Training in progress, epoch 3, checkpoint

last-checkpoint/global_step939/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5589178fb460218416d99611226366741d852e156dc4568df72ebba81341db7f
+size 15231238316

last-checkpoint/global_step939/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1697a79d1af1abe7c8749194dbaa3ab905ae7b3ae7b17a28286ccb5f6111807b
+size 15231238316

last-checkpoint/global_step939/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af326e1e2e016168d830f971644a733b853448d43d0bbe72c8fdd8054e40e1fd
+size 15231238316

last-checkpoint/global_step939/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76872aa85e4a81146d86f352af93960e7c08d2bb56304d4b4030a2b88cb4d345
+size 15231238316

last-checkpoint/global_step939/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99b1a3922c793b0c24d306251b0c1441e9f7a9cce14ecb23e7f75bc3c63f9af2
+size 167957

last-checkpoint/global_step939/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f71281b19ec81ffaa3d5f8398270ac4941b2ac635fe676ef886a45e5c46042cf
+size 167957

last-checkpoint/global_step939/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cbc34819d4db93c1a1169bbdce31406e3960e5dbe9738e8f1fc22a9e244180d
+size 167957

last-checkpoint/global_step939/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38e1230e3716503c60f04ce18fba282c382f9e189ddab60f41472b12e0e7d9c7
+size 167957

last-checkpoint/latest

@@ -1 +1 @@
-global_step626
+global_step939

last-checkpoint/model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5220eb9213d92273ad724abc5900289746d2cb77b95284ddfbc1789978c1c503
+oid sha256:a40d7be4dc87f6c09a609e4ee42df8e469cfeee851c4f699fbc5cbfc3fe8d456
 size 4877660776

last-checkpoint/model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a4ece371b15f6c0812d15e9a22f1404f87ba38802387a31efcfab55b66ca244
+oid sha256:92735b74728376dc5bc2f6d22e3362ffa846d056cce8bb762f76819bb1acc21d
 size 4932751008

last-checkpoint/model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70b70078ef2f4a26af456cc41dcdd4561f664904e45b74c9adf41eef88c7eff6
+oid sha256:9688067b0a3cb13619ccc8bf5cb94e5c585bd36e5d4071a99613e47ae5687ba6
 size 4330865200

last-checkpoint/model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2498f64d7e581f7a32717da94a3a00b35cd20e4dd7678b89a63ebe2a27dd2f48
+oid sha256:3d9f4472dcc1d160f0fbeaf813fe3a348f32b2d33407bc26a23b8f5e1e0638c3
 size 1089994880

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2307c03867cef25b5028feb9a23f80e784b9af9a615de13ddca560a6a90fb593
+oid sha256:ab7cf0d34d60dfda516cf9661904550e2e294e723edd07c25c738f05e8ba92d1
 size 15024

