Training in progress, step 682, checkpoint

last-checkpoint/global_step682/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:b5beda4b56a7a91484f971692d427a2268d48a4cc9ee57641666e2ec1fcd54c4
+size 13476835648

last-checkpoint/global_step682/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0645ce8ff780cd6aa5742e24248f941d49e958be43bd82b8e8e5ba331e955463
+size 13476835648

last-checkpoint/global_step682/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:da3f3f56a8c54822a8a00d0b58a0c1455157957226eb390200a565625da58a36
+size 13476835648

last-checkpoint/global_step682/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:a89cd07d540b7e68f0b8e7e51a1c52a7e062cf9a598ea85a4f1bf593b36e4098
+size 13476835648

last-checkpoint/global_step682/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5302e9f937abbdc558c1157c893863879a77d337dca79ef76dceb8aaa281769b
+size 150693

last-checkpoint/global_step682/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:937de52210755d6f3f9a3465565f7a01685102749fbda3c0f183fd3a4833b14a
+size 150693

last-checkpoint/global_step682/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8bd21fa8d9e5187765f0c614f77fb6ace6c3376c82b76ac9267bd63c2474011
+size 150693

last-checkpoint/global_step682/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ee275be2f3515a14f6a8f95fa623c837a903f6a7c93edbd62bc7924fc37d264
+size 150693

last-checkpoint/latest

@@ -1 +1 @@
-global_step558
+global_step682

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f57988b40ea7809dc08638e5caf7cc644808696ba431653d54751fa8e2ee7bef
+oid sha256:b540a3bdebafca7694656f77d0e558804e111c4c3b9b258038b1eaccbe984eec
 size 4938985352

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9637b70c600dc8d9de7c223cc42a2f547cec4117ec96d7971ad9596f125f6d4b
+oid sha256:23f9ad07280adb2d4b75e2e7438927d267ba0ae53691cf890abe923d1bb78778
 size 4947390880

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:57ab4e25910ea5f9e07b0b0aa8590e1bbbeec9b69fd598de221f66b43383d7c9
+oid sha256:bebbcf1f86afcb2a4c347004a55e474b02a6c86a2ffd1b17318f2c0f863e1cad
 size 3590488816

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb3d4793d6c7fda650513b6f3d24fe7afd643475a576b0104a85cb659ed441d3
+oid sha256:b4cf818939bfcb1de93c78e79be7390fcae46bb5c39166bd571b1f3a8b2cdeb3
 size 15024

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e44567105884943a8351ea71e9a71011cceee34a8f2a067be69948ea21390f49
+oid sha256:408ca302fdad670605f637bd7faeea541858831c4f2318e3fca2d0d709c7b218
 size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c3a6f11c1c82a313f717b58d29d20267d0bfc621c79116ac6f752e0c1ba1f50
+oid sha256:df380dcd2c8305fefe9f4ba88e7536beb3b5e94d863ac6f29eae326ddfef0f32
 size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aaa97839e10563397740560600a86ccaf5a6fe1e5f41084453cab0f732a05570
+oid sha256:51f78c929d775d81c8dd9feb87fd329a39e895b5f8b1d2cdbede2cbd8b45f49e
 size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:177a5448d699047d35d72636914adc373df5f511600959de3a1ed1ce15a52556
+oid sha256:3ec08ea6383804e8c05c7ff772c4aed53351dc10413de8f46c3a1998585e65c7
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.7856,
+  "epoch": 2.1824,
   "eval_steps": 62,
-  "global_step": 558,
+  "global_step": 682,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -991,6 +991,233 @@
       "eval_samples_per_second": 7.588,
       "eval_steps_per_second": 0.493,
       "step": 558
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.3508645377160027,
+      "learning_rate": 2.2327790973871732e-07,
+      "logits/generated": -0.8774440884590149,
+      "logits/real": -1.3573105335235596,
+      "logps/generated": -351.45697021484375,
+      "logps/real": -119.85599517822266,
+      "loss": 0.1099,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.625948905944824,
+      "rewards/margins": 17.333721160888672,
+      "rewards/real": 10.707772254943848,
+      "step": 560
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 0.3654629141180253,
+      "learning_rate": 2.173396674584323e-07,
+      "logits/generated": -0.8835099339485168,
+      "logits/real": -1.4046037197113037,
+      "logps/generated": -346.0320739746094,
+      "logps/real": -121.26094818115234,
+      "loss": 0.1059,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.1836042404174805,
+      "rewards/margins": 16.947559356689453,
+      "rewards/real": 10.763954162597656,
+      "step": 570
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.3990780890154353,
+      "learning_rate": 2.1140142517814726e-07,
+      "logits/generated": -0.8519056439399719,
+      "logits/real": -1.4273611307144165,
+      "logps/generated": -358.67572021484375,
+      "logps/real": -100.94701385498047,
+      "loss": 0.102,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.960443019866943,
+      "rewards/margins": 16.782718658447266,
+      "rewards/real": 9.822273254394531,
+      "step": 580
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 0.38514946196243705,
+      "learning_rate": 2.0546318289786222e-07,
+      "logits/generated": -0.8117479085922241,
+      "logits/real": -1.3695944547653198,
+      "logps/generated": -354.63836669921875,
+      "logps/real": -92.74488830566406,
+      "loss": 0.0995,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -7.354495048522949,
+      "rewards/margins": 16.78274154663086,
+      "rewards/real": 9.428245544433594,
+      "step": 590
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3924023599162714,
+      "learning_rate": 1.9952494061757718e-07,
+      "logits/generated": -0.8672367930412292,
+      "logits/real": -1.42955482006073,
+      "logps/generated": -358.61041259765625,
+      "logps/real": -108.92237854003906,
+      "loss": 0.1051,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -7.023502349853516,
+      "rewards/margins": 17.467809677124023,
+      "rewards/real": 10.444307327270508,
+      "step": 600
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 0.4306414835630299,
+      "learning_rate": 1.9358669833729216e-07,
+      "logits/generated": -0.8639044761657715,
+      "logits/real": -1.4375250339508057,
+      "logps/generated": -356.2347412109375,
+      "logps/real": -110.20645904541016,
+      "loss": 0.1031,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.728466987609863,
+      "rewards/margins": 17.04363441467285,
+      "rewards/real": 10.315167427062988,
+      "step": 610
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.3444950108876968,
+      "learning_rate": 1.876484560570071e-07,
+      "logits/generated": -0.9002432823181152,
+      "logits/real": -1.4994289875030518,
+      "logps/generated": -355.31854248046875,
+      "logps/real": -111.2657699584961,
+      "loss": 0.1058,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.687992095947266,
+      "rewards/margins": 16.953479766845703,
+      "rewards/real": 10.265488624572754,
+      "step": 620
+    },
+    {
+      "epoch": 1.984,
+      "eval_logits/generated": -0.8820578455924988,
+      "eval_logits/real": -1.4055497646331787,
+      "eval_logps/generated": -357.4770202636719,
+      "eval_logps/real": -105.37969207763672,
+      "eval_loss": 0.10766133666038513,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/generated": -7.480791091918945,
+      "eval_rewards/margins": 17.628650665283203,
+      "eval_rewards/real": 10.14786148071289,
+      "eval_runtime": 37.0785,
+      "eval_samples_per_second": 5.394,
+      "eval_steps_per_second": 0.351,
+      "step": 620
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.3780935929863943,
+      "learning_rate": 1.8171021377672207e-07,
+      "logits/generated": -0.8811729550361633,
+      "logits/real": -1.3763405084609985,
+      "logps/generated": -336.79595947265625,
+      "logps/real": -113.61727142333984,
+      "loss": 0.1078,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.133264064788818,
+      "rewards/margins": 16.935718536376953,
+      "rewards/real": 10.802453994750977,
+      "step": 630
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.3519668839255736,
+      "learning_rate": 1.7577197149643706e-07,
+      "logits/generated": -0.843653678894043,
+      "logits/real": -1.4190059900283813,
+      "logps/generated": -348.73944091796875,
+      "logps/real": -98.12186431884766,
+      "loss": 0.1021,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -7.104917049407959,
+      "rewards/margins": 16.62804412841797,
+      "rewards/real": 9.523126602172852,
+      "step": 640
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.3518523871601281,
+      "learning_rate": 1.6983372921615202e-07,
+      "logits/generated": -0.8700889348983765,
+      "logits/real": -1.4218956232070923,
+      "logps/generated": -359.59912109375,
+      "logps/real": -104.86639404296875,
+      "loss": 0.0964,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.900285243988037,
+      "rewards/margins": 17.489273071289062,
+      "rewards/real": 10.588987350463867,
+      "step": 650
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.33640275650323853,
+      "learning_rate": 1.6389548693586697e-07,
+      "logits/generated": -0.9481474161148071,
+      "logits/real": -1.4541842937469482,
+      "logps/generated": -341.5894470214844,
+      "logps/real": -103.59893798828125,
+      "loss": 0.1018,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.416829586029053,
+      "rewards/margins": 16.72159194946289,
+      "rewards/real": 10.30476188659668,
+      "step": 660
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 0.3422223879953069,
+      "learning_rate": 1.5795724465558193e-07,
+      "logits/generated": -0.9395056962966919,
+      "logits/real": -1.5020724534988403,
+      "logps/generated": -336.66485595703125,
+      "logps/real": -115.76351165771484,
+      "loss": 0.0999,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -6.646719455718994,
+      "rewards/margins": 17.59234619140625,
+      "rewards/real": 10.945624351501465,
+      "step": 670
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.4031943667703452,
+      "learning_rate": 1.520190023752969e-07,
+      "logits/generated": -0.8811619877815247,
+      "logits/real": -1.4061198234558105,
+      "logps/generated": -358.12457275390625,
+      "logps/real": -102.46736907958984,
+      "loss": 0.0995,
+      "rewards/accuracies": 1.0,
+      "rewards/generated": -7.244171142578125,
+      "rewards/margins": 17.505325317382812,
+      "rewards/real": 10.261154174804688,
+      "step": 680
+    },
+    {
+      "epoch": 2.1824,
+      "eval_logits/generated": -0.8780685067176819,
+      "eval_logits/real": -1.4101550579071045,
+      "eval_logps/generated": -354.6165771484375,
+      "eval_logps/real": -105.18895721435547,
+      "eval_loss": 0.10742899030447006,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/generated": -7.194748878479004,
+      "eval_rewards/margins": 17.36168670654297,
+      "eval_rewards/real": 10.166935920715332,
+      "eval_runtime": 25.4941,
+      "eval_samples_per_second": 7.845,
+      "eval_steps_per_second": 0.51,
+      "step": 682
     }
   ],
   "logging_steps": 10,