,
      "encoder_q-layer.6": 21939.2852,
      "encoder_q-layer.7": 21948.0039,
      "encoder_q-layer.8": 15483.4355,
      "encoder_q-layer.9": 3466.1262,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7898,
      "inbatch_pos_score": 1.4668,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 3.1498,
      "norm_diff": 0.0741,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 23188.2039,
      "preclip_grad_norm_avg": 0.0002,
      "q@queue_neg_score": 0.7905,
      "query_norm": 1.5473,
      "queue_k_norm": 1.6245,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9028,
      "sent_len_1": 66.7918,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7713,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 94000
    },
    {
      "accuracy": 54.1016,
      "active_queue_size": 16384.0,
      "cl_loss": 3.111,
      "doc_norm": 1.6204,
      "encoder_q-embeddings": 1343.7412,
      "encoder_q-layer.0": 903.291,
      "encoder_q-layer.1": 1061.5906,
      "encoder_q-layer.10": 1337.8779,
      "encoder_q-layer.11": 2904.1794,
      "encoder_q-layer.2": 1211.3588,
      "encoder_q-layer.3": 1291.2751,
      "encoder_q-layer.4": 1432.9548,
      "encoder_q-layer.5": 1508.8173,
      "encoder_q-layer.6": 1705.5363,
      "encoder_q-layer.7": 1847.0104,
      "encoder_q-layer.8": 1807.7113,
      "encoder_q-layer.9": 1329.3501,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7905,
      "inbatch_pos_score": 1.4502,
      "learning_rate": 3.277777777777778e-06,
      "loss": 3.111,
      "norm_diff": 0.0772,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2429.9482,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.79,
      "query_norm": 1.5433,
      "queue_k_norm": 1.6256,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9036,
      "sent_len_1": 67.0286,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4675,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 94100
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1068,
      "doc_norm": 1.6178,
      "encoder_q-embeddings": 536.3345,
      "encoder_q-layer.0": 339.3633,
      "encoder_q-layer.1": 361.5222,
      "encoder_q-layer.10": 638.2821,
      "encoder_q-layer.11": 1459.5365,
      "encoder_q-layer.2": 414.6568,
      "encoder_q-layer.3": 448.8432,
      "encoder_q-layer.4": 473.9182,
      "encoder_q-layer.5": 482.0149,
      "encoder_q-layer.6": 530.9382,
      "encoder_q-layer.7": 604.5976,
      "encoder_q-layer.8": 684.2759,
      "encoder_q-layer.9": 631.2968,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7906,
      "inbatch_pos_score": 1.4727,
      "learning_rate": 3.2222222222222222e-06,
      "loss": 3.1068,
      "norm_diff": 0.0648,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1022.9094,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7905,
      "query_norm": 1.553,
      "queue_k_norm": 1.6249,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9456,
      "sent_len_1": 66.5997,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.075,
      "stdk": 0.0488,
      "stdq": 0.0458,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 94200
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1422,
      "doc_norm": 1.6238,
      "encoder_q-embeddings": 1355.9813,
      "encoder_q-layer.0": 912.7049,
      "encoder_q-layer.1": 1048.6243,
      "encoder_q-layer.10": 634.7236,
      "encoder_q-layer.11": 1458.949,
      "encoder_q-layer.2": 1226.2485,
      "encoder_q-layer.3": 1330.4906,
      "encoder_q-layer.4": 1485.8499,
      "encoder_q-layer.5": 1392.8276,
      "encoder_q-layer.6": 1238.6206,
      "encoder_q-layer.7": 1421.9384,
      "encoder_q-layer.8": 1110.166,
      "encoder_q-layer.9": 712.8215,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7872,
      "inbatch_pos_score": 1.4658,
      "learning_rate": 3.166666666666667e-06,
      "loss": 3.1422,
      "norm_diff": 0.0752,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1844.333,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7886,
      "query_norm": 1.5485,
      "queue_k_norm": 1.6242,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8983,
      "sent_len_1": 66.8521,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.7138,
      "stdk": 0.0492,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 94300
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1493,
      "doc_norm": 1.6234,
      "encoder_q-embeddings": 444.0762,
      "encoder_q-layer.0": 302.3088,
      "encoder_q-layer.1": 312.144,
      "encoder_q-layer.10": 652.6085,
      "encoder_q-layer.11": 1481.9761,
      "encoder_q-layer.2": 346.6377,
      "encoder_q-layer.3": 368.5405,
      "encoder_q-layer.4": 392.3938,
      "encoder_q-layer.5": 411.3578,
      "encoder_q-layer.6": 488.4067,
      "encoder_q-layer.7": 580.5383,
      "encoder_q-layer.8": 690.2202,
      "encoder_q-layer.9": 639.8734,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7886,
      "inbatch_pos_score": 1.4668,
      "learning_rate": 3.111111111111111e-06,
      "loss": 3.1493,
      "norm_diff": 0.0754,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 985.7969,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7896,
      "query_norm": 1.548,
      "queue_k_norm": 1.624,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0211,
      "sent_len_1": 66.5483,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8887,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 94400
    },
    {
      "accuracy": 57.4219,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1271,
      "doc_norm": 1.6216,
      "encoder_q-embeddings": 568.223,
      "encoder_q-layer.0": 381.9745,
      "encoder_q-layer.1": 413.6579,
      "encoder_q-layer.10": 636.0231,
      "encoder_q-layer.11": 1423.1652,
      "encoder_q-layer.2": 477.5397,
      "encoder_q-layer.3": 511.2477,
      "encoder_q-layer.4": 550.0472,
      "encoder_q-layer.5": 589.6891,
      "encoder_q-layer.6": 670.3159,
      "encoder_q-layer.7": 758.991,
      "encoder_q-layer.8": 814.9535,
      "encoder_q-layer.9": 681.3532,
      "epoch": 0.92,
      "inbatch_neg_score": 0.79,
      "inbatch_pos_score": 1.4922,
      "learning_rate": 3.0555555555555556e-06,
      "loss": 3.1271,
      "norm_diff": 0.0544,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1074.6529,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7896,
      "query_norm": 1.5673,
      "queue_k_norm": 1.6239,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0858,
      "sent_len_1": 66.6789,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.575,
      "stdk": 0.049,
      "stdq": 0.0465,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 94500
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1279,
      "doc_norm": 1.6282,
      "encoder_q-embeddings": 490.946,
      "encoder_q-layer.0": 334.3576,
      "encoder_q-layer.1": 356.7077,
      "encoder_q-layer.10": 709.9531,
      "encoder_q-layer.11": 1519.8616,
      "encoder_q-layer.2": 393.6796,
      "encoder_q-layer.3": 412.8143,
      "encoder_q-layer.4": 445.3151,
      "encoder_q-layer.5": 462.7835,
      "encoder_q-layer.6": 506.2007,
      "encoder_q-layer.7": 597.9899,
      "encoder_q-layer.8": 691.6962,
      "encoder_q-layer.9": 625.018,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7901,
      "inbatch_pos_score": 1.459,
      "learning_rate": 3e-06,
      "loss": 3.1279,
      "norm_diff": 0.0877,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1008.5379,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.791,
      "query_norm": 1.5405,
      "queue_k_norm": 1.6239,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.794,
      "sent_len_1": 66.6427,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.5863,
      "stdk": 0.0493,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 94600
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.126,
      "doc_norm": 1.6271,
      "encoder_q-embeddings": 519.178,
      "encoder_q-layer.0": 341.6521,
      "encoder_q-layer.1": 367.6409,
      "encoder_q-layer.10": 635.8693,
      "encoder_q-layer.11": 1488.2173,
      "encoder_q-layer.2": 437.2106,
      "encoder_q-layer.3": 461.0767,
      "encoder_q-layer.4": 486.6225,
      "encoder_q-layer.5": 519.0058,
      "encoder_q-layer.6": 585.5477,
      "encoder_q-layer.7": 650.1013,
      "encoder_q-layer.8": 694.8998,
      "encoder_q-layer.9": 611.652,
      "epoch": 0.92,
      "inbatch_neg_score": 0.7903,
      "inbatch_pos_score": 1.4775,
      "learning_rate": 2.9444444444444445e-06,
      "loss": 3.126,
      "norm_diff": 0.0719,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1036.3615,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7905,
      "query_norm": 1.5552,
      "queue_k_norm": 1.624,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9421,
      "sent_len_1": 66.6292,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.4875,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 94700
    },
    {
      "accuracy": 53.9062,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1081,
      "doc_norm": 1.6279,
      "encoder_q-embeddings": 551.1251,
      "encoder_q-layer.0": 354.1718,
      "encoder_q-layer.1": 367.3044,
      "encoder_q-layer.10": 687.6435,
      "encoder_q-layer.11": 1524.2228,
      "encoder_q-layer.2": 425.2809,
      "encoder_q-layer.3": 437.1326,
      "encoder_q-layer.4": 468.0639,
      "encoder_q-layer.5": 497.2811,
      "encoder_q-layer.6": 544.1554,
      "encoder_q-layer.7": 622.7504,
      "encoder_q-layer.8": 688.8769,
      "encoder_q-layer.9": 621.6591,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7913,
      "inbatch_pos_score": 1.4678,
      "learning_rate": 2.888888888888889e-06,
      "loss": 3.1081,
      "norm_diff": 0.0836,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1046.8554,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7915,
      "query_norm": 1.5443,
      "queue_k_norm": 1.6261,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9151,
      "sent_len_1": 66.5891,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.925,
      "stdk": 0.0492,
      "stdq": 0.0454,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 94800
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1577,
      "doc_norm": 1.6208,
      "encoder_q-embeddings": 669.7737,
      "encoder_q-layer.0": 439.7318,
      "encoder_q-layer.1": 463.068,
      "encoder_q-layer.10": 621.6324,
      "encoder_q-layer.11": 1465.4304,
      "encoder_q-layer.2": 546.4568,
      "encoder_q-layer.3": 587.0074,
      "encoder_q-layer.4": 588.8514,
      "encoder_q-layer.5": 595.2225,
      "encoder_q-layer.6": 647.3148,
      "encoder_q-layer.7": 674.6348,
      "encoder_q-layer.8": 742.3953,
      "encoder_q-layer.9": 627.3733,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7903,
      "inbatch_pos_score": 1.4834,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 3.1577,
      "norm_diff": 0.0707,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1110.5988,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7905,
      "query_norm": 1.5501,
      "queue_k_norm": 1.6261,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8452,
      "sent_len_1": 66.8396,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1687,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 94900
    },
    {
      "accuracy": 55.3711,
      "active_queue_size": 16384.0,
      "cl_loss": 3.0899,
      "doc_norm": 1.6235,
      "encoder_q-embeddings": 635.0618,
      "encoder_q-layer.0": 422.7371,
      "encoder_q-layer.1": 469.0953,
      "encoder_q-layer.10": 660.4855,
      "encoder_q-layer.11": 1464.8101,
      "encoder_q-layer.2": 548.5352,
      "encoder_q-layer.3": 606.2968,
      "encoder_q-layer.4": 644.5683,
      "encoder_q-layer.5": 679.7496,
      "encoder_q-layer.6": 700.4047,
      "encoder_q-layer.7": 742.3655,
      "encoder_q-layer.8": 750.3468,
      "encoder_q-layer.9": 654.8229,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7913,
      "inbatch_pos_score": 1.4814,
      "learning_rate": 2.777777777777778e-06,
      "loss": 3.0899,
      "norm_diff": 0.0674,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1129.362,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7915,
      "query_norm": 1.5562,
      "queue_k_norm": 1.624,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9659,
      "sent_len_1": 66.8603,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0137,
      "stdk": 0.0491,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95000
    },
    {
      "accuracy": 56.7383,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1413,
      "doc_norm": 1.6222,
      "encoder_q-embeddings": 887.0527,
      "encoder_q-layer.0": 617.5422,
      "encoder_q-layer.1": 653.8419,
      "encoder_q-layer.10": 614.267,
      "encoder_q-layer.11": 1414.6835,
      "encoder_q-layer.2": 810.9359,
      "encoder_q-layer.3": 774.3558,
      "encoder_q-layer.4": 787.9551,
      "encoder_q-layer.5": 810.5151,
      "encoder_q-layer.6": 812.8145,
      "encoder_q-layer.7": 866.813,
      "encoder_q-layer.8": 826.743,
      "encoder_q-layer.9": 625.8743,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7913,
      "inbatch_pos_score": 1.4766,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 3.1413,
      "norm_diff": 0.0681,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1280.1927,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7905,
      "query_norm": 1.5541,
      "queue_k_norm": 1.6255,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9402,
      "sent_len_1": 66.5681,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.5725,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95100
    },
    {
      "accuracy": 57.3242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1342,
      "doc_norm": 1.6305,
      "encoder_q-embeddings": 585.3311,
      "encoder_q-layer.0": 388.8509,
      "encoder_q-layer.1": 435.0337,
      "encoder_q-layer.10": 663.8154,
      "encoder_q-layer.11": 1468.7059,
      "encoder_q-layer.2": 487.2504,
      "encoder_q-layer.3": 510.9774,
      "encoder_q-layer.4": 533.7061,
      "encoder_q-layer.5": 583.6622,
      "encoder_q-layer.6": 620.4731,
      "encoder_q-layer.7": 693.0012,
      "encoder_q-layer.8": 759.7797,
      "encoder_q-layer.9": 635.1432,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7899,
      "inbatch_pos_score": 1.4854,
      "learning_rate": 2.666666666666667e-06,
      "loss": 3.1342,
      "norm_diff": 0.0746,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1076.2069,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.791,
      "query_norm": 1.5558,
      "queue_k_norm": 1.6237,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1383,
      "sent_len_1": 66.9736,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6637,
      "stdk": 0.0494,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95200
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1341,
      "doc_norm": 1.6285,
      "encoder_q-embeddings": 538.3407,
      "encoder_q-layer.0": 365.214,
      "encoder_q-layer.1": 410.2883,
      "encoder_q-layer.10": 699.6381,
      "encoder_q-layer.11": 1536.7932,
      "encoder_q-layer.2": 476.0791,
      "encoder_q-layer.3": 519.0912,
      "encoder_q-layer.4": 616.8396,
      "encoder_q-layer.5": 666.0676,
      "encoder_q-layer.6": 848.8862,
      "encoder_q-layer.7": 988.8257,
      "encoder_q-layer.8": 1067.2605,
      "encoder_q-layer.9": 722.0337,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7927,
      "inbatch_pos_score": 1.4756,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 3.1341,
      "norm_diff": 0.0782,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1211.6145,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7925,
      "query_norm": 1.5503,
      "queue_k_norm": 1.6235,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8871,
      "sent_len_1": 66.4839,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5387,
      "stdk": 0.0493,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95300
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1252,
      "doc_norm": 1.6213,
      "encoder_q-embeddings": 554.8084,
      "encoder_q-layer.0": 394.3258,
      "encoder_q-layer.1": 409.9684,
      "encoder_q-layer.10": 640.1682,
      "encoder_q-layer.11": 1437.4579,
      "encoder_q-layer.2": 457.2804,
      "encoder_q-layer.3": 451.3904,
      "encoder_q-layer.4": 489.8665,
      "encoder_q-layer.5": 533.4288,
      "encoder_q-layer.6": 550.2481,
      "encoder_q-layer.7": 609.0526,
      "encoder_q-layer.8": 669.5361,
      "encoder_q-layer.9": 600.7408,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7921,
      "inbatch_pos_score": 1.4902,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 3.1252,
      "norm_diff": 0.0646,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1016.7517,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.793,
      "query_norm": 1.5568,
      "queue_k_norm": 1.6249,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8644,
      "sent_len_1": 66.502,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.8275,
      "stdk": 0.049,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95400
    },
    {
      "accuracy": 56.1523,
      "active_queue_size": 16384.0,
      "cl_loss": 3.0996,
      "doc_norm": 1.623,
      "encoder_q-embeddings": 529.6038,
      "encoder_q-layer.0": 370.7324,
      "encoder_q-layer.1": 416.7878,
      "encoder_q-layer.10": 709.1567,
      "encoder_q-layer.11": 1495.1127,
      "encoder_q-layer.2": 469.7054,
      "encoder_q-layer.3": 509.6913,
      "encoder_q-layer.4": 535.4832,
      "encoder_q-layer.5": 564.4781,
      "encoder_q-layer.6": 571.9897,
      "encoder_q-layer.7": 709.8152,
      "encoder_q-layer.8": 744.337,
      "encoder_q-layer.9": 654.9493,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7938,
      "inbatch_pos_score": 1.4873,
      "learning_rate": 2.5e-06,
      "loss": 3.0996,
      "norm_diff": 0.0673,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1053.6685,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7939,
      "query_norm": 1.5557,
      "queue_k_norm": 1.6251,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9508,
      "sent_len_1": 66.7772,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.1562,
      "stdk": 0.049,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95500
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1439,
      "doc_norm": 1.6248,
      "encoder_q-embeddings": 903.796,
      "encoder_q-layer.0": 638.885,
      "encoder_q-layer.1": 737.5099,
      "encoder_q-layer.10": 668.8021,
      "encoder_q-layer.11": 1452.6793,
      "encoder_q-layer.2": 919.2825,
      "encoder_q-layer.3": 995.9756,
      "encoder_q-layer.4": 1130.3774,
      "encoder_q-layer.5": 1184.4308,
      "encoder_q-layer.6": 1253.2374,
      "encoder_q-layer.7": 1433.2255,
      "encoder_q-layer.8": 1092.0228,
      "encoder_q-layer.9": 655.5226,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7945,
      "inbatch_pos_score": 1.4707,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 3.1439,
      "norm_diff": 0.0724,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1581.3633,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7949,
      "query_norm": 1.5524,