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50218cfaecdd818354e567b7167c13899e3b42297e7d8f58bd7e732cfa547800
+oid sha256:f7b183324e8227a51a9556d86b2ad893a8c4c52205ed4a737356c6611dac7353
 size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9814a66b49861f5495b06dae3be12ddf7185b88e2cae1fb808ca9efd99d5807f
+oid sha256:ac69e994090f4818cb1fa6f6cefa363178552c3c731c6507ff195bcb07fd5bef
 size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a7be93040a65e0a29975f6c70b94418e1fdf88423a50c58aa572141d3c92fbfc
+oid sha256:f68e61b63402f8afb1f69c960f7944965655dac11e3ccf29919c282f23931f86
 size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f7725b44c69725321786f1e58dccd7ea4d3ae5794ea47bd49c0d4a139aec266d
+oid sha256:fe7d17fbcfa5bee1bf6cb5aedab3ff5a70436912c200d7301d173be443809d63
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.0,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 626,
+  "global_step": 939,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -984,6 +984,487 @@
       "eval_samples_per_second": 17.858,
       "eval_steps_per_second": 0.625,
       "step": 626
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 0.00010862728759544983,
+      "learning_rate": 1.8284023668639053e-07,
+      "logits/chosen": -1.3046875,
+      "logits/rejected": -1.3125,
+      "logps/chosen": -112.0,
+      "logps/rejected": -410.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0,
+      "rewards/margins": 17.25,
+      "rewards/rejected": -16.25,
+      "step": 630
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 4.385121869847701e-05,
+      "learning_rate": 1.7692307692307693e-07,
+      "logits/chosen": -1.3359375,
+      "logits/rejected": -1.3515625,
+      "logps/chosen": -106.5,
+      "logps/rejected": -408.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.140625,
+      "rewards/margins": 17.5,
+      "rewards/rejected": -16.375,
+      "step": 640
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 0.0006704320313793596,
+      "learning_rate": 1.710059171597633e-07,
+      "logits/chosen": -1.3828125,
+      "logits/rejected": -1.2890625,
+      "logps/chosen": -109.5,
+      "logps/rejected": -412.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.984375,
+      "rewards/margins": 17.625,
+      "rewards/rejected": -16.75,
+      "step": 650
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 3.926751354406349e-05,
+      "learning_rate": 1.650887573964497e-07,
+      "logits/chosen": -1.296875,
+      "logits/rejected": -1.3984375,
+      "logps/chosen": -104.5,
+      "logps/rejected": -412.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.90625,
+      "rewards/margins": 17.875,
+      "rewards/rejected": -17.0,
+      "step": 660
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 7.442917181039221e-05,
+      "learning_rate": 1.591715976331361e-07,
+      "logits/chosen": -1.375,
+      "logits/rejected": -1.2890625,
+      "logps/chosen": -104.0,
+      "logps/rejected": -414.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.96875,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.25,
+      "step": 670
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 0.0001506499834402947,
+      "learning_rate": 1.5325443786982248e-07,
+      "logits/chosen": -1.2734375,
+      "logits/rejected": -1.3359375,
+      "logps/chosen": -104.5,
+      "logps/rejected": -420.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.96484375,
+      "rewards/margins": 18.625,
+      "rewards/rejected": -17.75,
+      "step": 680
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 4.3054154781333365e-05,
+      "learning_rate": 1.4733727810650885e-07,
+      "logits/chosen": -1.3984375,
+      "logits/rejected": -1.40625,
+      "logps/chosen": -103.5,
+      "logps/rejected": -416.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.1328125,
+      "rewards/margins": 18.125,
+      "rewards/rejected": -17.0,
+      "step": 690
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 3.607556222389406e-05,
+      "learning_rate": 1.4142011834319526e-07,
+      "logits/chosen": -1.40625,
+      "logits/rejected": -1.328125,
+      "logps/chosen": -112.5,
+      "logps/rejected": -412.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0703125,
+      "rewards/margins": 18.0,
+      "rewards/rejected": -16.875,
+      "step": 700
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 2.166580956783726e-05,
+      "learning_rate": 1.3550295857988164e-07,
+      "logits/chosen": -1.390625,
+      "logits/rejected": -1.34375,
+      "logps/chosen": -110.0,
+      "logps/rejected": -414.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0546875,
+      "rewards/margins": 18.0,
