Training in progress, step 186, checkpoint

last-checkpoint/global_step186/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38fc2b45962b4fe9ddd0170c1b0a3d82aa00681907099e1652658cde8f12d2c0
+size 13476835648

last-checkpoint/global_step186/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95fcfc53b9d9b55532f66c305d07c2ecad49fd633ab906f90e5e800aec141723
+size 13476835648

last-checkpoint/global_step186/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:df2f3a9250d014253df9e66b657eb820431e3c7e2ebda85da1233a9cbebe4128
+size 13476835648

last-checkpoint/global_step186/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e2e6f8a3378649c2cbf7da97f08509ff0a279daef4d15e17a5db1e6791b7e65
+size 13476835648

last-checkpoint/global_step186/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f30787e32a351f2fafa763bc3112306c77dd360bd809c192f5a312a49f34940c
+size 150693

last-checkpoint/global_step186/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3505ab73f642650cd46241792677675a984a76b2c3fdebac40e6b8d0af9cd51d
+size 150693

last-checkpoint/global_step186/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:830251cfc2c437586c0ae49bf1b66313be2c453ea7011d96d958fc6e9e364ac1
+size 150693

last-checkpoint/global_step186/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9aa8517485b38137ae2838bcc85f98965bb72cb092a20364df05e9ef39b1b5f
+size 150693

last-checkpoint/latest

@@ -1 +1 @@
-global_step62
+global_step186

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d563f5be7dd17173a26681e62428a54af2c147222112afedffc5793b6cc61cf0
+oid sha256:81392ca1341eff0062e670bf6bae66fb987670c5ea66d248165e16d5b2550583
 size 4938985352

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dbee048396d9ec9c865dc06f44037e638bd4de3701c4796e1e6b86e23984e4bd
+oid sha256:1006a496ee196194e313e59321565ad7eb47f0dda4e314fd48d34709bd92ff82
 size 4947390880

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5770faeae9c8035062d2b7125cc1c23337c3fa6091bb86f7deea44209df1a86b
+oid sha256:01d6dfe355522d2cd855421ed2ca07b299127e352ea22630b44c2a31747bb37c
 size 3590488816

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a4a1dc8b9f546123ce9b44c28828ae71d9d0ecd6f59c6b7b8337c73cb2e24c49
+oid sha256:50ca27603f7c4a4d9a980223e1a102922ac31561acfa7da4f28042e4e8b3528b
 size 15024

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a8b225b0323772b66570a16b5c7e6446dc3c9ee076597906a2c90e3c0a383bf
+oid sha256:c57f658c3d9b6de611da50021823095fc4067036fb9a4b34655cc2e34fbf5c2b
 size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c036a8d528905f7018266618c8bebe4add10c2778ef432fd0116369fa622d815
+oid sha256:109da2c363a1e09432fa2ea2b7dceb6f67d1f91975069fc669e71dd536d654d8
 size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b93d88429d97654e76d10b1ba4c36f29a47e210c65a03efdc136f04af8282797
+oid sha256:bda65e490958bfe3fe01157ed131eb2354f8fc19536842d27c998ecf29d48991
 size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5645f0c0c551384f392d5d4b3ba0069b3e59c71dc60c4952144026a9bc6705f2
+oid sha256:a4f13df3470221c87dbd73f49aa8200fcb840f387f1d9666c0df7b34d2fa99d1
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.1984,
+  "epoch": 0.5952,
   "eval_steps": 62,
-  "global_step": 62,
+  "global_step": 186,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -128,6 +128,218 @@
       "eval_samples_per_second": 7.763,
       "eval_steps_per_second": 0.505,
       "step": 62
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 20.968302955882685,
+      "learning_rate": 3.7234042553191484e-07,
+      "logits/generated": -0.8760935068130493,
+      "logits/real": -0.745302677154541,
+      "logps/generated": -156.32565307617188,
+      "logps/real": -164.9800567626953,
+      "loss": 0.846,
+      "rewards/accuracies": 0.5625,
+      "rewards/generated": 0.44186854362487793,
+      "rewards/margins": 0.04428985342383385,
+      "rewards/real": 0.4861583709716797,
+      "step": 70
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 21.48558093332416,
+      "learning_rate": 4.25531914893617e-07,
+      "logits/generated": -0.8265154957771301,
+      "logits/real": -0.6477177143096924,
+      "logps/generated": -155.42221069335938,
+      "logps/real": -172.77003479003906,
+      "loss": 0.8384,
+      "rewards/accuracies": 0.574999988079071,
+      "rewards/generated": 0.6306847333908081,
+      "rewards/margins": 0.037362463772296906,
+      "rewards/real": 0.6680471897125244,
+      "step": 80
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 21.521672577795613,
+      "learning_rate": 4.787234042553192e-07,
+      "logits/generated": -0.8268479108810425,
+      "logits/real": -0.7267503142356873,
+      "logps/generated": -157.39114379882812,
+      "logps/real": -170.5515594482422,
+      "loss": 0.8479,
+      "rewards/accuracies": 0.5375000238418579,
+      "rewards/generated": 0.7636477947235107,
+      "rewards/margins": 0.07054053246974945,
+      "rewards/real": 0.8341882824897766,
+      "step": 90
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.88045536925292,
+      "learning_rate": 4.96437054631829e-07,
+      "logits/generated": -0.8575566411018372,
+      "logits/real": -0.7087486982345581,
+      "logps/generated": -139.00509643554688,
+      "logps/real": -169.8658447265625,
+      "loss": 0.8168,
+      "rewards/accuracies": 0.637499988079071,
+      "rewards/generated": 0.8750035166740417,
+      "rewards/margins": 0.1635354459285736,
+      "rewards/real": 1.038538932800293,
+      "step": 100
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 20.692573860591445,
+      "learning_rate": 4.904988123515439e-07,
+      "logits/generated": -0.7912083864212036,
+      "logits/real": -0.5575038194656372,
+      "logps/generated": -153.8982391357422,
+      "logps/real": -178.22079467773438,
+      "loss": 0.8178,
+      "rewards/accuracies": 0.612500011920929,
+      "rewards/generated": 1.1632182598114014,
+      "rewards/margins": 0.14400726556777954,
+      "rewards/real": 1.3072254657745361,
+      "step": 110
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 18.074770811259548,
+      "learning_rate": 4.845605700712589e-07,
+      "logits/generated": -0.7757072448730469,
+      "logits/real": -0.7249404191970825,
+      "logps/generated": -153.55979919433594,
+      "logps/real": -163.69528198242188,
+      "loss": 0.7999,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/generated": 1.2686254978179932,
+      "rewards/margins": 0.22223016619682312,
+      "rewards/real": 1.4908555746078491,
+      "step": 120
+    },
+    {
+      "epoch": 0.3968,
+      "eval_logits/generated": -0.8056570887565613,
+      "eval_logits/real": -0.6066926717758179,
+      "eval_logps/generated": -147.1573486328125,
+      "eval_logps/real": -163.69662475585938,
+      "eval_loss": 0.832275927066803,
+      "eval_rewards/accuracies": 0.5192307829856873,
+      "eval_rewards/generated": 1.5154075622558594,
+      "eval_rewards/margins": 0.07085353136062622,
+      "eval_rewards/real": 1.5862611532211304,
+      "eval_runtime": 32.2601,
+      "eval_samples_per_second": 6.2,
+      "eval_steps_per_second": 0.403,
+      "step": 124
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 21.381904884595862,
+      "learning_rate": 4.786223277909738e-07,
+      "logits/generated": -0.7485866546630859,
+      "logits/real": -0.6831777095794678,
+      "logps/generated": -146.22817993164062,
+      "logps/real": -164.9325408935547,
+      "loss": 0.8166,
+      "rewards/accuracies": 0.612500011920929,
+      "rewards/generated": 1.4485336542129517,
+      "rewards/margins": 0.16419439017772675,
+      "rewards/real": 1.6127278804779053,
+      "step": 130
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 18.194356255882095,
+      "learning_rate": 4.7268408551068883e-07,
+      "logits/generated": -0.8496313095092773,
+      "logits/real": -0.6610928773880005,
+      "logps/generated": -142.645263671875,
+      "logps/real": -171.78109741210938,
+      "loss": 0.8028,
+      "rewards/accuracies": 0.637499988079071,
+      "rewards/generated": 1.6386501789093018,
+      "rewards/margins": 0.14944753050804138,
+      "rewards/real": 1.788097620010376,
+      "step": 140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.92923604166353,
+      "learning_rate": 4.667458432304038e-07,
+      "logits/generated": -0.8171635866165161,
+      "logits/real": -0.6865423917770386,
+      "logps/generated": -164.94351196289062,
+      "logps/real": -165.67474365234375,
+      "loss": 0.8137,
+      "rewards/accuracies": 0.5874999761581421,
+      "rewards/generated": 1.741497278213501,
+      "rewards/margins": 0.08408152312040329,
+      "rewards/real": 1.8255789279937744,
+      "step": 150
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 17.43597928536244,
+      "learning_rate": 4.6080760095011875e-07,
+      "logits/generated": -0.873740553855896,
+      "logits/real": -0.6691497564315796,
+      "logps/generated": -137.250244140625,
+      "logps/real": -168.40560913085938,
+      "loss": 0.7849,
+      "rewards/accuracies": 0.699999988079071,
+      "rewards/generated": 1.7626874446868896,
+      "rewards/margins": 0.25575488805770874,
+      "rewards/real": 2.018442392349243,
+      "step": 160
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 19.21556142017865,
+      "learning_rate": 4.548693586698337e-07,
+      "logits/generated": -0.8208873867988586,
+      "logits/real": -0.7359132766723633,
+      "logps/generated": -137.61297607421875,
+      "logps/real": -162.14431762695312,
+      "loss": 0.8094,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/generated": 1.9185206890106201,
+      "rewards/margins": 0.1673974096775055,
+      "rewards/real": 2.0859181880950928,
+      "step": 170
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 28.419435881651143,
+      "learning_rate": 4.4893111638954866e-07,
+      "logits/generated": -0.8659377098083496,
+      "logits/real": -0.8016618490219116,
+      "logps/generated": -138.0242156982422,
+      "logps/real": -151.81307983398438,
+      "loss": 0.7846,
+      "rewards/accuracies": 0.6000000238418579,
+      "rewards/generated": 2.0733604431152344,
+      "rewards/margins": 0.16894717514514923,
+      "rewards/real": 2.242307662963867,
+      "step": 180
+    },
+    {
+      "epoch": 0.5952,
+      "eval_logits/generated": -0.8685788512229919,
+      "eval_logits/real": -0.6861833333969116,
+      "eval_logps/generated": -139.17666625976562,
+      "eval_logps/real": -155.08929443359375,
+      "eval_loss": 0.7979166507720947,
+      "eval_rewards/accuracies": 0.557692289352417,
+      "eval_rewards/generated": 2.3134748935699463,
+      "eval_rewards/margins": 0.1335190087556839,
+      "eval_rewards/real": 2.446993827819824,
+      "eval_runtime": 24.1759,
+      "eval_samples_per_second": 8.273,
+      "eval_steps_per_second": 0.538,
+      "step": 186
     }
   ],
   "logging_steps": 10,