      "queue_k_norm": 1.625,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8521,
      "sent_len_1": 66.5261,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4375,
      "stdk": 0.0491,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95600
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1126,
      "doc_norm": 1.6223,
      "encoder_q-embeddings": 8563.6436,
      "encoder_q-layer.0": 5685.0005,
      "encoder_q-layer.1": 6085.9419,
      "encoder_q-layer.10": 671.2087,
      "encoder_q-layer.11": 1550.1488,
      "encoder_q-layer.2": 6863.8125,
      "encoder_q-layer.3": 7016.5239,
      "encoder_q-layer.4": 7301.9609,
      "encoder_q-layer.5": 7050.9326,
      "encoder_q-layer.6": 7035.1763,
      "encoder_q-layer.7": 8633.5469,
      "encoder_q-layer.8": 5967.6558,
      "encoder_q-layer.9": 1283.3192,
      "epoch": 0.93,
      "inbatch_neg_score": 0.7938,
      "inbatch_pos_score": 1.4658,
      "learning_rate": 2.388888888888889e-06,
      "loss": 3.1126,
      "norm_diff": 0.0743,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9772.355,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7939,
      "query_norm": 1.548,
      "queue_k_norm": 1.6264,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.065,
      "sent_len_1": 66.7889,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6912,
      "stdk": 0.049,
      "stdq": 0.0456,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 95700
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1389,
      "doc_norm": 1.6271,
      "encoder_q-embeddings": 712.3105,
      "encoder_q-layer.0": 511.8627,
      "encoder_q-layer.1": 545.293,
      "encoder_q-layer.10": 678.8977,
      "encoder_q-layer.11": 1529.5693,
      "encoder_q-layer.2": 643.7379,
      "encoder_q-layer.3": 684.3842,
      "encoder_q-layer.4": 755.1694,
      "encoder_q-layer.5": 802.2053,
      "encoder_q-layer.6": 879.5536,
      "encoder_q-layer.7": 867.3336,
      "encoder_q-layer.8": 830.3129,
      "encoder_q-layer.9": 654.4705,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7955,
      "inbatch_pos_score": 1.4648,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 3.1389,
      "norm_diff": 0.0737,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1260.5584,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7949,
      "query_norm": 1.5534,
      "queue_k_norm": 1.6264,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.146,
      "sent_len_1": 66.8189,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.1838,
      "stdk": 0.0492,
      "stdq": 0.0458,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 95800
    },
    {
      "accuracy": 53.3203,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1304,
      "doc_norm": 1.6278,
      "encoder_q-embeddings": 708.0601,
      "encoder_q-layer.0": 490.776,
      "encoder_q-layer.1": 519.9022,
      "encoder_q-layer.10": 635.7335,
      "encoder_q-layer.11": 1535.4153,
      "encoder_q-layer.2": 614.3389,
      "encoder_q-layer.3": 610.5404,
      "encoder_q-layer.4": 613.6389,
      "encoder_q-layer.5": 638.5349,
      "encoder_q-layer.6": 670.6155,
      "encoder_q-layer.7": 744.1726,
      "encoder_q-layer.8": 788.3381,
      "encoder_q-layer.9": 631.729,
      "epoch": 0.94,
      "inbatch_neg_score": 0.794,
      "inbatch_pos_score": 1.4648,
      "learning_rate": 2.277777777777778e-06,
      "loss": 3.1304,
      "norm_diff": 0.0733,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1179.5694,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7944,
      "query_norm": 1.5546,
      "queue_k_norm": 1.6256,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0049,
      "sent_len_1": 66.7911,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.5425,
      "stdk": 0.0492,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 95900
    },
    {
      "accuracy": 54.2969,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1319,
      "doc_norm": 1.6254,
      "encoder_q-embeddings": 465.5964,
      "encoder_q-layer.0": 313.2893,
      "encoder_q-layer.1": 317.8214,
      "encoder_q-layer.10": 614.6633,
      "encoder_q-layer.11": 1473.5298,
      "encoder_q-layer.2": 349.6156,
      "encoder_q-layer.3": 370.1321,
      "encoder_q-layer.4": 398.066,
      "encoder_q-layer.5": 417.4114,
      "encoder_q-layer.6": 469.3312,
      "encoder_q-layer.7": 562.9025,
      "encoder_q-layer.8": 682.2361,
      "encoder_q-layer.9": 634.7638,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7949,
      "inbatch_pos_score": 1.4775,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 3.1319,
      "norm_diff": 0.0706,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 982.619,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7944,
      "query_norm": 1.5548,
      "queue_k_norm": 1.6262,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.909,
      "sent_len_1": 66.7446,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 186.5288,
      "stdk": 0.0491,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96000
    },
    {
      "accuracy": 56.8359,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1176,
      "doc_norm": 1.6251,
      "encoder_q-embeddings": 534.8091,
      "encoder_q-layer.0": 356.9949,
      "encoder_q-layer.1": 397.0004,
      "encoder_q-layer.10": 666.274,
      "encoder_q-layer.11": 1502.5885,
      "encoder_q-layer.2": 451.3138,
      "encoder_q-layer.3": 438.6504,
      "encoder_q-layer.4": 475.6703,
      "encoder_q-layer.5": 488.7105,
      "encoder_q-layer.6": 519.5728,
      "encoder_q-layer.7": 634.561,
      "encoder_q-layer.8": 717.7823,
      "encoder_q-layer.9": 639.2882,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7961,
      "inbatch_pos_score": 1.4883,
      "learning_rate": 2.166666666666667e-06,
      "loss": 3.1176,
      "norm_diff": 0.0682,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1032.9689,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7954,
      "query_norm": 1.5569,
      "queue_k_norm": 1.6256,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8871,
      "sent_len_1": 66.6742,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3,
      "stdk": 0.0491,
      "stdq": 0.046,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96100
    },
    {
      "accuracy": 54.0039,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1512,
      "doc_norm": 1.6312,
      "encoder_q-embeddings": 1040.3597,
      "encoder_q-layer.0": 659.2542,
      "encoder_q-layer.1": 717.5479,
      "encoder_q-layer.10": 1293.0209,
      "encoder_q-layer.11": 3053.6755,
      "encoder_q-layer.2": 828.1411,
      "encoder_q-layer.3": 870.7198,
      "encoder_q-layer.4": 963.8295,
      "encoder_q-layer.5": 1002.6949,
      "encoder_q-layer.6": 1167.6378,
      "encoder_q-layer.7": 1354.8229,
      "encoder_q-layer.8": 1540.9169,
      "encoder_q-layer.9": 1308.0304,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7955,
      "inbatch_pos_score": 1.4531,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 3.1512,
      "norm_diff": 0.0898,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2125.2491,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7959,
      "query_norm": 1.5414,
      "queue_k_norm": 1.6263,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1866,
      "sent_len_1": 66.9829,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.5012,
      "stdk": 0.0494,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96200
    },
    {
      "accuracy": 60.0586,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1238,
      "doc_norm": 1.6344,
      "encoder_q-embeddings": 1069.6962,
      "encoder_q-layer.0": 686.888,
      "encoder_q-layer.1": 720.5658,
      "encoder_q-layer.10": 1245.1848,
      "encoder_q-layer.11": 2913.3389,
      "encoder_q-layer.2": 823.8362,
      "encoder_q-layer.3": 896.2906,
      "encoder_q-layer.4": 971.2831,
      "encoder_q-layer.5": 990.0278,
      "encoder_q-layer.6": 1159.7642,
      "encoder_q-layer.7": 1275.7542,
      "encoder_q-layer.8": 1466.2408,
      "encoder_q-layer.9": 1231.5459,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7951,
      "inbatch_pos_score": 1.5137,
      "learning_rate": 2.055555555555556e-06,
      "loss": 3.1238,
      "norm_diff": 0.0704,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2055.6705,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7954,
      "query_norm": 1.564,
      "queue_k_norm": 1.6256,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9192,
      "sent_len_1": 66.7571,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.0575,
      "stdk": 0.0495,
      "stdq": 0.0463,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96300
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1512,
      "doc_norm": 1.6268,
      "encoder_q-embeddings": 1023.031,
      "encoder_q-layer.0": 659.3455,
      "encoder_q-layer.1": 702.1702,
      "encoder_q-layer.10": 1343.1736,
      "encoder_q-layer.11": 3016.5334,
      "encoder_q-layer.2": 801.4954,
      "encoder_q-layer.3": 853.6965,
      "encoder_q-layer.4": 849.4606,
      "encoder_q-layer.5": 935.2035,
      "encoder_q-layer.6": 1015.8529,
      "encoder_q-layer.7": 1275.8925,
      "encoder_q-layer.8": 1421.6359,
      "encoder_q-layer.9": 1310.4215,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7964,
      "inbatch_pos_score": 1.4756,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 3.1512,
      "norm_diff": 0.0778,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2039.9316,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7954,
      "query_norm": 1.5491,
      "queue_k_norm": 1.6246,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.803,
      "sent_len_1": 66.7484,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5238,
      "stdk": 0.0491,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96400
    },
    {
      "accuracy": 57.9102,
      "active_queue_size": 16384.0,
      "cl_loss": 3.133,
      "doc_norm": 1.6205,
      "encoder_q-embeddings": 1000.5853,
      "encoder_q-layer.0": 655.7438,
      "encoder_q-layer.1": 672.6141,
      "encoder_q-layer.10": 1248.5323,
      "encoder_q-layer.11": 2938.5835,
      "encoder_q-layer.2": 768.9022,
      "encoder_q-layer.3": 793.8857,
      "encoder_q-layer.4": 865.8487,
      "encoder_q-layer.5": 901.8948,
      "encoder_q-layer.6": 1030.8973,
      "encoder_q-layer.7": 1214.2699,
      "encoder_q-layer.8": 1449.9358,
      "encoder_q-layer.9": 1243.8579,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7982,
      "inbatch_pos_score": 1.4941,
      "learning_rate": 1.9444444444444444e-06,
      "loss": 3.133,
      "norm_diff": 0.0688,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2004.0568,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7974,
      "query_norm": 1.5517,
      "queue_k_norm": 1.627,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.5743,
      "sent_len_1": 66.4734,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3075,
      "stdk": 0.0489,
      "stdq": 0.0457,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 96500
    },
    {
      "accuracy": 52.832,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1454,
      "doc_norm": 1.6247,
      "encoder_q-embeddings": 2428.9639,
      "encoder_q-layer.0": 1628.3413,
      "encoder_q-layer.1": 1872.749,
      "encoder_q-layer.10": 1332.3218,
      "encoder_q-layer.11": 2976.3494,
      "encoder_q-layer.2": 2319.9338,
      "encoder_q-layer.3": 2526.5972,
      "encoder_q-layer.4": 2833.4919,
      "encoder_q-layer.5": 2930.968,
      "encoder_q-layer.6": 3194.2434,
      "encoder_q-layer.7": 3998.7031,
      "encoder_q-layer.8": 3379.1235,
      "encoder_q-layer.9": 1488.2338,
      "epoch": 0.94,
      "inbatch_neg_score": 0.796,
      "inbatch_pos_score": 1.457,
      "learning_rate": 1.888888888888889e-06,
      "loss": 3.1454,
      "norm_diff": 0.0812,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4009.78,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7959,
      "query_norm": 1.5434,
      "queue_k_norm": 1.6272,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0721,
      "sent_len_1": 66.7491,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.1788,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 96600
    },
    {
      "accuracy": 58.8867,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1096,
      "doc_norm": 1.6271,
      "encoder_q-embeddings": 895.8323,
      "encoder_q-layer.0": 601.6376,
      "encoder_q-layer.1": 619.0306,
      "encoder_q-layer.10": 1338.7441,
      "encoder_q-layer.11": 3058.1116,
      "encoder_q-layer.2": 726.0092,
      "encoder_q-layer.3": 740.4032,
      "encoder_q-layer.4": 785.4417,
      "encoder_q-layer.5": 836.6,
      "encoder_q-layer.6": 990.3027,
      "encoder_q-layer.7": 1158.1127,
      "encoder_q-layer.8": 1400.0343,
      "encoder_q-layer.9": 1240.7571,
      "epoch": 0.94,
      "inbatch_neg_score": 0.7959,
      "inbatch_pos_score": 1.5166,
      "learning_rate": 1.8333333333333335e-06,
      "loss": 3.1096,
      "norm_diff": 0.0772,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1975.865,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7964,
      "query_norm": 1.5499,
      "queue_k_norm": 1.626,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9573,
      "sent_len_1": 66.9014,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.2688,
      "stdk": 0.0492,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96700
    },
    {
      "accuracy": 55.2734,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1476,
      "doc_norm": 1.6233,
      "encoder_q-embeddings": 1072.4094,
      "encoder_q-layer.0": 691.7136,
      "encoder_q-layer.1": 729.8138,
      "encoder_q-layer.10": 1228.2899,
      "encoder_q-layer.11": 2953.3293,
      "encoder_q-layer.2": 831.7639,
      "encoder_q-layer.3": 915.0835,
      "encoder_q-layer.4": 960.9299,
      "encoder_q-layer.5": 1000.2888,
      "encoder_q-layer.6": 1124.004,
      "encoder_q-layer.7": 1262.4054,
      "encoder_q-layer.8": 1439.8818,
      "encoder_q-layer.9": 1275.5792,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7975,
      "inbatch_pos_score": 1.4873,
      "learning_rate": 1.777777777777778e-06,
      "loss": 3.1476,
      "norm_diff": 0.0718,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2062.5698,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7974,
      "query_norm": 1.5515,
      "queue_k_norm": 1.6268,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.6743,
      "sent_len_1": 66.5766,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4175,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96800
    },
    {
      "accuracy": 58.4961,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1105,
      "doc_norm": 1.6296,
      "encoder_q-embeddings": 1379.7034,
      "encoder_q-layer.0": 909.0817,
      "encoder_q-layer.1": 971.2291,
      "encoder_q-layer.10": 1415.696,
      "encoder_q-layer.11": 3088.835,
      "encoder_q-layer.2": 1136.6156,
      "encoder_q-layer.3": 1161.111,
      "encoder_q-layer.4": 1254.5299,
      "encoder_q-layer.5": 1245.7073,
      "encoder_q-layer.6": 1340.5813,
      "encoder_q-layer.7": 1423.2598,
      "encoder_q-layer.8": 1567.0791,
      "encoder_q-layer.9": 1344.4124,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7964,
      "inbatch_pos_score": 1.4883,
      "learning_rate": 1.7222222222222222e-06,
      "loss": 3.1105,
      "norm_diff": 0.0818,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2285.5006,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7974,
      "query_norm": 1.5478,
      "queue_k_norm": 1.626,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1162,
      "sent_len_1": 66.9289,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1325,
      "stdk": 0.0493,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 96900
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1255,
      "doc_norm": 1.6216,
      "encoder_q-embeddings": 1034.2961,
      "encoder_q-layer.0": 679.8314,
      "encoder_q-layer.1": 736.861,
      "encoder_q-layer.10": 1232.0051,
      "encoder_q-layer.11": 2942.1125,
      "encoder_q-layer.2": 858.4785,
      "encoder_q-layer.3": 874.2191,
      "encoder_q-layer.4": 963.8414,
      "encoder_q-layer.5": 985.099,
      "encoder_q-layer.6": 1090.0382,
      "encoder_q-layer.7": 1264.8972,
      "encoder_q-layer.8": 1424.5657,
      "encoder_q-layer.9": 1251.1136,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7991,
      "inbatch_pos_score": 1.4854,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 3.1255,
      "norm_diff": 0.069,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2047.1265,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7979,
      "query_norm": 1.5526,
      "queue_k_norm": 1.6257,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9657,
      "sent_len_1": 66.802,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.4125,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 97000
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1147,
      "doc_norm": 1.6253,
      "encoder_q-embeddings": 1106.8784,
      "encoder_q-layer.0": 736.1497,
      "encoder_q-layer.1": 801.4086,
      "encoder_q-layer.10": 1374.5938,
      "encoder_q-layer.11": 2904.9907,
      "encoder_q-layer.2": 915.3994,
      "encoder_q-layer.3": 994.3898,
      "encoder_q-layer.4": 1078.8334,
      "encoder_q-layer.5": 1085.6284,
      "encoder_q-layer.6": 1187.9338,
      "encoder_q-layer.7": 1359.032,
      "encoder_q-layer.8": 1449.6924,
      "encoder_q-layer.9": 1200.0723,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7994,
      "inbatch_pos_score": 1.4873,
      "learning_rate": 1.6111111111111111e-06,
      "loss": 3.1147,
      "norm_diff": 0.0735,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2076.5025,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7988,
      "query_norm": 1.5519,
      "queue_k_norm": 1.6267,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9508,
      "sent_len_1": 66.7869,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.6525,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 97100
    },
    {
      "accuracy": 55.8594,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1367,
      "doc_norm": 1.6286,
      "encoder_q-embeddings": 937.454,
      "encoder_q-layer.0": 626.9391,
      "encoder_q-layer.1": 661.9055,
      "encoder_q-layer.10": 1294.2898,
      "encoder_q-layer.11": 3006.8459,
      "encoder_q-layer.2": 743.9683,
      "encoder_q-layer.3": 778.4683,
      "encoder_q-layer.4": 843.2409,
      "encoder_q-layer.5": 913.2504,
      "encoder_q-layer.6": 1059.5433,
      "encoder_q-layer.7": 1225.5563,
      "encoder_q-layer.8": 1404.5641,
      "encoder_q-layer.9": 1289.9438,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7981,
      "inbatch_pos_score": 1.4727,
      "learning_rate": 1.5555555555555556e-06,
      "loss": 3.1367,
      "norm_diff": 0.0848,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2001.1335,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.5438,
      "queue_k_norm": 1.6274,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1752,
      "sent_len_1": 66.6953,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8113,
      "stdk": 0.0492,
      "stdq": 0.0454,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 97200
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1255,
      "doc_norm": 1.6357,
      "encoder_q-embeddings": 1087.1,
      "encoder_q-layer.0": 738.4236,
      "encoder_q-layer.1": 806.7236,
      "encoder_q-layer.10": 1271.7927,
      "encoder_q-layer.11": 2914.8582,
      "encoder_q-layer.2": 919.8989,
      "encoder_q-layer.3": 990.1647,
      "encoder_q-layer.4": 1055.4711,
      "encoder_q-layer.5": 1170.6295,
      "encoder_q-layer.6": 1284.8486,
      "encoder_q-layer.7": 1536.2731,
      "encoder_q-layer.8": 1572.863,
      "encoder_q-layer.9": 1283.5388,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7978,
      "inbatch_pos_score": 1.5049,
      "learning_rate": 1.5e-06,
      "loss": 3.1255,
      "norm_diff": 0.0806,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2153.7599,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.555,
      "queue_k_norm": 1.6272,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8852,
      "sent_len_1": 66.8179,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.1362,
      "stdk": 0.0495,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 97300
    },
    {
      "accuracy": 55.6641,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1482,
      "doc_norm": 1.6264,
      "encoder_q-embeddings": 1091.1449,
      "encoder_q-layer.0": 720.5469,
      "encoder_q-layer.1": 796.7941,
      "encoder_q-layer.10": 1239.9541,
      "encoder_q-layer.11": 2924.6785,
      "encoder_q-layer.2": 921.5251,
      "encoder_q-layer.3": 987.7806,
      "encoder_q-layer.4": 1046.2655,
      "encoder_q-layer.5": 1129.0002,
      "encoder_q-layer.6": 1217.1869,
      "encoder_q-layer.7": 1265.1534,
      "encoder_q-layer.8": 1371.9103,
      "encoder_q-layer.9": 1196.717,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7984,
      "inbatch_pos_score": 1.4854,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 3.1482,
      "norm_diff": 0.0725,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2079.6347,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7979,
      "query_norm": 1.5539,
      "queue_k_norm": 1.627,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8417,
      "sent_len_1": 66.5906,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.305,
      "stdk": 0.0491,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 97400
    },
    {
      "accuracy": 54.7852,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1654,
      "doc_norm": 1.631,
      "encoder_q-embeddings": 1137.7904,
      "encoder_q-layer.0": 789.3634,
      "encoder_q-layer.1": 910.019,
      "encoder_q-layer.10": 1375.2484,
      "encoder_q-layer.11": 3036.7383,
      "encoder_q-layer.2": 1068.483,
      "encoder_q-layer.3": 1139.3661,
      "encoder_q-layer.4": 1246.1377,
      "encoder_q-layer.5": 1352.0104,
      "encoder_q-layer.6": 1366.3429,
      "encoder_q-layer.7": 1501.8373,
      "encoder_q-layer.8": 1448.3319,
      "encoder_q-layer.9": 1210.7128,
      "epoch": 0.95,
      "inbatch_neg_score": 0.799,
      "inbatch_pos_score": 1.4824,
      "learning_rate": 1.388888888888889e-06,
      "loss": 3.1654,
      "norm_diff": 0.0866,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2227.7758,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.5445,
      "queue_k_norm": 1.6272,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0257,
      "sent_len_1": 66.7733,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.2237,
      "stdk": 0.0493,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 97500
    },
    {
      "accuracy": 56.0547,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1276,
      "doc_norm": 1.6267,
      "encoder_q-embeddings": 1003.2155,
      "encoder_q-layer.0": 683.9483,
      "encoder_q-layer.1": 712.5925,
      "encoder_q-layer.10": 1225.3713,
      "encoder_q-layer.11": 2938.7166,
      "encoder_q-layer.2": 819.6984,
      "encoder_q-layer.3": 879.6201,
      "encoder_q-layer.4": 984.6255,
      "encoder_q-layer.5": 1032.6326,
      "encoder_q-layer.6": 1148.5179,
      "encoder_q-layer.7": 1349.6171,
      "encoder_q-layer.8": 1468.3857,
      "encoder_q-layer.9": 1209.1268,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7989,
      "inbatch_pos_score": 1.4648,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 3.1276,
      "norm_diff": 0.0858,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2064.4211,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7988,
      "query_norm": 1.5409,
      "queue_k_norm": 1.6265,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0727,
      "sent_len_1": 66.8338,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7088,
      "stdk": 0.0491,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 97600
    },
    {
      "accuracy": 56.3477,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1425,
      "doc_norm": 1.6264,
      "encoder_q-embeddings": 1273.8069,
      "encoder_q-layer.0": 847.3696,
      "encoder_q-layer.1": 959.4254,
      "encoder_q-layer.10": 1327.9318,
      "encoder_q-layer.11": 2920.1465,
      "encoder_q-layer.2": 1052.8948,
      "encoder_q-layer.3": 1156.9491,
      "encoder_q-layer.4": 1308.0109,
      "encoder_q-layer.5": 1456.3784,
      "encoder_q-layer.6": 1474.1522,
      "encoder_q-layer.7": 1539.7584,
      "encoder_q-layer.8": 1543.7385,
      "encoder_q-layer.9": 1295.0756,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7987,
      "inbatch_pos_score": 1.498,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 3.1425,
      "norm_diff": 0.0624,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2260.2361,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.564,
      "queue_k_norm": 1.6276,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0138,
      "sent_len_1": 66.8099,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.3088,
      "stdk": 0.0491,
      "stdq": 0.0463,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 97700
    },
    {
      "accuracy": 56.4453,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1271,
      "doc_norm": 1.6231,
      "encoder_q-embeddings": 1776.0162,
      "encoder_q-layer.0": 1193.3483,
      "encoder_q-layer.1": 1287.1792,
      "encoder_q-layer.10": 1232.3416,
      "encoder_q-layer.11": 2841.4285,
      "encoder_q-layer.2": 1489.8282,
      "encoder_q-layer.3": 1344.4718,
      "encoder_q-layer.4": 1333.2289,
      "encoder_q-layer.5": 1089.1038,
      "encoder_q-layer.6": 1032.9999,
      "encoder_q-layer.7": 1178.9534,
      "encoder_q-layer.8": 1296.1869,
      "encoder_q-layer.9": 1201.9382,
      "epoch": 0.95,
      "inbatch_neg_score": 0.7976,
      "inbatch_pos_score": 1.4883,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 3.1271,
      "norm_diff": 0.0699,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2327.3118,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.5532,
      "queue_k_norm": 1.628,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8478,
      "sent_len_1": 66.7556,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.7175,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 97800
    },
    {
      "accuracy": 53.5156,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1341,
      "doc_norm": 1.6252,
      "encoder_q-embeddings": 1520.3242,
      "encoder_q-layer.0": 1136.8524,
      "encoder_q-layer.1": 1277.5979,
      "encoder_q-layer.10": 1307.8444,
      "encoder_q-layer.11": 3081.2432,
      "encoder_q-layer.2": 1368.1368,
      "encoder_q-layer.3": 1401.3473,
      "encoder_q-layer.4": 1537.7885,
      "encoder_q-layer.5": 1487.6257,
      "encoder_q-layer.6": 1502.0864,
      "encoder_q-layer.7": 1656.632,
      "encoder_q-layer.8": 1565.696,
      "encoder_q-layer.9": 1308.951,
      "epoch": 0.96,
      "inbatch_neg_score": 0.8004,
      "inbatch_pos_score": 1.4639,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 3.1341,
      "norm_diff": 0.0813,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2523.8068,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7988,
      "query_norm": 1.5438,
      "queue_k_norm": 1.6295,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9144,
      "sent_len_1": 66.8402,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.7012,
      "stdk": 0.049,
      "stdq": 0.0454,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 97900
    },
    {
      "accuracy": 54.1992,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1197,
      "doc_norm": 1.625,
      "encoder_q-embeddings": 1256.0249,
      "encoder_q-layer.0": 833.632,
      "encoder_q-layer.1": 937.6289,
      "encoder_q-layer.10": 1213.2897,
      "encoder_q-layer.11": 2972.5955,
      "encoder_q-layer.2": 1146.119,
      "encoder_q-layer.3": 1189.2988,
      "encoder_q-layer.4": 1334.484,
      "encoder_q-layer.5": 1417.0861,
      "encoder_q-layer.6": 1478.2388,
      "encoder_q-layer.7": 1738.4373,
      "encoder_q-layer.8": 1682.3075,
      "encoder_q-layer.9": 1300.598,
      "epoch": 0.96,
      "inbatch_neg_score": 0.8008,
      "inbatch_pos_score": 1.4688,
      "learning_rate": 1.1111111111111112e-06,
      "loss": 3.1197,
      "norm_diff": 0.0783,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2347.7945,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7998,
      "query_norm": 1.5467,
      "queue_k_norm": 1.6271,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8556,
      "sent_len_1": 66.683,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.6838,
      "stdk": 0.049,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 98000
    },
    {
      "accuracy": 52.1484,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1427,
      "doc_norm": 1.6194,
      "encoder_q-embeddings": 2529.1221,
      "encoder_q-layer.0": 1625.2609,
      "encoder_q-layer.1": 1734.0674,
      "encoder_q-layer.10": 1380.8546,
      "encoder_q-layer.11": 3121.3818,
      "encoder_q-layer.2": 1897.4531,
      "encoder_q-layer.3": 2144.8906,
      "encoder_q-layer.4": 2497.7539,
      "encoder_q-layer.5": 2543.1848,
      "encoder_q-layer.6": 2715.6736,
      "encoder_q-layer.7": 2853.9067,
      "encoder_q-layer.8": 2208.7708,
      "encoder_q-layer.9": 1390.8766,
      "epoch": 0.96,
      "inbatch_neg_score": 0.8014,
      "inbatch_pos_score": 1.4551,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 3.1427,
      "norm_diff": 0.0774,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3510.7584,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8003,
      "query_norm": 1.542,
      "queue_k_norm": 1.6269,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2879,
      "sent_len_1": 66.976,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.015,
      "stdk": 0.0488,
      "stdq": 0.0453,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 98100
    },
    {
      "accuracy": 57.0312,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1358,
      "doc_norm": 1.6271,
      "encoder_q-embeddings": 2122.8474,
      "encoder_q-layer.0": 1398.7823,
      "encoder_q-layer.1": 1452.5569,
      "encoder_q-layer.10": 2627.7222,
      "encoder_q-layer.11": 5939.418,
      "encoder_q-layer.2": 1619.7357,
      "encoder_q-layer.3": 1714.2858,
      "encoder_q-layer.4": 1808.1337,
      "encoder_q-layer.5": 1904.4634,
      "encoder_q-layer.6": 2131.8274,
      "encoder_q-layer.7": 2466.4387,
      "encoder_q-layer.8": 2750.6521,
      "encoder_q-layer.9": 2564.8376,
      "epoch": 0.96,
      "inbatch_neg_score": 0.7975,
      "inbatch_pos_score": 1.4922,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 3.1358,
      "norm_diff": 0.069,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4071.7704,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7988,
      "query_norm": 1.558,
      "queue_k_norm": 1.6284,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0622,
      "sent_len_1": 66.7078,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.3425,
      "stdk": 0.0491,
      "stdq": 0.046,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 98200
    },
    {
      "accuracy": 54.8828,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1007,
      "doc_norm": 1.6293,
      "encoder_q-embeddings": 9076.2988,
      "encoder_q-layer.0": 6417.002,
      "encoder_q-layer.1": 7157.5718,
      "encoder_q-layer.10": 2388.9006,
      "encoder_q-layer.11": 5953.5908,
      "encoder_q-layer.2": 8745.0195,
      "encoder_q-layer.3": 7446.7686,
      "encoder_q-layer.4": 7703.9102,
      "encoder_q-layer.5": 6153.229,
      "encoder_q-layer.6": 4745.6431,
      "encoder_q-layer.7": 4861.5713,
      "encoder_q-layer.8": 3974.219,
      "encoder_q-layer.9": 2536.0923,
      "epoch": 0.96,
      "inbatch_neg_score": 0.8003,
      "inbatch_pos_score": 1.4785,
      "learning_rate": 9.444444444444445e-07,
      "loss": 3.1007,
      "norm_diff": 0.0845,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 9812.5887,
      "preclip_grad_norm_avg": 0.0001,
      "q@queue_neg_score": 0.7998,
      "query_norm": 1.5448,
      "queue_k_norm": 1.628,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.2252,
      "sent_len_1": 67.0728,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 191.505,
      "stdk": 0.0492,
      "stdq": 0.0454,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 98300
    },
    {
      "accuracy": 56.25,
      "active_queue_size": 16384.0,
      "cl_loss": 3.11,
      "doc_norm": 1.6224,
      "encoder_q-embeddings": 2195.0552,
      "encoder_q-layer.0": 1523.8815,
      "encoder_q-layer.1": 1700.2957,
      "encoder_q-layer.10": 2521.437,
      "encoder_q-layer.11": 6004.335,
      "encoder_q-layer.2": 2039.2294,
      "encoder_q-layer.3": 2134.2932,
      "encoder_q-layer.4": 2262.9504,
      "encoder_q-layer.5": 2381.4158,
      "encoder_q-layer.6": 2499.1975,
      "encoder_q-layer.7": 2775.4429,
      "encoder_q-layer.8": 3019.9565,
      "encoder_q-layer.9": 2418.0374,
      "epoch": 0.96,
      "inbatch_neg_score": 0.799,
      "inbatch_pos_score": 1.4902,
      "learning_rate": 8.88888888888889e-07,
      "loss": 3.11,
      "norm_diff": 0.0696,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 4344.7306,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.5528,
      "queue_k_norm": 1.6296,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9862,
      "sent_len_1": 66.9573,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.3663,
      "stdk": 0.0489,
      "stdq": 0.0458,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 98400
    },
    {
      "accuracy": 61.8164,
      "active_queue_size": 16384.0,
      "cl_loss": 3.122,
      "doc_norm": 1.6289,
      "encoder_q-embeddings": 1684.2423,
      "encoder_q-layer.0": 1119.2596,
      "encoder_q-layer.1": 1143.9159,
      "encoder_q-layer.10": 2616.3984,
      "encoder_q-layer.11": 5631.6655,
      "encoder_q-layer.2": 1282.3613,
      "encoder_q-layer.3": 1326.8058,
      "encoder_q-layer.4": 1399.1775,
      "encoder_q-layer.5": 1487.8159,
      "encoder_q-layer.6": 1783.7587,
      "encoder_q-layer.7": 2034.3096,
      "encoder_q-layer.8": 2402.0479,
      "encoder_q-layer.9": 2303.1887,
      "epoch": 0.96,
      "inbatch_neg_score": 0.7968,
      "inbatch_pos_score": 1.5713,
      "learning_rate": 8.333333333333333e-07,
      "loss": 3.122,
      "norm_diff": 0.0523,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3622.8317,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7979,
      "query_norm": 1.5765,
      "queue_k_norm": 1.6288,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9663,
      "sent_len_1": 66.8059,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.8462,
      "stdk": 0.0492,
      "stdq": 0.0469,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 98500
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1534,
      "doc_norm": 1.6281,
      "encoder_q-embeddings": 1240.9238,
      "encoder_q-layer.0": 805.614,
      "encoder_q-layer.1": 917.7928,
      "encoder_q-layer.10": 1338.5502,
      "encoder_q-layer.11": 3054.9529,
      "encoder_q-layer.2": 1103.6761,
      "encoder_q-layer.3": 1210.7523,
      "encoder_q-layer.4": 1348.4546,
      "encoder_q-layer.5": 1376.5708,
      "encoder_q-layer.6": 1474.2139,
      "encoder_q-layer.7": 1582.7113,
      "encoder_q-layer.8": 1650.1355,
      "encoder_q-layer.9": 1309.3778,
      "epoch": 0.96,
      "inbatch_neg_score": 0.7992,
      "inbatch_pos_score": 1.4854,
      "learning_rate": 7.777777777777778e-07,
      "loss": 3.1534,
      "norm_diff": 0.0774,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 2340.9762,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7993,
      "query_norm": 1.5507,
      "queue_k_norm": 1.6285,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9041,
      "sent_len_1": 66.8534,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.3688,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 98600
    },
    {
      "accuracy": 57.3242,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1374,
      "doc_norm": 1.6229,
      "encoder_q-embeddings": 2149.7021,
      "encoder_q-layer.0": 1463.6196,
      "encoder_q-layer.1": 1689.7904,
      "encoder_q-layer.10": 1271.963,
      "encoder_q-layer.11": 3062.9973,
      "encoder_q-layer.2": 1977.5349,
      "encoder_q-layer.3": 2271.2505,
      "encoder_q-layer.4": 2577.2527,
      "encoder_q-layer.5": 2799.7515,
      "encoder_q-layer.6": 3248.8948,
      "encoder_q-layer.7": 3202.9636,
      "encoder_q-layer.8": 2740.9436,
      "encoder_q-layer.9": 1644.2767,
      "epoch": 0.96,
      "inbatch_neg_score": 0.7992,
      "inbatch_pos_score": 1.4941,
      "learning_rate": 7.222222222222222e-07,
      "loss": 3.1374,
      "norm_diff": 0.074,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3651.0146,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7993,
      "query_norm": 1.5489,
      "queue_k_norm": 1.6274,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8583,
      "sent_len_1": 66.8656,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9425,
      "stdk": 0.0489,
      "stdq": 0.0456,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 98700
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1228,
      "doc_norm": 1.6353,
      "encoder_q-embeddings": 2815.772,
      "encoder_q-layer.0": 1941.7177,
      "encoder_q-layer.1": 2119.9092,
      "encoder_q-layer.10": 1365.4369,
      "encoder_q-layer.11": 3197.7788,
      "encoder_q-layer.2": 2345.6003,
      "encoder_q-layer.3": 2460.8271,
      "encoder_q-layer.4": 2667.4785,
      "encoder_q-layer.5": 2478.2627,
      "encoder_q-layer.6": 2820.7776,
      "encoder_q-layer.7": 2966.2373,
      "encoder_q-layer.8": 2592.6677,
      "encoder_q-layer.9": 1420.16,
      "epoch": 0.96,
      "inbatch_neg_score": 0.8012,
      "inbatch_pos_score": 1.4629,
      "learning_rate": 6.666666666666667e-07,
      "loss": 3.1228,
      "norm_diff": 0.0967,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 3805.0481,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8018,
      "query_norm": 1.5386,
      "queue_k_norm": 1.6266,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.1485,
      "sent_len_1": 66.7839,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 190.65,
      "stdk": 0.0495,
      "stdq": 0.0451,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 98800
    },
    {
      "accuracy": 57.7148,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1449,
      "doc_norm": 1.6276,
      "encoder_q-embeddings": 505.4417,
      "encoder_q-layer.0": 332.3542,
      "encoder_q-layer.1": 372.5187,
      "encoder_q-layer.10": 630.0106,
      "encoder_q-layer.11": 1490.8594,
      "encoder_q-layer.2": 404.2332,
      "encoder_q-layer.3": 421.2356,
      "encoder_q-layer.4": 456.2198,
      "encoder_q-layer.5": 464.0778,
      "encoder_q-layer.6": 545.048,
      "encoder_q-layer.7": 639.564,
      "encoder_q-layer.8": 708.5679,
      "encoder_q-layer.9": 623.1411,
      "epoch": 0.97,
      "inbatch_neg_score": 0.7991,
      "inbatch_pos_score": 1.4844,
      "learning_rate": 6.111111111111112e-07,
      "loss": 3.1449,
      "norm_diff": 0.0776,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1029.6532,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7998,
      "query_norm": 1.55,
      "queue_k_norm": 1.626,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7609,
      "sent_len_1": 66.7925,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.5575,
      "stdk": 0.0491,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 98900
    },
    {
      "accuracy": 54.4922,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1427,
      "doc_norm": 1.6257,
      "encoder_q-embeddings": 608.6304,
      "encoder_q-layer.0": 400.6755,
      "encoder_q-layer.1": 461.8012,
      "encoder_q-layer.10": 660.9629,
      "encoder_q-layer.11": 1544.9276,
      "encoder_q-layer.2": 543.2045,
      "encoder_q-layer.3": 569.2464,
      "encoder_q-layer.4": 582.9301,
      "encoder_q-layer.5": 610.5192,
      "encoder_q-layer.6": 685.8011,
      "encoder_q-layer.7": 679.9697,
      "encoder_q-layer.8": 723.1649,
      "encoder_q-layer.9": 646.743,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8004,
      "inbatch_pos_score": 1.4512,
      "learning_rate": 5.555555555555556e-07,
      "loss": 3.1427,
      "norm_diff": 0.0859,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1124.8756,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7998,
      "query_norm": 1.5399,
      "queue_k_norm": 1.6268,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8827,
      "sent_len_1": 66.7448,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.0337,
      "stdk": 0.049,
      "stdq": 0.0452,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 99000
    },
    {
      "accuracy": 57.6172,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1169,
      "doc_norm": 1.625,
      "encoder_q-embeddings": 487.1332,
      "encoder_q-layer.0": 313.3098,
      "encoder_q-layer.1": 347.4796,
      "encoder_q-layer.10": 608.6552,
      "encoder_q-layer.11": 1494.085,
      "encoder_q-layer.2": 405.4938,
      "encoder_q-layer.3": 409.9704,
      "encoder_q-layer.4": 440.6172,
      "encoder_q-layer.5": 447.0775,
      "encoder_q-layer.6": 505.5657,
      "encoder_q-layer.7": 565.4058,
      "encoder_q-layer.8": 677.6221,
      "encoder_q-layer.9": 601.1205,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8007,
      "inbatch_pos_score": 1.498,
      "learning_rate": 5.000000000000001e-07,
      "loss": 3.1169,
      "norm_diff": 0.0746,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 997.4508,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7993,
      "query_norm": 1.5504,
      "queue_k_norm": 1.6275,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8697,
      "sent_len_1": 66.6521,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4563,
      "stdk": 0.049,
      "stdq": 0.0457,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 99100
    },
    {
      "accuracy": 58.7891,
      "active_queue_size": 16384.0,
      "cl_loss": 3.141,
      "doc_norm": 1.6252,
      "encoder_q-embeddings": 1485.5378,
      "encoder_q-layer.0": 971.4633,
      "encoder_q-layer.1": 1100.7433,
      "encoder_q-layer.10": 656.392,
      "encoder_q-layer.11": 1472.4589,
      "encoder_q-layer.2": 1290.7911,
      "encoder_q-layer.3": 1323.7185,
      "encoder_q-layer.4": 1414.8651,
      "encoder_q-layer.5": 1342.6287,
      "encoder_q-layer.6": 1052.2322,
      "encoder_q-layer.7": 999.6823,
      "encoder_q-layer.8": 883.146,
      "encoder_q-layer.9": 640.2141,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8014,
      "inbatch_pos_score": 1.5059,
      "learning_rate": 4.444444444444445e-07,
      "loss": 3.141,
      "norm_diff": 0.0722,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1783.9466,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7998,
      "query_norm": 1.553,
      "queue_k_norm": 1.6258,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.059,
      "sent_len_1": 66.6816,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.0913,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0491,
      "stdqueue_q": 0.0,
      "step": 99200
    },
    {
      "accuracy": 57.5195,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1111,
      "doc_norm": 1.6232,
      "encoder_q-embeddings": 531.3,
      "encoder_q-layer.0": 369.3612,
      "encoder_q-layer.1": 417.9042,
      "encoder_q-layer.10": 605.1995,
      "encoder_q-layer.11": 1401.8605,
      "encoder_q-layer.2": 493.4672,
      "encoder_q-layer.3": 533.6185,
      "encoder_q-layer.4": 572.4954,
      "encoder_q-layer.5": 606.019,
      "encoder_q-layer.6": 683.3998,
      "encoder_q-layer.7": 745.4193,
      "encoder_q-layer.8": 798.1882,
      "encoder_q-layer.9": 621.493,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8005,
      "inbatch_pos_score": 1.5039,
      "learning_rate": 3.888888888888889e-07,
      "loss": 3.1111,
      "norm_diff": 0.0666,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1049.1002,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7998,
      "query_norm": 1.5566,
      "queue_k_norm": 1.6301,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0489,
      "sent_len_1": 66.867,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.6425,
      "stdk": 0.0489,
      "stdq": 0.046,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 99300
    },
    {
      "accuracy": 53.8086,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1356,
      "doc_norm": 1.6264,
      "encoder_q-embeddings": 544.3378,
      "encoder_q-layer.0": 360.8826,
      "encoder_q-layer.1": 396.7446,
      "encoder_q-layer.10": 680.9468,
      "encoder_q-layer.11": 1480.412,
      "encoder_q-layer.2": 444.3055,
      "encoder_q-layer.3": 458.0107,
      "encoder_q-layer.4": 471.8878,
      "encoder_q-layer.5": 500.0557,
      "encoder_q-layer.6": 550.7263,
      "encoder_q-layer.7": 609.4535,
      "encoder_q-layer.8": 689.3629,
      "encoder_q-layer.9": 618.5887,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8011,
      "inbatch_pos_score": 1.4443,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 3.1356,
      "norm_diff": 0.0908,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1040.6837,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8003,
      "query_norm": 1.5357,
      "queue_k_norm": 1.628,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9281,
      "sent_len_1": 66.732,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.2188,
      "stdk": 0.0491,
      "stdq": 0.045,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 99400
    },
    {
      "accuracy": 59.2773,
      "active_queue_size": 16384.0,
      "cl_loss": 3.126,
      "doc_norm": 1.6303,
      "encoder_q-embeddings": 683.7311,
      "encoder_q-layer.0": 466.8259,
      "encoder_q-layer.1": 563.6986,
      "encoder_q-layer.10": 647.4399,
      "encoder_q-layer.11": 1406.7869,
      "encoder_q-layer.2": 685.073,
      "encoder_q-layer.3": 708.8535,
      "encoder_q-layer.4": 781.5479,
      "encoder_q-layer.5": 825.6468,
      "encoder_q-layer.6": 810.918,
      "encoder_q-layer.7": 917.209,
      "encoder_q-layer.8": 962.4536,
      "encoder_q-layer.9": 640.0618,
      "epoch": 0.97,
      "inbatch_neg_score": 0.7983,
      "inbatch_pos_score": 1.5176,
      "learning_rate": 2.777777777777778e-07,
      "loss": 3.126,
      "norm_diff": 0.0682,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1238.2339,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7983,
      "query_norm": 1.5621,
      "queue_k_norm": 1.6277,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.7305,
      "sent_len_1": 66.7192,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.8862,
      "stdk": 0.0492,
      "stdq": 0.0462,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 99500
    },
    {
      "accuracy": 52.7344,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1359,
      "doc_norm": 1.6268,
      "encoder_q-embeddings": 534.6382,
      "encoder_q-layer.0": 362.5091,
      "encoder_q-layer.1": 399.9092,
      "encoder_q-layer.10": 683.9824,
      "encoder_q-layer.11": 1520.4178,
      "encoder_q-layer.2": 440.8601,
      "encoder_q-layer.3": 481.3538,
      "encoder_q-layer.4": 542.3274,
      "encoder_q-layer.5": 542.2966,
      "encoder_q-layer.6": 588.402,
      "encoder_q-layer.7": 684.8657,
      "encoder_q-layer.8": 734.9099,
      "encoder_q-layer.9": 642.4219,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8013,
      "inbatch_pos_score": 1.46,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 3.1359,
      "norm_diff": 0.0802,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1076.4753,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8013,
      "query_norm": 1.5466,
      "queue_k_norm": 1.6286,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.963,
      "sent_len_1": 66.7582,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.84,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 99600
    },
    {
      "accuracy": 55.1758,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1285,
      "doc_norm": 1.6267,
      "encoder_q-embeddings": 509.1297,
      "encoder_q-layer.0": 342.2402,
      "encoder_q-layer.1": 366.0719,
      "encoder_q-layer.10": 688.582,
      "encoder_q-layer.11": 1499.7104,
      "encoder_q-layer.2": 403.1726,
      "encoder_q-layer.3": 416.599,
      "encoder_q-layer.4": 469.459,
      "encoder_q-layer.5": 477.6151,
      "encoder_q-layer.6": 530.834,
      "encoder_q-layer.7": 578.3519,
      "encoder_q-layer.8": 671.5649,
      "encoder_q-layer.9": 621.3677,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8006,
      "inbatch_pos_score": 1.4785,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 3.1285,
      "norm_diff": 0.0795,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1015.968,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8008,
      "query_norm": 1.5472,
      "queue_k_norm": 1.6279,
      "queue_ptr": 4096.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.8921,
      "sent_len_1": 66.7904,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 188.795,
      "stdk": 0.0491,
      "stdq": 0.0455,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 99700
    },
    {
      "accuracy": 57.1289,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1258,
      "doc_norm": 1.6335,
      "encoder_q-embeddings": 468.8048,
      "encoder_q-layer.0": 309.5146,
      "encoder_q-layer.1": 334.0506,
      "encoder_q-layer.10": 697.8617,
      "encoder_q-layer.11": 1488.3055,
      "encoder_q-layer.2": 372.6438,
      "encoder_q-layer.3": 387.1603,
      "encoder_q-layer.4": 418.9378,
      "encoder_q-layer.5": 420.2871,
      "encoder_q-layer.6": 487.7356,
      "encoder_q-layer.7": 578.3751,
      "encoder_q-layer.8": 680.5081,
      "encoder_q-layer.9": 616.0317,
      "epoch": 0.97,
      "inbatch_neg_score": 0.8023,
      "inbatch_pos_score": 1.502,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 3.1258,
      "norm_diff": 0.078,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 984.3736,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8008,
      "query_norm": 1.5555,
      "queue_k_norm": 1.6271,
      "queue_ptr": 8192.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 64.9765,
      "sent_len_1": 66.7563,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 189.4525,
      "stdk": 0.0494,
      "stdq": 0.0459,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 99800
    },
    {
      "accuracy": 55.957,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1336,
      "doc_norm": 1.6221,
      "encoder_q-embeddings": 1234.7236,
      "encoder_q-layer.0": 860.2324,
      "encoder_q-layer.1": 1042.6755,
      "encoder_q-layer.10": 684.2162,
      "encoder_q-layer.11": 1423.9255,
      "encoder_q-layer.2": 1192.1025,
      "encoder_q-layer.3": 1288.9143,
      "encoder_q-layer.4": 1396.2981,
      "encoder_q-layer.5": 1508.1541,
      "encoder_q-layer.6": 1568.2279,
      "encoder_q-layer.7": 1458.5782,
      "encoder_q-layer.8": 1125.9252,
      "encoder_q-layer.9": 659.9648,
      "epoch": 0.98,
      "inbatch_neg_score": 0.7996,
      "inbatch_pos_score": 1.4941,
      "learning_rate": 5.555555555555556e-08,
      "loss": 3.1336,
      "norm_diff": 0.0588,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 1844.0064,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.7993,
      "query_norm": 1.5633,
      "queue_k_norm": 1.6291,
      "queue_ptr": 12288.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.0928,
      "sent_len_1": 66.9947,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 192.5962,
      "stdk": 0.0489,
      "stdq": 0.0463,
      "stdqueue_k": 0.0493,
      "stdqueue_q": 0.0,
      "step": 99900
    },
    {
      "accuracy": 55.4688,
      "active_queue_size": 16384.0,
      "cl_loss": 3.1272,
      "doc_norm": 1.6259,
      "encoder_q-embeddings": 494.1666,
      "encoder_q-layer.0": 326.925,
      "encoder_q-layer.1": 333.7584,
      "encoder_q-layer.10": 614.2658,
      "encoder_q-layer.11": 1479.3057,
      "encoder_q-layer.2": 380.183,
      "encoder_q-layer.3": 389.9003,
      "encoder_q-layer.4": 425.15,
      "encoder_q-layer.5": 430.5507,
      "encoder_q-layer.6": 479.9345,
      "encoder_q-layer.7": 559.0207,
      "encoder_q-layer.8": 643.0842,
      "encoder_q-layer.9": 603.2275,
      "epoch": 0.98,
      "inbatch_neg_score": 0.8001,
      "inbatch_pos_score": 1.4961,
      "learning_rate": 0.0,
      "loss": 3.1272,
      "norm_diff": 0.074,
      "norm_loss": 0.0,
      "postclip_grad_norm": 1.0,
      "preclip_grad_norm": 982.2622,
      "preclip_grad_norm_avg": 0.0,
      "q@queue_neg_score": 0.8003,
      "query_norm": 1.5519,
      "queue_k_norm": 1.6274,
      "queue_ptr": 0.0,
      "queue_q_norm": 0.0,
      "sent_len_0": 65.112,
      "sent_len_1": 66.5937,
      "sent_max_len_0": 128.0,
      "sent_max_len_1": 187.9462,
      "stdk": 0.049,
      "stdq": 0.0458,
      "stdqueue_k": 0.0492,
      "stdqueue_q": 0.0,
      "step": 100000
    },
    {
      "dev_runtime": 38.0401,
      "dev_samples_per_second": 1.682,
      "dev_steps_per_second": 0.026,
      "epoch": 0.98,
      "step": 100000,
      "test_accuracy": 94.62890625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.30579674243927,
      "test_doc_norm": 1.6166253089904785,
      "test_inbatch_neg_score": 1.167407512664795,
      "test_inbatch_pos_score": 2.140146255493164,
      "test_loss": 0.30579674243927,
      "test_loss_align": 0.9694724082946777,
      "test_loss_unif": 2.8071036338806152,
      "test_loss_unif_q@queue": 2.8071038722991943,
      "test_norm_diff": 0.006956934928894043,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.7821612358093262,
      "test_query_norm": 1.616883635520935,
      "test_queue_k_norm": 1.6272437572479248,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.043169643729925156,
      "test_stdq": 0.04317212849855423,
      "test_stdqueue_k": 0.04919165372848511,
      "test_stdqueue_q": 0.0
    },
    {
      "dev_runtime": 38.0401,
      "dev_samples_per_second": 1.682,
      "dev_steps_per_second": 0.026,
      "epoch": 0.98,
      "eval_beir-arguana_ndcg@10": 0.39138,
      "eval_beir-arguana_recall@10": 0.66216,
      "eval_beir-arguana_recall@100": 0.93599,
      "eval_beir-arguana_recall@20": 0.7909,
      "eval_beir-avg_ndcg@10": 0.3805048333333333,
      "eval_beir-avg_recall@10": 0.4477325,
      "eval_beir-avg_recall@100": 0.6274010833333333,
      "eval_beir-avg_recall@20": 0.5071286666666668,
      "eval_beir-cqadupstack_ndcg@10": 0.27104833333333334,
      "eval_beir-cqadupstack_recall@10": 0.363245,
      "eval_beir-cqadupstack_recall@100": 0.5892308333333333,
      "eval_beir-cqadupstack_recall@20": 0.4280566666666667,
      "eval_beir-fiqa_ndcg@10": 0.22343,
      "eval_beir-fiqa_recall@10": 0.28095,
      "eval_beir-fiqa_recall@100": 0.53699,
      "eval_beir-fiqa_recall@20": 0.3412,
      "eval_beir-nfcorpus_ndcg@10": 0.30319,
      "eval_beir-nfcorpus_recall@10": 0.1525,
      "eval_beir-nfcorpus_recall@100": 0.27586,
      "eval_beir-nfcorpus_recall@20": 0.18001,
      "eval_beir-nq_ndcg@10": 0.27168,
      "eval_beir-nq_recall@10": 0.44928,
      "eval_beir-nq_recall@100": 0.79179,
      "eval_beir-nq_recall@20": 0.56817,
      "eval_beir-quora_ndcg@10": 0.82685,
      "eval_beir-quora_recall@10": 0.91814,
      "eval_beir-quora_recall@100": 0.98627,
      "eval_beir-quora_recall@20": 0.95101,
      "eval_beir-scidocs_ndcg@10": 0.14703,
      "eval_beir-scidocs_recall@10": 0.15518,
      "eval_beir-scidocs_recall@100": 0.35427,
      "eval_beir-scidocs_recall@20": 0.21462,
      "eval_beir-scifact_ndcg@10": 0.64459,
      "eval_beir-scifact_recall@10": 0.78733,
      "eval_beir-scifact_recall@100": 0.93422,
      "eval_beir-scifact_recall@20": 0.84011,
      "eval_beir-trec-covid_ndcg@10": 0.53589,
      "eval_beir-trec-covid_recall@10": 0.572,
      "eval_beir-trec-covid_recall@100": 0.4254,
      "eval_beir-trec-covid_recall@20": 0.542,
      "eval_beir-webis-touche2020_ndcg@10": 0.18996,
      "eval_beir-webis-touche2020_recall@10": 0.13654,
      "eval_beir-webis-touche2020_recall@100": 0.44399,
      "eval_beir-webis-touche2020_recall@20": 0.21521,
      "eval_senteval-avg_sts": 0.7273805222302221,
      "eval_senteval-sickr_spearman": 0.6974671648657597,
      "eval_senteval-stsb_spearman": 0.7572938795946844,
      "step": 100000,
      "test_accuracy": 94.62890625,
      "test_active_queue_size": 16384.0,
      "test_cl_loss": 0.30579674243927,
      "test_doc_norm": 1.6166253089904785,
      "test_inbatch_neg_score": 1.167407512664795,
      "test_inbatch_pos_score": 2.140146255493164,
      "test_loss": 0.30579674243927,
      "test_loss_align": 0.9694724082946777,
      "test_loss_unif": 2.8071036338806152,
      "test_loss_unif_q@queue": 2.8071038722991943,
      "test_norm_diff": 0.006956934928894043,
      "test_norm_loss": 0.0,
      "test_q@queue_neg_score": 0.7821612358093262,
      "test_query_norm": 1.616883635520935,
      "test_queue_k_norm": 1.6272437572479248,
      "test_queue_ptr": 0.0,
      "test_queue_q_norm": 0.0,
      "test_stdk": 0.043169643729925156,
      "test_stdq": 0.04317212849855423,
      "test_stdqueue_k": 0.04919165372848511,
      "test_stdqueue_q": 0.0
    },
    {
      "epoch": 0.98,
      "step": 100000,
      "total_flos": 0,
      "train_runtime": 248357.4129,
      "train_samples_per_second": 0.403
    }
  ],
  "max_steps": 100000,
  "num_train_epochs": 1,
  "total_flos": 0,
  "trial_name": null,
  "trial_params": null
}