| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 351, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008547008547008548, | |
| "grad_norm": 0.475104957818985, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9491, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017094017094017096, | |
| "grad_norm": 0.3602507710456848, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7401, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02564102564102564, | |
| "grad_norm": 0.49503466486930847, | |
| "learning_rate": 6e-05, | |
| "loss": 0.9861, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03418803418803419, | |
| "grad_norm": 0.3425951898097992, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3363, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.042735042735042736, | |
| "grad_norm": 0.5006900429725647, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8529, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05128205128205128, | |
| "grad_norm": 0.414726585149765, | |
| "learning_rate": 0.00012, | |
| "loss": 1.2524, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05982905982905983, | |
| "grad_norm": 0.3771184980869293, | |
| "learning_rate": 0.00014, | |
| "loss": 0.8486, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06837606837606838, | |
| "grad_norm": 0.3893500566482544, | |
| "learning_rate": 0.00016, | |
| "loss": 0.8931, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 0.33733996748924255, | |
| "learning_rate": 0.00018, | |
| "loss": 0.7253, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08547008547008547, | |
| "grad_norm": 0.6951685547828674, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9938, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09401709401709402, | |
| "grad_norm": 0.6600769758224487, | |
| "learning_rate": 0.0001999976474595967, | |
| "loss": 0.9379, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10256410256410256, | |
| "grad_norm": 0.3751288056373596, | |
| "learning_rate": 0.00019999058994907564, | |
| "loss": 0.7716, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.4716266393661499, | |
| "learning_rate": 0.00019997882780049847, | |
| "loss": 0.8637, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11965811965811966, | |
| "grad_norm": 0.5008829832077026, | |
| "learning_rate": 0.0001999623615672837, | |
| "loss": 0.9545, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1282051282051282, | |
| "grad_norm": 0.46447134017944336, | |
| "learning_rate": 0.00019994119202418098, | |
| "loss": 1.0945, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.13675213675213677, | |
| "grad_norm": 0.5583352446556091, | |
| "learning_rate": 0.00019991532016723439, | |
| "loss": 0.7752, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1452991452991453, | |
| "grad_norm": 0.4787881076335907, | |
| "learning_rate": 0.00019988474721373568, | |
| "loss": 1.1693, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.4033676087856293, | |
| "learning_rate": 0.00019984947460216707, | |
| "loss": 0.6522, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1623931623931624, | |
| "grad_norm": 0.5240854620933533, | |
| "learning_rate": 0.00019980950399213344, | |
| "loss": 0.7717, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 0.41608908772468567, | |
| "learning_rate": 0.00019976483726428422, | |
| "loss": 0.6719, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1794871794871795, | |
| "grad_norm": 0.48278939723968506, | |
| "learning_rate": 0.0001997154765202251, | |
| "loss": 0.6905, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.18803418803418803, | |
| "grad_norm": 0.445468544960022, | |
| "learning_rate": 0.00019966142408241901, | |
| "loss": 0.7572, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.19658119658119658, | |
| "grad_norm": 0.4958309233188629, | |
| "learning_rate": 0.00019960268249407675, | |
| "loss": 0.8273, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.20512820512820512, | |
| "grad_norm": 0.46474188566207886, | |
| "learning_rate": 0.00019953925451903756, | |
| "loss": 0.776, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.21367521367521367, | |
| "grad_norm": 0.3750772774219513, | |
| "learning_rate": 0.0001994711431416389, | |
| "loss": 1.4004, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.4821198284626007, | |
| "learning_rate": 0.00019939835156657616, | |
| "loss": 1.2073, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.5062704682350159, | |
| "learning_rate": 0.00019932088321875172, | |
| "loss": 0.7346, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.23931623931623933, | |
| "grad_norm": 0.5169610977172852, | |
| "learning_rate": 0.00019923874174311394, | |
| "loss": 0.6212, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.24786324786324787, | |
| "grad_norm": 0.4412098526954651, | |
| "learning_rate": 0.0001991519310044857, | |
| "loss": 1.0973, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 0.49315598607063293, | |
| "learning_rate": 0.00019906045508738228, | |
| "loss": 0.9767, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.26495726495726496, | |
| "grad_norm": 0.4327487647533417, | |
| "learning_rate": 0.0001989643182958196, | |
| "loss": 0.682, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.27350427350427353, | |
| "grad_norm": 0.45163172483444214, | |
| "learning_rate": 0.00019886352515311134, | |
| "loss": 0.7315, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.28205128205128205, | |
| "grad_norm": 0.45258256793022156, | |
| "learning_rate": 0.0001987580804016563, | |
| "loss": 0.9502, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2905982905982906, | |
| "grad_norm": 0.4901167154312134, | |
| "learning_rate": 0.00019864798900271532, | |
| "loss": 0.8431, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.29914529914529914, | |
| "grad_norm": 0.4742427170276642, | |
| "learning_rate": 0.0001985332561361776, | |
| "loss": 0.658, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.4435608983039856, | |
| "learning_rate": 0.00019841388720031727, | |
| "loss": 0.6167, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3162393162393162, | |
| "grad_norm": 0.49269428849220276, | |
| "learning_rate": 0.00019828988781153917, | |
| "loss": 0.7105, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3247863247863248, | |
| "grad_norm": 0.47239962220191956, | |
| "learning_rate": 0.00019816126380411476, | |
| "loss": 0.7264, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.4605489671230316, | |
| "learning_rate": 0.00019802802122990758, | |
| "loss": 0.744, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 0.5027369260787964, | |
| "learning_rate": 0.00019789016635808837, | |
| "loss": 0.7997, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3504273504273504, | |
| "grad_norm": 0.4595533311367035, | |
| "learning_rate": 0.00019774770567484022, | |
| "loss": 0.9184, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.358974358974359, | |
| "grad_norm": 0.4321463108062744, | |
| "learning_rate": 0.00019760064588305345, | |
| "loss": 0.6633, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.36752136752136755, | |
| "grad_norm": 0.4019186198711395, | |
| "learning_rate": 0.00019744899390201006, | |
| "loss": 0.7046, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.37606837606837606, | |
| "grad_norm": 0.4053269326686859, | |
| "learning_rate": 0.0001972927568670583, | |
| "loss": 1.0849, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.4526562988758087, | |
| "learning_rate": 0.00019713194212927696, | |
| "loss": 0.7877, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.39316239316239315, | |
| "grad_norm": 0.47216200828552246, | |
| "learning_rate": 0.00019696655725512933, | |
| "loss": 0.7337, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4017094017094017, | |
| "grad_norm": 0.469186395406723, | |
| "learning_rate": 0.00019679661002610743, | |
| "loss": 0.8048, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.41025641025641024, | |
| "grad_norm": 0.40787142515182495, | |
| "learning_rate": 0.00019662210843836574, | |
| "loss": 0.7041, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4188034188034188, | |
| "grad_norm": 0.4881182610988617, | |
| "learning_rate": 0.0001964430607023449, | |
| "loss": 0.7163, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "grad_norm": 0.5976874828338623, | |
| "learning_rate": 0.00019625947524238563, | |
| "loss": 1.0021, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4358974358974359, | |
| "grad_norm": 0.49746468663215637, | |
| "learning_rate": 0.00019607136069633212, | |
| "loss": 0.6709, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.43088001012802124, | |
| "learning_rate": 0.0001958787259151258, | |
| "loss": 0.6773, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.452991452991453, | |
| "grad_norm": 0.5117547512054443, | |
| "learning_rate": 0.00019568157996238884, | |
| "loss": 0.6801, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.39666032791137695, | |
| "learning_rate": 0.0001954799321139975, | |
| "loss": 0.9736, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4700854700854701, | |
| "grad_norm": 0.4586123526096344, | |
| "learning_rate": 0.00019527379185764612, | |
| "loss": 0.6431, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.47863247863247865, | |
| "grad_norm": 0.5679425001144409, | |
| "learning_rate": 0.00019506316889240027, | |
| "loss": 1.1387, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.48717948717948717, | |
| "grad_norm": 0.4713611304759979, | |
| "learning_rate": 0.00019484807312824067, | |
| "loss": 0.5671, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.49572649572649574, | |
| "grad_norm": 0.5581649541854858, | |
| "learning_rate": 0.0001946285146855968, | |
| "loss": 0.7273, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5042735042735043, | |
| "grad_norm": 0.4809207320213318, | |
| "learning_rate": 0.0001944045038948709, | |
| "loss": 0.7237, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 0.4750754237174988, | |
| "learning_rate": 0.00019417605129595157, | |
| "loss": 0.6509, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5213675213675214, | |
| "grad_norm": 0.4904468059539795, | |
| "learning_rate": 0.0001939431676377183, | |
| "loss": 0.6897, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5299145299145299, | |
| "grad_norm": 0.4743252396583557, | |
| "learning_rate": 0.0001937058638775353, | |
| "loss": 0.6232, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.38650548458099365, | |
| "learning_rate": 0.00019346415118073632, | |
| "loss": 1.2282, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5470085470085471, | |
| "grad_norm": 0.5221203565597534, | |
| "learning_rate": 0.00019321804092009906, | |
| "loss": 0.6979, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.4398798644542694, | |
| "learning_rate": 0.00019296754467531014, | |
| "loss": 0.6671, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5641025641025641, | |
| "grad_norm": 0.5590323209762573, | |
| "learning_rate": 0.00019271267423242024, | |
| "loss": 0.6642, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5726495726495726, | |
| "grad_norm": 0.512601912021637, | |
| "learning_rate": 0.00019245344158328972, | |
| "loss": 0.7476, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5811965811965812, | |
| "grad_norm": 0.49666309356689453, | |
| "learning_rate": 0.0001921898589250242, | |
| "loss": 0.6581, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5897435897435898, | |
| "grad_norm": 0.48546653985977173, | |
| "learning_rate": 0.0001919219386594007, | |
| "loss": 0.7066, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5982905982905983, | |
| "grad_norm": 0.46102452278137207, | |
| "learning_rate": 0.00019164969339228422, | |
| "loss": 0.7214, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6068376068376068, | |
| "grad_norm": 0.5350074768066406, | |
| "learning_rate": 0.00019137313593303463, | |
| "loss": 0.7814, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.4764799177646637, | |
| "learning_rate": 0.00019109227929390378, | |
| "loss": 1.3498, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6239316239316239, | |
| "grad_norm": 0.4472093880176544, | |
| "learning_rate": 0.00019080713668942356, | |
| "loss": 0.6558, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6324786324786325, | |
| "grad_norm": 0.4892171025276184, | |
| "learning_rate": 0.00019051772153578389, | |
| "loss": 0.663, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6410256410256411, | |
| "grad_norm": 0.46784690022468567, | |
| "learning_rate": 0.00019022404745020163, | |
| "loss": 0.6452, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6495726495726496, | |
| "grad_norm": 0.4115755558013916, | |
| "learning_rate": 0.00018992612825027976, | |
| "loss": 0.6968, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6581196581196581, | |
| "grad_norm": 0.5248686075210571, | |
| "learning_rate": 0.0001896239779533575, | |
| "loss": 0.8097, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.4202110171318054, | |
| "learning_rate": 0.00018931761077585035, | |
| "loss": 0.6617, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6752136752136753, | |
| "grad_norm": 0.418204128742218, | |
| "learning_rate": 0.00018900704113258165, | |
| "loss": 0.6448, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 0.43573886156082153, | |
| "learning_rate": 0.00018869228363610404, | |
| "loss": 0.6327, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.48871275782585144, | |
| "learning_rate": 0.00018837335309601213, | |
| "loss": 0.7072, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7008547008547008, | |
| "grad_norm": 0.4110111892223358, | |
| "learning_rate": 0.00018805026451824546, | |
| "loss": 1.2326, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7094017094017094, | |
| "grad_norm": 0.40106821060180664, | |
| "learning_rate": 0.00018772303310438275, | |
| "loss": 1.1263, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.717948717948718, | |
| "grad_norm": 0.4688105583190918, | |
| "learning_rate": 0.00018739167425092644, | |
| "loss": 1.1007, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7264957264957265, | |
| "grad_norm": 0.4649762809276581, | |
| "learning_rate": 0.00018705620354857833, | |
| "loss": 0.5994, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7350427350427351, | |
| "grad_norm": 0.46150311827659607, | |
| "learning_rate": 0.00018671663678150607, | |
| "loss": 0.5989, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7435897435897436, | |
| "grad_norm": 0.46822255849838257, | |
| "learning_rate": 0.0001863729899266004, | |
| "loss": 0.6286, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7521367521367521, | |
| "grad_norm": 0.5017993450164795, | |
| "learning_rate": 0.0001860252791527236, | |
| "loss": 1.0247, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7606837606837606, | |
| "grad_norm": 0.3973097801208496, | |
| "learning_rate": 0.00018567352081994852, | |
| "loss": 1.1795, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.410843163728714, | |
| "learning_rate": 0.00018531773147878895, | |
| "loss": 1.0421, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.4612373113632202, | |
| "learning_rate": 0.0001849579278694209, | |
| "loss": 0.6935, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7863247863247863, | |
| "grad_norm": 0.46664804220199585, | |
| "learning_rate": 0.00018459412692089494, | |
| "loss": 0.6416, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7948717948717948, | |
| "grad_norm": 0.4491286277770996, | |
| "learning_rate": 0.0001842263457503397, | |
| "loss": 0.5875, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8034188034188035, | |
| "grad_norm": 0.5161553621292114, | |
| "learning_rate": 0.00018385460166215638, | |
| "loss": 0.8172, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.811965811965812, | |
| "grad_norm": 0.500079870223999, | |
| "learning_rate": 0.00018347891214720477, | |
| "loss": 0.6465, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8205128205128205, | |
| "grad_norm": 0.3862932622432709, | |
| "learning_rate": 0.00018309929488198012, | |
| "loss": 1.1162, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8290598290598291, | |
| "grad_norm": 0.44776979088783264, | |
| "learning_rate": 0.00018271576772778154, | |
| "loss": 0.6599, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8376068376068376, | |
| "grad_norm": 0.4439445734024048, | |
| "learning_rate": 0.00018232834872987147, | |
| "loss": 1.1616, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.491223007440567, | |
| "learning_rate": 0.00018193705611662696, | |
| "loss": 0.7391, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 0.5422267913818359, | |
| "learning_rate": 0.0001815419082986815, | |
| "loss": 0.7363, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8632478632478633, | |
| "grad_norm": 0.4128502905368805, | |
| "learning_rate": 0.00018114292386805936, | |
| "loss": 1.0206, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8717948717948718, | |
| "grad_norm": 0.3821941316127777, | |
| "learning_rate": 0.00018074012159730032, | |
| "loss": 1.1002, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8803418803418803, | |
| "grad_norm": 0.46259158849716187, | |
| "learning_rate": 0.00018033352043857675, | |
| "loss": 0.9668, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.42356961965560913, | |
| "learning_rate": 0.00017992313952280172, | |
| "loss": 1.0531, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8974358974358975, | |
| "grad_norm": 0.45854252576828003, | |
| "learning_rate": 0.00017950899815872892, | |
| "loss": 1.1803, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.905982905982906, | |
| "grad_norm": 0.5974142551422119, | |
| "learning_rate": 0.00017909111583204422, | |
| "loss": 1.0416, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9145299145299145, | |
| "grad_norm": 0.6633700728416443, | |
| "learning_rate": 0.0001786695122044487, | |
| "loss": 0.8394, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.4408910274505615, | |
| "learning_rate": 0.0001782442071127338, | |
| "loss": 1.0641, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9316239316239316, | |
| "grad_norm": 0.6098408699035645, | |
| "learning_rate": 0.0001778152205678477, | |
| "loss": 1.0964, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9401709401709402, | |
| "grad_norm": 0.5177589654922485, | |
| "learning_rate": 0.00017738257275395404, | |
| "loss": 0.767, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9487179487179487, | |
| "grad_norm": 0.46795642375946045, | |
| "learning_rate": 0.00017694628402748202, | |
| "loss": 0.6871, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9572649572649573, | |
| "grad_norm": 0.5054307579994202, | |
| "learning_rate": 0.0001765063749161688, | |
| "loss": 0.7332, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9658119658119658, | |
| "grad_norm": 0.4986553490161896, | |
| "learning_rate": 0.00017606286611809353, | |
| "loss": 0.7099, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.9743589743589743, | |
| "grad_norm": 0.5066249966621399, | |
| "learning_rate": 0.00017561577850070355, | |
| "loss": 0.8318, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.9829059829059829, | |
| "grad_norm": 0.4113486707210541, | |
| "learning_rate": 0.00017516513309983253, | |
| "loss": 0.6118, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9914529914529915, | |
| "grad_norm": 0.5578283667564392, | |
| "learning_rate": 0.00017471095111871074, | |
| "loss": 0.7386, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4336143732070923, | |
| "learning_rate": 0.0001742532539269674, | |
| "loss": 0.631, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0085470085470085, | |
| "grad_norm": 0.4282875955104828, | |
| "learning_rate": 0.00017379206305962526, | |
| "loss": 0.5196, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.017094017094017, | |
| "grad_norm": 0.49017685651779175, | |
| "learning_rate": 0.00017332740021608722, | |
| "loss": 0.5137, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 0.48112180829048157, | |
| "learning_rate": 0.00017285928725911562, | |
| "loss": 0.5312, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0341880341880343, | |
| "grad_norm": 0.5431344509124756, | |
| "learning_rate": 0.00017238774621380337, | |
| "loss": 0.5791, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0427350427350428, | |
| "grad_norm": 0.5359887480735779, | |
| "learning_rate": 0.00017191279926653761, | |
| "loss": 1.1568, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0512820512820513, | |
| "grad_norm": 0.5204887390136719, | |
| "learning_rate": 0.00017143446876395602, | |
| "loss": 0.973, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0598290598290598, | |
| "grad_norm": 0.5024672150611877, | |
| "learning_rate": 0.00017095277721189528, | |
| "loss": 0.9468, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.0683760683760684, | |
| "grad_norm": 0.5556107759475708, | |
| "learning_rate": 0.00017046774727433222, | |
| "loss": 0.6045, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.5047200918197632, | |
| "learning_rate": 0.00016997940177231722, | |
| "loss": 0.5355, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.0854700854700854, | |
| "grad_norm": 0.4635385274887085, | |
| "learning_rate": 0.00016948776368290084, | |
| "loss": 1.0928, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0940170940170941, | |
| "grad_norm": 0.4944803714752197, | |
| "learning_rate": 0.00016899285613805246, | |
| "loss": 0.4621, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.1025641025641026, | |
| "grad_norm": 0.5671278834342957, | |
| "learning_rate": 0.00016849470242357196, | |
| "loss": 0.5215, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.5651219487190247, | |
| "learning_rate": 0.00016799332597799413, | |
| "loss": 0.5574, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1196581196581197, | |
| "grad_norm": 0.8755828738212585, | |
| "learning_rate": 0.00016748875039148593, | |
| "loss": 0.8745, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1282051282051282, | |
| "grad_norm": 0.5328382253646851, | |
| "learning_rate": 0.0001669809994047364, | |
| "loss": 0.9376, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1367521367521367, | |
| "grad_norm": 0.5357273817062378, | |
| "learning_rate": 0.0001664700969078398, | |
| "loss": 0.5562, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1452991452991452, | |
| "grad_norm": 0.4480111300945282, | |
| "learning_rate": 0.00016595606693917142, | |
| "loss": 0.9475, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.5380228757858276, | |
| "learning_rate": 0.00016543893368425666, | |
| "loss": 0.5098, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1623931623931625, | |
| "grad_norm": 0.5581599473953247, | |
| "learning_rate": 0.00016491872147463306, | |
| "loss": 0.4822, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.170940170940171, | |
| "grad_norm": 0.6443746089935303, | |
| "learning_rate": 0.00016439545478670543, | |
| "loss": 1.4373, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.1794871794871795, | |
| "grad_norm": 0.5481683611869812, | |
| "learning_rate": 0.00016386915824059427, | |
| "loss": 0.4893, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.188034188034188, | |
| "grad_norm": 0.4814392626285553, | |
| "learning_rate": 0.00016333985659897735, | |
| "loss": 0.4537, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1965811965811965, | |
| "grad_norm": 0.6673360466957092, | |
| "learning_rate": 0.00016280757476592466, | |
| "loss": 0.5764, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.205128205128205, | |
| "grad_norm": 0.5545169711112976, | |
| "learning_rate": 0.0001622723377857265, | |
| "loss": 0.5101, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2136752136752136, | |
| "grad_norm": 0.5839561820030212, | |
| "learning_rate": 0.00016173417084171536, | |
| "loss": 1.0742, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.5815286636352539, | |
| "learning_rate": 0.00016119309925508078, | |
| "loss": 0.4653, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.6954501867294312, | |
| "learning_rate": 0.0001606491484836782, | |
| "loss": 0.524, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.2393162393162394, | |
| "grad_norm": 0.6594898104667664, | |
| "learning_rate": 0.00016010234412083086, | |
| "loss": 0.5977, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2478632478632479, | |
| "grad_norm": 0.5977997779846191, | |
| "learning_rate": 0.00015955271189412598, | |
| "loss": 0.5959, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2564102564102564, | |
| "grad_norm": 0.6280649900436401, | |
| "learning_rate": 0.00015900027766420393, | |
| "loss": 0.4784, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.264957264957265, | |
| "grad_norm": 0.6320593953132629, | |
| "learning_rate": 0.00015844506742354164, | |
| "loss": 0.5877, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.2735042735042734, | |
| "grad_norm": 0.5736891031265259, | |
| "learning_rate": 0.00015788710729522953, | |
| "loss": 0.8057, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.282051282051282, | |
| "grad_norm": 0.8064723014831543, | |
| "learning_rate": 0.00015732642353174259, | |
| "loss": 0.899, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2905982905982907, | |
| "grad_norm": 0.5785207152366638, | |
| "learning_rate": 0.0001567630425137049, | |
| "loss": 0.4336, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.2991452991452992, | |
| "grad_norm": 0.7277658581733704, | |
| "learning_rate": 0.00015619699074864864, | |
| "loss": 0.5308, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.7484418153762817, | |
| "learning_rate": 0.00015562829486976673, | |
| "loss": 0.5529, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.3162393162393162, | |
| "grad_norm": 0.6898430585861206, | |
| "learning_rate": 0.00015505698163465986, | |
| "loss": 0.6197, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.3247863247863247, | |
| "grad_norm": 0.7089406847953796, | |
| "learning_rate": 0.00015448307792407734, | |
| "loss": 0.5021, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.6021527051925659, | |
| "learning_rate": 0.00015390661074065256, | |
| "loss": 0.5186, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.341880341880342, | |
| "grad_norm": 0.7267798185348511, | |
| "learning_rate": 0.00015332760720763232, | |
| "loss": 0.9724, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3504273504273505, | |
| "grad_norm": 0.7818686366081238, | |
| "learning_rate": 0.00015274609456760073, | |
| "loss": 0.6234, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.358974358974359, | |
| "grad_norm": 0.6912621259689331, | |
| "learning_rate": 0.00015216210018119733, | |
| "loss": 0.6456, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.3675213675213675, | |
| "grad_norm": 0.6164488792419434, | |
| "learning_rate": 0.00015157565152583002, | |
| "loss": 0.4671, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.376068376068376, | |
| "grad_norm": 0.6033616065979004, | |
| "learning_rate": 0.0001509867761943818, | |
| "loss": 0.7344, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.6088240146636963, | |
| "learning_rate": 0.00015039550189391298, | |
| "loss": 0.5298, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.393162393162393, | |
| "grad_norm": 0.6237228512763977, | |
| "learning_rate": 0.0001498018564443571, | |
| "loss": 0.792, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4017094017094016, | |
| "grad_norm": 0.7559857964515686, | |
| "learning_rate": 0.0001492058677772123, | |
| "loss": 0.6565, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4102564102564101, | |
| "grad_norm": 0.636070728302002, | |
| "learning_rate": 0.000148607563934227, | |
| "loss": 0.5548, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4188034188034189, | |
| "grad_norm": 0.48738163709640503, | |
| "learning_rate": 0.00014800697306608044, | |
| "loss": 0.4279, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4273504273504274, | |
| "grad_norm": 0.6380481719970703, | |
| "learning_rate": 0.00014740412343105828, | |
| "loss": 0.8164, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.435897435897436, | |
| "grad_norm": 0.5830632448196411, | |
| "learning_rate": 0.00014679904339372302, | |
| "loss": 0.4907, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.6315392851829529, | |
| "learning_rate": 0.00014619176142357935, | |
| "loss": 0.4647, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.452991452991453, | |
| "grad_norm": 0.6549903750419617, | |
| "learning_rate": 0.0001455823060937347, | |
| "loss": 0.5897, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.6652940511703491, | |
| "learning_rate": 0.00014497070607955476, | |
| "loss": 0.5145, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.4700854700854702, | |
| "grad_norm": 0.6678252220153809, | |
| "learning_rate": 0.00014435699015731448, | |
| "loss": 0.5141, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.4786324786324787, | |
| "grad_norm": 0.6491101384162903, | |
| "learning_rate": 0.00014374118720284388, | |
| "loss": 0.605, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.4871794871794872, | |
| "grad_norm": 0.547728955745697, | |
| "learning_rate": 0.00014312332619016965, | |
| "loss": 0.7185, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.4957264957264957, | |
| "grad_norm": 0.6576856970787048, | |
| "learning_rate": 0.0001425034361901516, | |
| "loss": 0.55, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.5042735042735043, | |
| "grad_norm": 0.7433821558952332, | |
| "learning_rate": 0.00014188154636911524, | |
| "loss": 1.1253, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5128205128205128, | |
| "grad_norm": 0.6701263189315796, | |
| "learning_rate": 0.0001412576859874791, | |
| "loss": 0.5022, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5213675213675213, | |
| "grad_norm": 0.6024699211120605, | |
| "learning_rate": 0.00014063188439837832, | |
| "loss": 0.4485, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5299145299145298, | |
| "grad_norm": 0.6067913770675659, | |
| "learning_rate": 0.0001400041710462833, | |
| "loss": 0.4409, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.6062759757041931, | |
| "learning_rate": 0.0001393745754656146, | |
| "loss": 0.5231, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.547008547008547, | |
| "grad_norm": 0.5581098794937134, | |
| "learning_rate": 0.00013874312727935292, | |
| "loss": 0.4631, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.6857714056968689, | |
| "learning_rate": 0.00013810985619764572, | |
| "loss": 0.9211, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.564102564102564, | |
| "grad_norm": 0.5674791932106018, | |
| "learning_rate": 0.00013747479201640914, | |
| "loss": 0.4676, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.5726495726495726, | |
| "grad_norm": 0.6834592223167419, | |
| "learning_rate": 0.00013683796461592604, | |
| "loss": 0.5552, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.5811965811965814, | |
| "grad_norm": 0.5935745239257812, | |
| "learning_rate": 0.00013619940395944027, | |
| "loss": 0.5349, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5897435897435899, | |
| "grad_norm": 0.6285943984985352, | |
| "learning_rate": 0.00013555914009174663, | |
| "loss": 0.5011, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.5982905982905984, | |
| "grad_norm": 0.6977561712265015, | |
| "learning_rate": 0.00013491720313777756, | |
| "loss": 0.8628, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.606837606837607, | |
| "grad_norm": 0.6670892834663391, | |
| "learning_rate": 0.00013427362330118543, | |
| "loss": 0.5293, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.6611135005950928, | |
| "learning_rate": 0.0001336284308629216, | |
| "loss": 0.6343, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.623931623931624, | |
| "grad_norm": 0.6946349740028381, | |
| "learning_rate": 0.00013298165617981172, | |
| "loss": 0.6632, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6324786324786325, | |
| "grad_norm": 0.7274101972579956, | |
| "learning_rate": 0.00013233332968312715, | |
| "loss": 0.7332, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.641025641025641, | |
| "grad_norm": 0.696225106716156, | |
| "learning_rate": 0.0001316834818771535, | |
| "loss": 0.5666, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.6495726495726495, | |
| "grad_norm": 0.7337691187858582, | |
| "learning_rate": 0.00013103214333775521, | |
| "loss": 0.5729, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.658119658119658, | |
| "grad_norm": 0.7068138122558594, | |
| "learning_rate": 0.00013037934471093682, | |
| "loss": 0.5094, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.7041600346565247, | |
| "learning_rate": 0.00012972511671140125, | |
| "loss": 0.4844, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6752136752136753, | |
| "grad_norm": 0.7568795084953308, | |
| "learning_rate": 0.00012906949012110456, | |
| "loss": 0.727, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.6837606837606838, | |
| "grad_norm": 0.7036489248275757, | |
| "learning_rate": 0.00012841249578780757, | |
| "loss": 0.6276, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.5621996521949768, | |
| "learning_rate": 0.00012775416462362457, | |
| "loss": 0.5462, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7008547008547008, | |
| "grad_norm": 0.6641893982887268, | |
| "learning_rate": 0.00012709452760356884, | |
| "loss": 1.1367, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7094017094017095, | |
| "grad_norm": 0.6324275732040405, | |
| "learning_rate": 0.00012643361576409516, | |
| "loss": 0.5462, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.717948717948718, | |
| "grad_norm": 0.674049973487854, | |
| "learning_rate": 0.00012577146020163968, | |
| "loss": 0.5571, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.7264957264957266, | |
| "grad_norm": 0.726768970489502, | |
| "learning_rate": 0.00012510809207115666, | |
| "loss": 0.5019, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.735042735042735, | |
| "grad_norm": 0.6050737500190735, | |
| "learning_rate": 0.00012444354258465268, | |
| "loss": 0.4863, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.7435897435897436, | |
| "grad_norm": 0.7358708381652832, | |
| "learning_rate": 0.00012377784300971807, | |
| "loss": 0.78, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.7521367521367521, | |
| "grad_norm": 0.6856046319007874, | |
| "learning_rate": 0.0001231110246680558, | |
| "loss": 1.0646, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.7606837606837606, | |
| "grad_norm": 0.7350341081619263, | |
| "learning_rate": 0.00012244311893400763, | |
| "loss": 0.5927, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.7623452544212341, | |
| "learning_rate": 0.00012177415723307808, | |
| "loss": 0.549, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.588233470916748, | |
| "learning_rate": 0.00012110417104045575, | |
| "loss": 0.4276, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.7863247863247862, | |
| "grad_norm": 0.6118550300598145, | |
| "learning_rate": 0.00012043319187953241, | |
| "loss": 0.5884, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.7948717948717947, | |
| "grad_norm": 0.6763346195220947, | |
| "learning_rate": 0.00011976125132041974, | |
| "loss": 0.5695, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8034188034188035, | |
| "grad_norm": 0.8064243793487549, | |
| "learning_rate": 0.00011908838097846404, | |
| "loss": 0.6683, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.811965811965812, | |
| "grad_norm": 0.6810288429260254, | |
| "learning_rate": 0.00011841461251275867, | |
| "loss": 0.7366, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.8205128205128205, | |
| "grad_norm": 0.7595717310905457, | |
| "learning_rate": 0.00011773997762465429, | |
| "loss": 0.5537, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.8290598290598292, | |
| "grad_norm": 0.6144272089004517, | |
| "learning_rate": 0.0001170645080562676, | |
| "loss": 0.5992, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.8376068376068377, | |
| "grad_norm": 0.5802556872367859, | |
| "learning_rate": 0.00011638823558898762, | |
| "loss": 0.4745, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.6331381797790527, | |
| "learning_rate": 0.00011571119204198037, | |
| "loss": 0.4894, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.8547008547008548, | |
| "grad_norm": 0.6273657083511353, | |
| "learning_rate": 0.00011503340927069189, | |
| "loss": 0.6066, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.8632478632478633, | |
| "grad_norm": 0.5358060598373413, | |
| "learning_rate": 0.00011435491916534919, | |
| "loss": 0.5186, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.8717948717948718, | |
| "grad_norm": 0.6037499904632568, | |
| "learning_rate": 0.00011367575364946006, | |
| "loss": 0.4939, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.8803418803418803, | |
| "grad_norm": 0.6647982001304626, | |
| "learning_rate": 0.00011299594467831078, | |
| "loss": 0.511, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.7059252858161926, | |
| "learning_rate": 0.00011231552423746283, | |
| "loss": 0.5783, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.8974358974358974, | |
| "grad_norm": 0.5569174289703369, | |
| "learning_rate": 0.00011163452434124773, | |
| "loss": 0.4994, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.9059829059829059, | |
| "grad_norm": 0.5955557227134705, | |
| "learning_rate": 0.00011095297703126093, | |
| "loss": 0.4769, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.9145299145299144, | |
| "grad_norm": 0.6245276927947998, | |
| "learning_rate": 0.00011027091437485404, | |
| "loss": 0.5901, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.6680919528007507, | |
| "learning_rate": 0.00010958836846362621, | |
| "loss": 0.6266, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9316239316239316, | |
| "grad_norm": 0.6182616353034973, | |
| "learning_rate": 0.00010890537141191417, | |
| "loss": 0.4838, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.9401709401709402, | |
| "grad_norm": 0.7662301659584045, | |
| "learning_rate": 0.00010822195535528106, | |
| "loss": 0.7301, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.9487179487179487, | |
| "grad_norm": 0.5705377459526062, | |
| "learning_rate": 0.00010753815244900458, | |
| "loss": 0.469, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.9572649572649574, | |
| "grad_norm": 0.6762925386428833, | |
| "learning_rate": 0.00010685399486656406, | |
| "loss": 0.5276, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.965811965811966, | |
| "grad_norm": 0.6223552227020264, | |
| "learning_rate": 0.00010616951479812658, | |
| "loss": 0.5262, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9743589743589745, | |
| "grad_norm": 0.662918746471405, | |
| "learning_rate": 0.00010548474444903247, | |
| "loss": 0.4567, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.982905982905983, | |
| "grad_norm": 0.6021672487258911, | |
| "learning_rate": 0.00010479971603828, | |
| "loss": 0.4697, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.9914529914529915, | |
| "grad_norm": 0.5436458587646484, | |
| "learning_rate": 0.00010411446179700943, | |
| "loss": 0.4631, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8974222540855408, | |
| "learning_rate": 0.00010342901396698659, | |
| "loss": 0.5085, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.0085470085470085, | |
| "grad_norm": 0.582740843296051, | |
| "learning_rate": 0.00010274340479908568, | |
| "loss": 0.2909, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.017094017094017, | |
| "grad_norm": 0.5627701282501221, | |
| "learning_rate": 0.00010205766655177215, | |
| "loss": 0.3259, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.0256410256410255, | |
| "grad_norm": 0.5415629148483276, | |
| "learning_rate": 0.00010137183148958463, | |
| "loss": 0.4473, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.034188034188034, | |
| "grad_norm": 0.6142169833183289, | |
| "learning_rate": 0.00010068593188161697, | |
| "loss": 0.2966, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.0427350427350426, | |
| "grad_norm": 0.7116908431053162, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3941, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 0.6427211165428162, | |
| "learning_rate": 9.931406811838308e-05, | |
| "loss": 0.2969, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0598290598290596, | |
| "grad_norm": 0.6511048674583435, | |
| "learning_rate": 9.862816851041541e-05, | |
| "loss": 0.2805, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.0683760683760686, | |
| "grad_norm": 0.6265400648117065, | |
| "learning_rate": 9.79423334482279e-05, | |
| "loss": 0.4247, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.7180808186531067, | |
| "learning_rate": 9.725659520091433e-05, | |
| "loss": 0.3414, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.0854700854700856, | |
| "grad_norm": 0.7465853095054626, | |
| "learning_rate": 9.657098603301346e-05, | |
| "loss": 0.4229, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.094017094017094, | |
| "grad_norm": 0.7189355492591858, | |
| "learning_rate": 9.588553820299056e-05, | |
| "loss": 0.3304, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.1025641025641026, | |
| "grad_norm": 0.8560303449630737, | |
| "learning_rate": 9.520028396172003e-05, | |
| "loss": 0.5008, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 0.8229368329048157, | |
| "learning_rate": 9.451525555096753e-05, | |
| "loss": 0.3663, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.1196581196581197, | |
| "grad_norm": 0.6783272624015808, | |
| "learning_rate": 9.383048520187344e-05, | |
| "loss": 0.2915, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.128205128205128, | |
| "grad_norm": 0.7634493708610535, | |
| "learning_rate": 9.314600513343595e-05, | |
| "loss": 0.3216, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.1367521367521367, | |
| "grad_norm": 0.8328527212142944, | |
| "learning_rate": 9.246184755099545e-05, | |
| "loss": 0.7065, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.1452991452991452, | |
| "grad_norm": 0.8876810073852539, | |
| "learning_rate": 9.177804464471898e-05, | |
| "loss": 0.7159, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.9150386452674866, | |
| "learning_rate": 9.109462858808586e-05, | |
| "loss": 0.7384, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.1623931623931623, | |
| "grad_norm": 0.6746718287467957, | |
| "learning_rate": 9.041163153637381e-05, | |
| "loss": 0.3506, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.1709401709401708, | |
| "grad_norm": 0.8253422379493713, | |
| "learning_rate": 8.972908562514598e-05, | |
| "loss": 0.3898, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.1794871794871793, | |
| "grad_norm": 0.8796607851982117, | |
| "learning_rate": 8.904702296873912e-05, | |
| "loss": 0.3889, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.1880341880341883, | |
| "grad_norm": 0.7309861183166504, | |
| "learning_rate": 8.836547565875227e-05, | |
| "loss": 0.3186, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.1965811965811968, | |
| "grad_norm": 0.7765249609947205, | |
| "learning_rate": 8.76844757625372e-05, | |
| "loss": 0.9032, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.2051282051282053, | |
| "grad_norm": 0.7523384094238281, | |
| "learning_rate": 8.70040553216892e-05, | |
| "loss": 0.3416, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.213675213675214, | |
| "grad_norm": 0.8088672757148743, | |
| "learning_rate": 8.632424635053997e-05, | |
| "loss": 0.4688, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.6530531644821167, | |
| "learning_rate": 8.564508083465079e-05, | |
| "loss": 0.3132, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.7539976239204407, | |
| "learning_rate": 8.496659072930813e-05, | |
| "loss": 0.3367, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.2393162393162394, | |
| "grad_norm": 0.7007617950439453, | |
| "learning_rate": 8.428880795801965e-05, | |
| "loss": 0.2973, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.247863247863248, | |
| "grad_norm": 1.2382545471191406, | |
| "learning_rate": 8.36117644110124e-05, | |
| "loss": 0.4853, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.2564102564102564, | |
| "grad_norm": 0.7213827967643738, | |
| "learning_rate": 8.293549194373243e-05, | |
| "loss": 0.2965, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.264957264957265, | |
| "grad_norm": 0.6621086001396179, | |
| "learning_rate": 8.226002237534572e-05, | |
| "loss": 0.3074, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.2735042735042734, | |
| "grad_norm": 1.5596137046813965, | |
| "learning_rate": 8.158538748724139e-05, | |
| "loss": 0.5787, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.282051282051282, | |
| "grad_norm": 0.9275439381599426, | |
| "learning_rate": 8.091161902153595e-05, | |
| "loss": 0.3738, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.2905982905982905, | |
| "grad_norm": 0.7352764010429382, | |
| "learning_rate": 8.023874867958027e-05, | |
| "loss": 0.6238, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.299145299145299, | |
| "grad_norm": 1.0265077352523804, | |
| "learning_rate": 7.95668081204676e-05, | |
| "loss": 0.4505, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.697374701499939, | |
| "learning_rate": 7.889582895954427e-05, | |
| "loss": 0.3394, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.316239316239316, | |
| "grad_norm": 0.6912069916725159, | |
| "learning_rate": 7.822584276692191e-05, | |
| "loss": 0.2912, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.324786324786325, | |
| "grad_norm": 0.7488755583763123, | |
| "learning_rate": 7.755688106599241e-05, | |
| "loss": 0.3848, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.798455536365509, | |
| "learning_rate": 7.688897533194424e-05, | |
| "loss": 0.3819, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.341880341880342, | |
| "grad_norm": 0.8598634004592896, | |
| "learning_rate": 7.622215699028196e-05, | |
| "loss": 0.3361, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.3504273504273505, | |
| "grad_norm": 0.6687372326850891, | |
| "learning_rate": 7.555645741534736e-05, | |
| "loss": 0.3035, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.358974358974359, | |
| "grad_norm": 1.1965564489364624, | |
| "learning_rate": 7.489190792884338e-05, | |
| "loss": 0.6589, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.3675213675213675, | |
| "grad_norm": 0.9742981195449829, | |
| "learning_rate": 7.422853979836034e-05, | |
| "loss": 0.5927, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.376068376068376, | |
| "grad_norm": 0.6811902523040771, | |
| "learning_rate": 7.356638423590485e-05, | |
| "loss": 0.3195, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.7968785762786865, | |
| "learning_rate": 7.290547239643117e-05, | |
| "loss": 0.3516, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.393162393162393, | |
| "grad_norm": 0.7076959013938904, | |
| "learning_rate": 7.224583537637544e-05, | |
| "loss": 0.3486, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.4017094017094016, | |
| "grad_norm": 1.0984399318695068, | |
| "learning_rate": 7.158750421219244e-05, | |
| "loss": 0.8063, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.41025641025641, | |
| "grad_norm": 0.8261274099349976, | |
| "learning_rate": 7.093050987889547e-05, | |
| "loss": 0.3128, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.4188034188034186, | |
| "grad_norm": 0.784706175327301, | |
| "learning_rate": 7.027488328859876e-05, | |
| "loss": 0.3127, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.427350427350427, | |
| "grad_norm": 0.7766672968864441, | |
| "learning_rate": 6.96206552890632e-05, | |
| "loss": 0.3918, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.435897435897436, | |
| "grad_norm": 0.868733286857605, | |
| "learning_rate": 6.896785666224481e-05, | |
| "loss": 0.3289, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.9634539484977722, | |
| "learning_rate": 6.831651812284652e-05, | |
| "loss": 0.3334, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.452991452991453, | |
| "grad_norm": 0.78970867395401, | |
| "learning_rate": 6.766667031687286e-05, | |
| "loss": 0.3411, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.8855091333389282, | |
| "learning_rate": 6.701834382018832e-05, | |
| "loss": 0.3223, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.47008547008547, | |
| "grad_norm": 0.8204054832458496, | |
| "learning_rate": 6.637156913707839e-05, | |
| "loss": 0.2914, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.4786324786324787, | |
| "grad_norm": 0.775288462638855, | |
| "learning_rate": 6.572637669881458e-05, | |
| "loss": 0.3122, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4871794871794872, | |
| "grad_norm": 1.1425946950912476, | |
| "learning_rate": 6.508279686222243e-05, | |
| "loss": 0.6664, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.4957264957264957, | |
| "grad_norm": 0.8591251373291016, | |
| "learning_rate": 6.444085990825338e-05, | |
| "loss": 0.3222, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.5042735042735043, | |
| "grad_norm": 0.7236760258674622, | |
| "learning_rate": 6.380059604055974e-05, | |
| "loss": 0.3038, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.5128205128205128, | |
| "grad_norm": 1.1851248741149902, | |
| "learning_rate": 6.316203538407397e-05, | |
| "loss": 0.4802, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.5213675213675213, | |
| "grad_norm": 0.8601669073104858, | |
| "learning_rate": 6.252520798359092e-05, | |
| "loss": 0.3428, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.52991452991453, | |
| "grad_norm": 0.760200023651123, | |
| "learning_rate": 6.18901438023543e-05, | |
| "loss": 0.2776, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.7901735305786133, | |
| "learning_rate": 6.125687272064713e-05, | |
| "loss": 0.3185, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.547008547008547, | |
| "grad_norm": 0.8040541410446167, | |
| "learning_rate": 6.0625424534385425e-05, | |
| "loss": 0.3315, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 0.9885798096656799, | |
| "learning_rate": 5.9995828953716695e-05, | |
| "loss": 0.4315, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.564102564102564, | |
| "grad_norm": 0.7921980619430542, | |
| "learning_rate": 5.936811560162169e-05, | |
| "loss": 0.3407, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5726495726495724, | |
| "grad_norm": 0.770325243473053, | |
| "learning_rate": 5.87423140125209e-05, | |
| "loss": 0.2901, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.5811965811965814, | |
| "grad_norm": 0.6994922161102295, | |
| "learning_rate": 5.811845363088477e-05, | |
| "loss": 0.2576, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.58974358974359, | |
| "grad_norm": 0.8995521664619446, | |
| "learning_rate": 5.749656380984844e-05, | |
| "loss": 0.3407, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.5982905982905984, | |
| "grad_norm": 0.8891424536705017, | |
| "learning_rate": 5.687667380983037e-05, | |
| "loss": 0.3127, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.606837606837607, | |
| "grad_norm": 0.7055900692939758, | |
| "learning_rate": 5.625881279715615e-05, | |
| "loss": 0.293, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.9877127408981323, | |
| "learning_rate": 5.5643009842685554e-05, | |
| "loss": 0.7147, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.623931623931624, | |
| "grad_norm": 0.8487789034843445, | |
| "learning_rate": 5.502929392044528e-05, | |
| "loss": 0.3721, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.6324786324786325, | |
| "grad_norm": 0.8070208430290222, | |
| "learning_rate": 5.4417693906265365e-05, | |
| "loss": 0.262, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.641025641025641, | |
| "grad_norm": 0.9167418479919434, | |