+      "rewards/rejected": -17.0,
+      "step": 710
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 0.0007768824282097016,
+      "learning_rate": 1.2958579881656802e-07,
+      "logits/chosen": -1.34375,
+      "logits/rejected": -1.328125,
+      "logps/chosen": -114.0,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.9921875,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.25,
+      "step": 720
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 1.3261401519699353e-05,
+      "learning_rate": 1.2366863905325443e-07,
+      "logits/chosen": -1.28125,
+      "logits/rejected": -1.3203125,
+      "logps/chosen": -103.0,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.140625,
+      "rewards/margins": 18.5,
+      "rewards/rejected": -17.375,
+      "step": 730
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 0.0005719833094700048,
+      "learning_rate": 1.1775147928994082e-07,
+      "logits/chosen": -1.40625,
+      "logits/rejected": -1.3671875,
+      "logps/chosen": -113.0,
+      "logps/rejected": -420.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0234375,
+      "rewards/margins": 18.625,
+      "rewards/rejected": -17.5,
+      "step": 740
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 4.7443275644552423e-05,
+      "learning_rate": 1.1183431952662721e-07,
+      "logits/chosen": -1.40625,
+      "logits/rejected": -1.3046875,
+      "logps/chosen": -118.0,
+      "logps/rejected": -416.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.796875,
+      "rewards/margins": 18.375,
+      "rewards/rejected": -17.625,
+      "step": 750
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 1.663293459020266e-05,
+      "learning_rate": 1.059171597633136e-07,
+      "logits/chosen": -1.3828125,
+      "logits/rejected": -1.2578125,
+      "logps/chosen": -104.0,
+      "logps/rejected": -428.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.96875,
+      "rewards/margins": 19.0,
+      "rewards/rejected": -18.0,
+      "step": 760
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 0.00010957271577923361,
+      "learning_rate": 1e-07,
+      "logits/chosen": -1.265625,
+      "logits/rejected": -1.2890625,
+      "logps/chosen": -90.0,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.1328125,
+      "rewards/margins": 18.5,
+      "rewards/rejected": -17.375,
+      "step": 770
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 0.0003475483172869348,
+      "learning_rate": 9.408284023668639e-08,
+      "logits/chosen": -1.2734375,
+      "logits/rejected": -1.296875,
+      "logps/chosen": -107.0,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.93359375,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.25,
+      "step": 780
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 3.990205582948748e-05,
+      "learning_rate": 8.816568047337278e-08,
+      "logits/chosen": -1.3828125,
+      "logits/rejected": -1.34375,
+      "logps/chosen": -102.5,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.109375,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.125,
+      "step": 790
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 3.0246347907445157e-05,
+      "learning_rate": 8.224852071005916e-08,
+      "logits/chosen": -1.5703125,
+      "logits/rejected": -1.3046875,
+      "logps/chosen": -120.0,
+      "logps/rejected": -414.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.94140625,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.375,
+      "step": 800
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 1.7382836083050494e-05,
+      "learning_rate": 7.633136094674555e-08,
+      "logits/chosen": -1.3671875,
+      "logits/rejected": -1.328125,
+      "logps/chosen": -102.5,
+      "logps/rejected": -420.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0625,
+      "rewards/margins": 18.5,
+      "rewards/rejected": -17.5,
+      "step": 810
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 1.1141783507026842e-05,
+      "learning_rate": 7.041420118343195e-08,
+      "logits/chosen": -1.4140625,
+      "logits/rejected": -1.359375,
+      "logps/chosen": -111.0,
+      "logps/rejected": -422.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0234375,
+      "rewards/margins": 18.625,
+      "rewards/rejected": -17.625,
+      "step": 820
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 5.6649362804211735e-05,
+      "learning_rate": 6.449704142011835e-08,
+      "logits/chosen": -1.4375,
+      "logits/rejected": -1.3359375,
+      "logps/chosen": -112.0,
+      "logps/rejected": -428.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0078125,
+      "rewards/margins": 19.25,
+      "rewards/rejected": -18.25,
+      "step": 830
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 4.1718570421510905e-05,
+      "learning_rate": 5.857988165680473e-08,
+      "logits/chosen": -1.3515625,
+      "logits/rejected": -1.375,
+      "logps/chosen": -104.0,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.140625,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.125,
+      "step": 840
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 4.584929902297974e-05,
+      "learning_rate": 5.266272189349112e-08,
+      "logits/chosen": -1.3203125,
+      "logits/rejected": -1.2890625,
+      "logps/chosen": -111.5,
+      "logps/rejected": -420.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.953125,
+      "rewards/margins": 18.5,
+      "rewards/rejected": -17.625,
+      "step": 850
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 0.0008958556766046752,
+      "learning_rate": 4.674556213017751e-08,
+      "logits/chosen": -1.3671875,
+      "logits/rejected": -1.3984375,
+      "logps/chosen": -107.5,
+      "logps/rejected": -414.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0625,
+      "rewards/margins": 18.25,
+      "rewards/rejected": -17.25,
+      "step": 860
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 5.93299351526953e-06,
+      "learning_rate": 4.082840236686391e-08,
+      "logits/chosen": -1.2734375,
+      "logits/rejected": -1.3125,
+      "logps/chosen": -117.0,
+      "logps/rejected": -416.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0390625,
+      "rewards/margins": 18.5,
+      "rewards/rejected": -17.5,
+      "step": 870
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 0.00014938750953581524,
+      "learning_rate": 3.4911242603550294e-08,
+      "logits/chosen": -1.375,
+      "logits/rejected": -1.359375,
+      "logps/chosen": -96.0,
+      "logps/rejected": -424.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.125,
+      "rewards/margins": 18.875,
+      "rewards/rejected": -17.75,
+      "step": 880
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 0.00013613017137118136,
+      "learning_rate": 2.8994082840236687e-08,
+      "logits/chosen": -1.4140625,
+      "logits/rejected": -1.2890625,
+      "logps/chosen": -118.0,
+      "logps/rejected": -418.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.9609375,
+      "rewards/margins": 18.375,
+      "rewards/rejected": -17.375,
+      "step": 890
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 0.0003341949440027235,
+      "learning_rate": 2.3076923076923076e-08,
+      "logits/chosen": -1.421875,
+      "logits/rejected": -1.3515625,
+      "logps/chosen": -110.0,
+      "logps/rejected": -426.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.09375,
+      "rewards/margins": 18.75,
+      "rewards/rejected": -17.625,
+      "step": 900
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 3.7993345789288195e-05,
+      "learning_rate": 1.7159763313609465e-08,
+      "logits/chosen": -1.359375,
+      "logits/rejected": -1.3515625,
+      "logps/chosen": -122.0,
+      "logps/rejected": -424.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 1.0,
+      "rewards/margins": 18.625,
+      "rewards/rejected": -17.625,
+      "step": 910
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 5.6917102751922324e-05,
+      "learning_rate": 1.1242603550295858e-08,
+      "logits/chosen": -1.2890625,
+      "logits/rejected": -1.34375,
+      "logps/chosen": -107.0,
+      "logps/rejected": -430.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.88671875,
+      "rewards/margins": 19.0,
+      "rewards/rejected": -18.0,
+      "step": 920
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 0.00018882592412388543,
+      "learning_rate": 5.325443786982248e-09,
+      "logits/chosen": -1.4375,
+      "logits/rejected": -1.3125,
+      "logps/chosen": -108.0,
+      "logps/rejected": -416.0,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.96484375,
+      "rewards/margins": 18.0,
+      "rewards/rejected": -17.125,
+      "step": 930
+    },
+    {
+      "epoch": 3.0,
+      "eval_logits/chosen": -1.2421875,
+      "eval_logits/rejected": -1.2734375,
+      "eval_logps/chosen": -91.5,
+      "eval_logps/rejected": -432.0,
+      "eval_loss": 2.9754883712485025e-07,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/chosen": 0.99609375,
+      "eval_rewards/margins": 19.125,
+      "eval_rewards/rejected": -18.125,
+      "eval_runtime": 13.8353,
+      "eval_samples_per_second": 14.456,
+      "eval_steps_per_second": 0.506,
+      "step": 939
     }
   ],
   "logging_steps": 10,
@@ -998,7 +1479,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }