Training in progress, step 310, checkpoint

last-checkpoint/global_step310/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60f8137d4761f8c2bd9ab99f5be668bd702519ce7ac64c78eda8e3fc2620c25a
+size 13476835648

last-checkpoint/global_step310/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85aaeb10b7115ba1c856de21eec3583959c79d28815cf8c670dca3b3f00c9bac
+size 13476835648

last-checkpoint/global_step310/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d956dae0da7c7df185dd20bd6f860ce87327578ca48626be6aa1254f9468aab
+size 13476835648

last-checkpoint/global_step310/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:448f39763b00dd35a0267adee83056d8b9b2e421836a35a5d3580a9a570c9075
+size 13476835648

last-checkpoint/global_step310/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dfaccb97630865bfa96a9f8fc8e1eaac2f641bb5486f7f1b34ba82548d9a217
+size 150693

last-checkpoint/global_step310/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c296a38893fbec54c8fb4c320b4ea180b982277ef291f058b5ffca0caf27bd2
+size 150693

last-checkpoint/global_step310/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a19de9850756f925991116920eef332af9868426f05cf741786bffa261e1b51
+size 150693

last-checkpoint/global_step310/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b775b848751a6b5c74313b60cd24e3e2d525b738411897a1a952fc106bb1799
+size 150693

last-checkpoint/latest

@@ -1 +1 @@
-global_step186
+global_step310

last-checkpoint/model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:313d5167651e3868509dd3619ee1c2d07051385b98649993872036ea95d4aba0
+oid sha256:7629983c481eb9200b556ac83ed8d209475fc4ef48f27d568a5cb34179c3b672
 size 4938985352

last-checkpoint/model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:845c39a62c9cb44d23ae87ad70b7d3a9dc92f88b09fb653b4375a2ca9ac136c3
+oid sha256:7708922bcf1e99c851124c87b0ec55e1c4b40e5c0cea43b42302293a17729487
 size 4947390880

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61a944543e23aed815241148692fce42317e0a4e455f648feb2320a8353247ad
+oid sha256:44492addf0ff01b4318cdfbe27b38d73decff5f3a716bd9a106eee463d36f582
 size 3590488816

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50ca27603f7c4a4d9a980223e1a102922ac31561acfa7da4f28042e4e8b3528b
+oid sha256:85a44fde7a8ca235658f035c0812d344f55779640520411739b17e214177e846
 size 15024

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c57f658c3d9b6de611da50021823095fc4067036fb9a4b34655cc2e34fbf5c2b
+oid sha256:2caad91201b0df46e56af9eecef567ed2037350c38fdc81837724a4727d0b050
 size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:109da2c363a1e09432fa2ea2b7dceb6f67d1f91975069fc669e71dd536d654d8
+oid sha256:73e988587366c84fd1b4fc5931610543c6da84cf501cd98c78745fd102dd5024
 size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bda65e490958bfe3fe01157ed131eb2354f8fc19536842d27c998ecf29d48991
+oid sha256:55af3bb04c01d99df727583b8b7a09241b4e47d58ff1b9cf5532d3fa0d878b52
 size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a4f13df3470221c87dbd73f49aa8200fcb840f387f1d9666c0df7b34d2fa99d1
+oid sha256:deebfd046c912c12a5ec3992f182a7898d0a095fd57e8891bf19ccf9e61ba445
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.5952,
+  "epoch": 0.992,
   "eval_steps": 62,
-  "global_step": 186,
+  "global_step": 310,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -340,6 +340,233 @@
       "eval_samples_per_second": 7.758,
       "eval_steps_per_second": 0.504,
       "step": 186
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.39320789077558344,
+      "learning_rate": 4.429928741092636e-07,
+      "logits/generated": -0.8406023979187012,
+      "logits/real": -1.2843071222305298,
+      "logps/generated": -316.73291015625,
+      "logps/real": -109.3148193359375,
+      "loss": 0.1103,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -4.513542175292969,
+      "rewards/margins": 14.111068725585938,
+      "rewards/real": 9.597526550292969,
+      "step": 190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34126559985788285,
+      "learning_rate": 4.3705463182897863e-07,
+      "logits/generated": -0.8079656362533569,
+      "logits/real": -1.2299845218658447,
+      "logps/generated": -333.931396484375,
+      "logps/real": -113.5406723022461,
+      "loss": 0.1079,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.271535873413086,
+      "rewards/margins": 14.859156608581543,
+      "rewards/real": 9.587618827819824,
+      "step": 200
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.37211551392957165,
+      "learning_rate": 4.311163895486936e-07,
+      "logits/generated": -0.8258484601974487,
+      "logits/real": -1.2792431116104126,
+      "logps/generated": -326.34539794921875,
+      "logps/real": -110.4327621459961,
+      "loss": 0.1062,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -4.544249534606934,
+      "rewards/margins": 13.822877883911133,
+      "rewards/real": 9.278627395629883,
+      "step": 210
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.40528624225351695,
+      "learning_rate": 4.251781472684085e-07,
+      "logits/generated": -0.9005807638168335,
+      "logits/real": -1.367760419845581,
+      "logps/generated": -334.90667724609375,
+      "logps/real": -108.89170837402344,
+      "loss": 0.1082,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.304292678833008,
+      "rewards/margins": 14.677830696105957,
+      "rewards/real": 9.37353801727295,
+      "step": 220
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.4299824458037775,
+      "learning_rate": 4.192399049881235e-07,
+      "logits/generated": -0.871414065361023,
+      "logits/real": -1.2680821418762207,
+      "logps/generated": -327.0166015625,
+      "logps/real": -125.11517333984375,
+      "loss": 0.1201,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.102071285247803,
+      "rewards/margins": 15.333366394042969,
+      "rewards/real": 10.231294631958008,
+      "step": 230
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.3200345894830024,
+      "learning_rate": 4.1330166270783846e-07,
+      "logits/generated": -0.8396976590156555,
+      "logits/real": -1.2651922702789307,
+      "logps/generated": -349.18804931640625,
+      "logps/real": -117.8864517211914,
+      "loss": 0.1094,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.743192672729492,
+      "rewards/margins": 15.722867965698242,
+      "rewards/real": 9.979674339294434,
+      "step": 240
+    },
+    {
+      "epoch": 0.7936,
+      "eval_logits/generated": -0.8538265228271484,
+      "eval_logits/real": -1.2560151815414429,
+      "eval_logps/generated": -337.5219421386719,
+      "eval_logps/real": -108.52310180664062,
+      "eval_loss": 0.11097312718629837,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/generated": -5.485285758972168,
+      "eval_rewards/margins": 15.318806648254395,
+      "eval_rewards/real": 9.83352279663086,
+      "eval_runtime": 32.0233,
+      "eval_samples_per_second": 6.245,
+      "eval_steps_per_second": 0.406,
+      "step": 248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37657892572843127,
+      "learning_rate": 4.0736342042755347e-07,
+      "logits/generated": -0.8381448984146118,
+      "logits/real": -1.293889045715332,
+      "logps/generated": -344.9915466308594,
+      "logps/real": -111.15998840332031,
+      "loss": 0.1113,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.526784896850586,
+      "rewards/margins": 15.003395080566406,
+      "rewards/real": 9.476609230041504,
+      "step": 250
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.3794924953311087,
+      "learning_rate": 4.0142517814726837e-07,
+      "logits/generated": -0.8616489171981812,
+      "logits/real": -1.342829704284668,
+      "logps/generated": -333.8689270019531,
+      "logps/real": -110.18806457519531,
+      "loss": 0.1068,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.346303939819336,
+      "rewards/margins": 15.174580574035645,
+      "rewards/real": 9.828276634216309,
+      "step": 260
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.38432288454636715,
+      "learning_rate": 3.9548693586698333e-07,
+      "logits/generated": -0.8341599702835083,
+      "logits/real": -1.2853684425354004,
+      "logps/generated": -340.5439147949219,
+      "logps/real": -111.25767517089844,
+      "loss": 0.1119,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.1574320793151855,
+      "rewards/margins": 14.693034172058105,
+      "rewards/real": 9.535603523254395,
+      "step": 270
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.4988860977100256,
+      "learning_rate": 3.8954869358669834e-07,
+      "logits/generated": -0.842201828956604,
+      "logits/real": -1.3558833599090576,
+      "logps/generated": -338.8299560546875,
+      "logps/real": -110.4022216796875,
+      "loss": 0.1111,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.072219371795654,
+      "rewards/margins": 14.635177612304688,
+      "rewards/real": 9.562958717346191,
+      "step": 280
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.38841887432580663,
+      "learning_rate": 3.836104513064133e-07,
+      "logits/generated": -0.8865805864334106,
+      "logits/real": -1.2532027959823608,
+      "logps/generated": -338.3769226074219,
+      "logps/real": -120.43833923339844,
+      "loss": 0.1084,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.086993217468262,
+      "rewards/margins": 15.214749336242676,
+      "rewards/real": 10.127756118774414,
+      "step": 290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3616732188669415,
+      "learning_rate": 3.7767220902612825e-07,
+      "logits/generated": -0.8655015826225281,
+      "logits/real": -1.3518953323364258,
+      "logps/generated": -325.9730529785156,
+      "logps/real": -95.09385681152344,
+      "loss": 0.1028,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.991705417633057,
+      "rewards/margins": 14.74760627746582,
+      "rewards/real": 8.755899429321289,
+      "step": 300
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.39359159204366234,
+      "learning_rate": 3.717339667458432e-07,
+      "logits/generated": -0.8543545603752136,
+      "logits/real": -1.3398587703704834,
+      "logps/generated": -338.96466064453125,
+      "logps/real": -116.5450668334961,
+      "loss": 0.1115,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -5.719244480133057,
+      "rewards/margins": 15.74243450164795,
+      "rewards/real": 10.02319049835205,
+      "step": 310
+    },
+    {
+      "epoch": 0.992,
+      "eval_logits/generated": -0.865822434425354,
+      "eval_logits/real": -1.33043372631073,
+      "eval_logps/generated": -347.4965515136719,
+      "eval_logps/real": -107.73165130615234,
+      "eval_loss": 0.11000669747591019,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/generated": -6.482741832733154,
+      "eval_rewards/margins": 16.39541244506836,
+      "eval_rewards/real": 9.912667274475098,
+      "eval_runtime": 24.7895,
+      "eval_samples_per_second": 8.068,
+      "eval_steps_per_second": 0.524,
+      "step": 310
     }
   ],
   "logging_steps": 10,