| "learning_rate": 5.380823857642069e-05, | |
| "loss": 0.3475, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.6495726495726495, | |
| "grad_norm": 1.1027212142944336, | |
| "learning_rate": 5.3200956606277006e-05, | |
| "loss": 0.547, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.658119658119658, | |
| "grad_norm": 0.7494267821311951, | |
| "learning_rate": 5.259587656894174e-05, | |
| "loss": 0.2841, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 1.0230811834335327, | |
| "learning_rate": 5.199302693391959e-05, | |
| "loss": 0.3595, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.6752136752136755, | |
| "grad_norm": 1.4321858882904053, | |
| "learning_rate": 5.139243606577302e-05, | |
| "loss": 0.483, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.683760683760684, | |
| "grad_norm": 0.9643809199333191, | |
| "learning_rate": 5.0794132222787707e-05, | |
| "loss": 0.3236, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.832219660282135, | |
| "learning_rate": 5.019814355564292e-05, | |
| "loss": 0.3292, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.700854700854701, | |
| "grad_norm": 0.8470568656921387, | |
| "learning_rate": 4.960449810608705e-05, | |
| "loss": 0.4273, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.7094017094017095, | |
| "grad_norm": 0.7562103867530823, | |
| "learning_rate": 4.90132238056182e-05, | |
| "loss": 0.3075, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.717948717948718, | |
| "grad_norm": 0.941655158996582, | |
| "learning_rate": 4.8424348474170014e-05, | |
| "loss": 0.8919, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.7264957264957266, | |
| "grad_norm": 0.9257362484931946, | |
| "learning_rate": 4.783789981880267e-05, | |
| "loss": 0.3352, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.735042735042735, | |
| "grad_norm": 0.7269126772880554, | |
| "learning_rate": 4.725390543239929e-05, | |
| "loss": 0.249, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.7435897435897436, | |
| "grad_norm": 0.9031948447227478, | |
| "learning_rate": 4.667239279236768e-05, | |
| "loss": 0.4783, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.752136752136752, | |
| "grad_norm": 0.8617339134216309, | |
| "learning_rate": 4.609338925934743e-05, | |
| "loss": 0.3422, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.7606837606837606, | |
| "grad_norm": 1.2087595462799072, | |
| "learning_rate": 4.551692207592265e-05, | |
| "loss": 0.5638, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.7378338575363159, | |
| "learning_rate": 4.494301836534016e-05, | |
| "loss": 0.2687, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 0.9566189646720886, | |
| "learning_rate": 4.4371705130233275e-05, | |
| "loss": 0.3597, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.786324786324786, | |
| "grad_norm": 0.7672934532165527, | |
| "learning_rate": 4.380300925135138e-05, | |
| "loss": 0.284, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.7948717948717947, | |
| "grad_norm": 0.7886463403701782, | |
| "learning_rate": 4.3236957486295115e-05, | |
| "loss": 0.8525, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.8034188034188032, | |
| "grad_norm": 0.9983607530593872, | |
| "learning_rate": 4.267357646825746e-05, | |
| "loss": 0.7077, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.8119658119658117, | |
| "grad_norm": 0.8718456029891968, | |
| "learning_rate": 4.211289270477047e-05, | |
| "loss": 0.2778, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.8205128205128203, | |
| "grad_norm": 0.8060352802276611, | |
| "learning_rate": 4.1554932576458415e-05, | |
| "loss": 0.3293, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.8290598290598292, | |
| "grad_norm": 1.0714794397354126, | |
| "learning_rate": 4.0999722335796075e-05, | |
| "loss": 0.6121, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.8376068376068377, | |
| "grad_norm": 0.8349295258522034, | |
| "learning_rate": 4.044728810587406e-05, | |
| "loss": 0.2984, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.8635120987892151, | |
| "learning_rate": 3.989765587916914e-05, | |
| "loss": 0.2869, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.8547008547008548, | |
| "grad_norm": 0.8403009176254272, | |
| "learning_rate": 3.935085151632185e-05, | |
| "loss": 0.3936, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.8632478632478633, | |
| "grad_norm": 0.7469919919967651, | |
| "learning_rate": 3.8806900744919205e-05, | |
| "loss": 0.2383, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.871794871794872, | |
| "grad_norm": 1.0146729946136475, | |
| "learning_rate": 3.826582915828468e-05, | |
| "loss": 0.4251, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.8803418803418803, | |
| "grad_norm": 0.8476770520210266, | |
| "learning_rate": 3.7727662214273495e-05, | |
| "loss": 0.3068, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.6913571357727051, | |
| "learning_rate": 3.719242523407539e-05, | |
| "loss": 0.2446, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.8974358974358974, | |
| "grad_norm": 1.0405776500701904, | |
| "learning_rate": 3.666014340102268e-05, | |
| "loss": 0.6346, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.905982905982906, | |
| "grad_norm": 0.9416986703872681, | |
| "learning_rate": 3.613084175940578e-05, | |
| "loss": 0.6073, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.9145299145299144, | |
| "grad_norm": 0.8243383765220642, | |
| "learning_rate": 3.5604545213294616e-05, | |
| "loss": 0.6594, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.8304813504219055, | |
| "learning_rate": 3.508127852536698e-05, | |
| "loss": 0.5282, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.931623931623932, | |
| "grad_norm": 0.9133498072624207, | |
| "learning_rate": 3.456106631574336e-05, | |
| "loss": 0.4149, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.9401709401709404, | |
| "grad_norm": 0.7665524482727051, | |
| "learning_rate": 3.4043933060828605e-05, | |
| "loss": 0.3206, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.948717948717949, | |
| "grad_norm": 0.8704233169555664, | |
| "learning_rate": 3.352990309216022e-05, | |
| "loss": 0.6664, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.9572649572649574, | |
| "grad_norm": 1.0126529932022095, | |
| "learning_rate": 3.3019000595263574e-05, | |
| "loss": 0.3356, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.965811965811966, | |
| "grad_norm": 0.8295007944107056, | |
| "learning_rate": 3.251124960851408e-05, | |
| "loss": 0.3067, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.9743589743589745, | |
| "grad_norm": 0.9717605113983154, | |
| "learning_rate": 3.200667402200586e-05, | |
| "loss": 0.3353, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.982905982905983, | |
| "grad_norm": 0.8787976503372192, | |
| "learning_rate": 3.1505297576428075e-05, | |
| "loss": 0.3135, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.9914529914529915, | |
| "grad_norm": 0.7986027002334595, | |
| "learning_rate": 3.100714386194757e-05, | |
| "loss": 0.3116, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8115022778511047, | |
| "learning_rate": 3.0512236317099175e-05, | |
| "loss": 0.2797, | |
| "step": 351 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 468, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 117, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.670263710121984e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |