| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8403361344537815, | |
| "eval_steps": 500, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 641.9296875, | |
| "epoch": 0.01680672268907563, | |
| "grad_norm": 0.1829434037208557, | |
| "kl": 0.0, | |
| "learning_rate": 5e-07, | |
| "loss": -0.0027, | |
| "num_tokens": 143935.0, | |
| "reward": 0.015625, | |
| "reward_std": 0.0289318785071373, | |
| "rewards/format_reward": 0.015625, | |
| "step": 1 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 558.2109375, | |
| "epoch": 0.03361344537815126, | |
| "grad_norm": 0.3770347237586975, | |
| "kl": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0243, | |
| "num_tokens": 275234.0, | |
| "reward": 0.03125, | |
| "reward_std": 0.0883883461356163, | |
| "rewards/format_reward": 0.03125, | |
| "step": 2 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 577.3203125, | |
| "epoch": 0.05042016806722689, | |
| "grad_norm": 0.2114293873310089, | |
| "kl": 0.00027871131896972656, | |
| "learning_rate": 9.990365154573716e-07, | |
| "loss": 0.0109, | |
| "num_tokens": 411211.0, | |
| "reward": 0.015625, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.015625, | |
| "step": 3 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 551.625, | |
| "epoch": 0.06722689075630252, | |
| "grad_norm": 0.32700982689857483, | |
| "kl": 0.00025844573974609375, | |
| "learning_rate": 9.961501876182148e-07, | |
| "loss": 0.0038, | |
| "num_tokens": 541427.0, | |
| "reward": 0.0234375, | |
| "reward_std": 0.06629125960171223, | |
| "rewards/format_reward": 0.0234375, | |
| "step": 4 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 577.4296875, | |
| "epoch": 0.08403361344537816, | |
| "grad_norm": 0.2744513154029846, | |
| "kl": 0.0002903938293457031, | |
| "learning_rate": 9.913533761814537e-07, | |
| "loss": 0.0185, | |
| "num_tokens": 674322.0, | |
| "reward": 0.0390625, | |
| "reward_std": 0.0765409953892231, | |
| "rewards/format_reward": 0.0390625, | |
| "step": 5 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 582.046875, | |
| "epoch": 0.10084033613445378, | |
| "grad_norm": 0.2994139492511749, | |
| "kl": 0.0002753734588623047, | |
| "learning_rate": 9.846666218300807e-07, | |
| "loss": 0.011, | |
| "num_tokens": 808448.0, | |
| "reward": 0.03125, | |
| "reward_std": 0.07312605157494545, | |
| "rewards/format_reward": 0.03125, | |
| "step": 6 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 567.9296875, | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 0.2810344696044922, | |
| "kl": 0.0003495216369628906, | |
| "learning_rate": 9.761185582727976e-07, | |
| "loss": 0.0199, | |
| "num_tokens": 936975.0, | |
| "reward": 0.0234375, | |
| "reward_std": 0.051028965041041374, | |
| "rewards/format_reward": 0.0234375, | |
| "step": 7 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 565.2578125, | |
| "epoch": 0.13445378151260504, | |
| "grad_norm": 0.3799228370189667, | |
| "kl": 0.00041484832763671875, | |
| "learning_rate": 9.657457896300791e-07, | |
| "loss": 0.022, | |
| "num_tokens": 1071784.0, | |
| "reward": 0.046875, | |
| "reward_std": 0.1173202246427536, | |
| "rewards/format_reward": 0.046875, | |
| "step": 8 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 534.75, | |
| "epoch": 0.15126050420168066, | |
| "grad_norm": 0.42587605118751526, | |
| "kl": 0.0007534027099609375, | |
| "learning_rate": 9.535927336897097e-07, | |
| "loss": 0.0392, | |
| "num_tokens": 1194536.0, | |
| "reward": 0.0625, | |
| "reward_std": 0.14283225685358047, | |
| "rewards/format_reward": 0.0625, | |
| "step": 9 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 621.3828125, | |
| "epoch": 0.16806722689075632, | |
| "grad_norm": 0.41017991304397583, | |
| "kl": 0.0006995201110839844, | |
| "learning_rate": 9.397114317029974e-07, | |
| "loss": -0.0271, | |
| "num_tokens": 1336961.0, | |
| "reward": 0.09375, | |
| "reward_std": 0.13781969621777534, | |
| "rewards/format_reward": 0.09375, | |
| "step": 10 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 537.6796875, | |
| "epoch": 0.18487394957983194, | |
| "grad_norm": 0.6829501986503601, | |
| "kl": 0.0015192031860351562, | |
| "learning_rate": 9.241613255361454e-07, | |
| "loss": 0.0224, | |
| "num_tokens": 1462080.0, | |
| "reward": 0.1953125, | |
| "reward_std": 0.2961437702178955, | |
| "rewards/format_reward": 0.1953125, | |
| "step": 11 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 590.59375, | |
| "epoch": 0.20168067226890757, | |
| "grad_norm": 0.484163761138916, | |
| "kl": 0.0014421939849853516, | |
| "learning_rate": 9.070090031310558e-07, | |
| "loss": 0.0102, | |
| "num_tokens": 1600932.0, | |
| "reward": 0.109375, | |
| "reward_std": 0.12703318148851395, | |
| "rewards/format_reward": 0.109375, | |
| "step": 12 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 565.2109375, | |
| "epoch": 0.2184873949579832, | |
| "grad_norm": 0.5853496789932251, | |
| "kl": 0.002685546875, | |
| "learning_rate": 8.883279133655398e-07, | |
| "loss": 0.0463, | |
| "num_tokens": 1733479.0, | |
| "reward": 0.203125, | |
| "reward_std": 0.2335668858140707, | |
| "rewards/format_reward": 0.203125, | |
| "step": 13 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 489.96875, | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 0.7053346037864685, | |
| "kl": 0.00472259521484375, | |
| "learning_rate": 8.681980515339463e-07, | |
| "loss": -0.0509, | |
| "num_tokens": 1851611.0, | |
| "reward": 0.34375, | |
| "reward_std": 0.2740466818213463, | |
| "rewards/format_reward": 0.34375, | |
| "step": 14 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 555.671875, | |
| "epoch": 0.25210084033613445, | |
| "grad_norm": 0.6062747240066528, | |
| "kl": 0.0043487548828125, | |
| "learning_rate": 8.46705616795031e-07, | |
| "loss": 0.0142, | |
| "num_tokens": 1982657.0, | |
| "reward": 0.4140625, | |
| "reward_std": 0.30509532056748867, | |
| "rewards/format_reward": 0.4140625, | |
| "step": 15 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 567.4453125, | |
| "epoch": 0.2689075630252101, | |
| "grad_norm": 0.5138593912124634, | |
| "kl": 0.004817962646484375, | |
| "learning_rate": 8.239426430539243e-07, | |
| "loss": -0.0185, | |
| "num_tokens": 2112490.0, | |
| "reward": 0.328125, | |
| "reward_std": 0.2001592293381691, | |
| "rewards/format_reward": 0.328125, | |
| "step": 16 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 497.5859375, | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.5945853590965271, | |
| "kl": 0.00463104248046875, | |
| "learning_rate": 8.00006604858821e-07, | |
| "loss": -0.0218, | |
| "num_tokens": 2233365.0, | |
| "reward": 0.40625, | |
| "reward_std": 0.25224902853369713, | |
| "rewards/format_reward": 0.40625, | |
| "step": 17 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 670.125, | |
| "epoch": 0.3025210084033613, | |
| "grad_norm": 0.27369990944862366, | |
| "kl": 0.000995635986328125, | |
| "learning_rate": 7.75e-07, | |
| "loss": 0.0205, | |
| "num_tokens": 2383277.0, | |
| "reward": 0.09375, | |
| "reward_std": 0.05444391071796417, | |
| "rewards/format_reward": 0.09375, | |
| "step": 18 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 602.7265625, | |
| "epoch": 0.31932773109243695, | |
| "grad_norm": 0.5327508449554443, | |
| "kl": 0.006078243255615234, | |
| "learning_rate": 7.490299105985506e-07, | |
| "loss": -0.0041, | |
| "num_tokens": 2521058.0, | |
| "reward": 0.296875, | |
| "reward_std": 0.15650184080004692, | |
| "rewards/format_reward": 0.296875, | |
| "step": 19 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 473.875, | |
| "epoch": 0.33613445378151263, | |
| "grad_norm": 0.5724947452545166, | |
| "kl": 0.008075714111328125, | |
| "learning_rate": 7.222075445642904e-07, | |
| "loss": 0.0444, | |
| "num_tokens": 2641082.0, | |
| "reward": 0.4453125, | |
| "reward_std": 0.2212003916501999, | |
| "rewards/format_reward": 0.4453125, | |
| "step": 20 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 660.5703125, | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 0.3499065935611725, | |
| "kl": 0.0032248497009277344, | |
| "learning_rate": 6.946477593864227e-07, | |
| "loss": -0.0102, | |
| "num_tokens": 2788211.0, | |
| "reward": 0.203125, | |
| "reward_std": 0.08337578736245632, | |
| "rewards/format_reward": 0.203125, | |
| "step": 21 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 566.859375, | |
| "epoch": 0.3697478991596639, | |
| "grad_norm": 0.2869742810726166, | |
| "kl": 0.00702667236328125, | |
| "learning_rate": 6.664685702961344e-07, | |
| "loss": 0.0002, | |
| "num_tokens": 2917449.0, | |
| "reward": 0.4296875, | |
| "reward_std": 0.022097086533904076, | |
| "rewards/format_reward": 0.4296875, | |
| "step": 22 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 603.03125, | |
| "epoch": 0.3865546218487395, | |
| "grad_norm": 0.22740638256072998, | |
| "kl": 0.005024433135986328, | |
| "learning_rate": 6.377906449072577e-07, | |
| "loss": -0.0122, | |
| "num_tokens": 3056589.0, | |
| "reward": 0.28125, | |
| "reward_std": 0.05444390885531902, | |
| "rewards/format_reward": 0.28125, | |
| "step": 23 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 525.0234375, | |
| "epoch": 0.40336134453781514, | |
| "grad_norm": 0.486358642578125, | |
| "kl": 0.005306243896484375, | |
| "learning_rate": 6.087367864990232e-07, | |
| "loss": 0.0533, | |
| "num_tokens": 3187176.0, | |
| "reward": 0.3828125, | |
| "reward_std": 0.1207351740449667, | |
| "rewards/format_reward": 0.3828125, | |
| "step": 24 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 540.1953125, | |
| "epoch": 0.42016806722689076, | |
| "grad_norm": 0.3638220429420471, | |
| "kl": 0.007305145263671875, | |
| "learning_rate": 5.794314081535644e-07, | |
| "loss": -0.0038, | |
| "num_tokens": 3314737.0, | |
| "reward": 0.4140625, | |
| "reward_std": 0.09522314183413982, | |
| "rewards/format_reward": 0.4140625, | |
| "step": 25 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 497.3125, | |
| "epoch": 0.4369747899159664, | |
| "grad_norm": 0.5379955768585205, | |
| "kl": 0.014757633209228516, | |
| "learning_rate": 5.5e-07, | |
| "loss": -0.0513, | |
| "num_tokens": 3436265.0, | |
| "reward": 0.6015625, | |
| "reward_std": 0.19728107377886772, | |
| "rewards/format_reward": 0.6015625, | |
| "step": 26 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 627.1875, | |
| "epoch": 0.453781512605042, | |
| "grad_norm": 0.26467689871788025, | |
| "kl": 0.005634307861328125, | |
| "learning_rate": 5.205685918464355e-07, | |
| "loss": -0.0003, | |
| "num_tokens": 3577265.0, | |
| "reward": 0.3515625, | |
| "reward_std": 0.05102896690368652, | |
| "rewards/format_reward": 0.3515625, | |
| "step": 27 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 543.90625, | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.3865010142326355, | |
| "kl": 0.0108489990234375, | |
| "learning_rate": 4.912632135009769e-07, | |
| "loss": 0.0253, | |
| "num_tokens": 3703797.0, | |
| "reward": 0.5390625, | |
| "reward_std": 0.05102896690368652, | |
| "rewards/format_reward": 0.5390625, | |
| "step": 28 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 668.0078125, | |
| "epoch": 0.48739495798319327, | |
| "grad_norm": 0.29972589015960693, | |
| "kl": 0.0080108642578125, | |
| "learning_rate": 4.6220935509274227e-07, | |
| "loss": 0.0094, | |
| "num_tokens": 3850094.0, | |
| "reward": 0.3515625, | |
| "reward_std": 0.051028965041041374, | |
| "rewards/format_reward": 0.3515625, | |
| "step": 29 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 607.1875, | |
| "epoch": 0.5042016806722689, | |
| "grad_norm": 0.32297083735466003, | |
| "kl": 0.007590293884277344, | |
| "learning_rate": 4.3353142970386557e-07, | |
| "loss": 0.0134, | |
| "num_tokens": 3991174.0, | |
| "reward": 0.4140625, | |
| "reward_std": 0.06629125960171223, | |
| "rewards/format_reward": 0.4140625, | |
| "step": 30 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 599.109375, | |
| "epoch": 0.5210084033613446, | |
| "grad_norm": 0.004667076747864485, | |
| "kl": 0.005215167999267578, | |
| "learning_rate": 4.053522406135774e-07, | |
| "loss": 0.0001, | |
| "num_tokens": 4129020.0, | |
| "reward": 0.1875, | |
| "reward_std": 0.0, | |
| "rewards/format_reward": 0.1875, | |
| "step": 31 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 553.5546875, | |
| "epoch": 0.5378151260504201, | |
| "grad_norm": 0.23914550244808197, | |
| "kl": 0.006618499755859375, | |
| "learning_rate": 3.777924554357096e-07, | |
| "loss": 0.0173, | |
| "num_tokens": 4263587.0, | |
| "reward": 0.4296875, | |
| "reward_std": 0.022097086533904076, | |
| "rewards/format_reward": 0.4296875, | |
| "step": 32 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 614.109375, | |
| "epoch": 0.5546218487394958, | |
| "grad_norm": 0.38585954904556274, | |
| "kl": 0.00874185562133789, | |
| "learning_rate": 3.5097008940144955e-07, | |
| "loss": 0.0186, | |
| "num_tokens": 4404113.0, | |
| "reward": 0.359375, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.359375, | |
| "step": 33 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 503.859375, | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.2774820923805237, | |
| "kl": 0.0081939697265625, | |
| "learning_rate": 3.250000000000001e-07, | |
| "loss": 0.009, | |
| "num_tokens": 4530399.0, | |
| "reward": 0.421875, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.421875, | |
| "step": 34 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 510.8046875, | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.3793410360813141, | |
| "kl": 0.007872581481933594, | |
| "learning_rate": 2.999933951411791e-07, | |
| "loss": 0.0096, | |
| "num_tokens": 4656958.0, | |
| "reward": 0.359375, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.359375, | |
| "step": 35 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 554.125, | |
| "epoch": 0.6050420168067226, | |
| "grad_norm": 0.23132917284965515, | |
| "kl": 0.0100250244140625, | |
| "learning_rate": 2.760573569460757e-07, | |
| "loss": -0.0074, | |
| "num_tokens": 4788182.0, | |
| "reward": 0.484375, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.484375, | |
| "step": 36 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 503.140625, | |
| "epoch": 0.6218487394957983, | |
| "grad_norm": 0.41905921697616577, | |
| "kl": 0.0104522705078125, | |
| "learning_rate": 2.53294383204969e-07, | |
| "loss": 0.004, | |
| "num_tokens": 4910232.0, | |
| "reward": 0.40625, | |
| "reward_std": 0.0883883461356163, | |
| "rewards/format_reward": 0.40625, | |
| "step": 37 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 488.4453125, | |
| "epoch": 0.6386554621848739, | |
| "grad_norm": 0.009937948547303677, | |
| "kl": 0.0146942138671875, | |
| "learning_rate": 2.3180194846605364e-07, | |
| "loss": 0.0001, | |
| "num_tokens": 5028785.0, | |
| "reward": 0.5625, | |
| "reward_std": 0.0, | |
| "rewards/format_reward": 0.5625, | |
| "step": 38 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 535.7578125, | |
| "epoch": 0.6554621848739496, | |
| "grad_norm": 0.40260764956474304, | |
| "kl": 0.00785064697265625, | |
| "learning_rate": 2.1167208663446022e-07, | |
| "loss": -0.0086, | |
| "num_tokens": 5156906.0, | |
| "reward": 0.3515625, | |
| "reward_std": 0.06629125960171223, | |
| "rewards/format_reward": 0.3515625, | |
| "step": 39 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 584.6796875, | |
| "epoch": 0.6722689075630253, | |
| "grad_norm": 0.3577350676059723, | |
| "kl": 0.008663177490234375, | |
| "learning_rate": 1.929909968689442e-07, | |
| "loss": 0.0166, | |
| "num_tokens": 5292793.0, | |
| "reward": 0.3125, | |
| "reward_std": 0.0731260534375906, | |
| "rewards/format_reward": 0.3125, | |
| "step": 40 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 494.53125, | |
| "epoch": 0.6890756302521008, | |
| "grad_norm": 0.5300049781799316, | |
| "kl": 0.016998291015625, | |
| "learning_rate": 1.758386744638546e-07, | |
| "loss": -0.0005, | |
| "num_tokens": 5413677.0, | |
| "reward": 0.5859375, | |
| "reward_std": 0.11048543266952038, | |
| "rewards/format_reward": 0.5859375, | |
| "step": 41 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 454.0859375, | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.6002023220062256, | |
| "kl": 0.012908935546875, | |
| "learning_rate": 1.6028856829700258e-07, | |
| "loss": 0.0249, | |
| "num_tokens": 5530672.0, | |
| "reward": 0.578125, | |
| "reward_std": 0.1173202209174633, | |
| "rewards/format_reward": 0.578125, | |
| "step": 42 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 536.7890625, | |
| "epoch": 0.7226890756302521, | |
| "grad_norm": 0.40132713317871094, | |
| "kl": 0.01432037353515625, | |
| "learning_rate": 1.464072663102903e-07, | |
| "loss": 0.0289, | |
| "num_tokens": 5657981.0, | |
| "reward": 0.4765625, | |
| "reward_std": 0.06629125960171223, | |
| "rewards/format_reward": 0.4765625, | |
| "step": 43 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 477.53125, | |
| "epoch": 0.7394957983193278, | |
| "grad_norm": 0.5269466042518616, | |
| "kl": 0.011138916015625, | |
| "learning_rate": 1.3425421036992097e-07, | |
| "loss": 0.0338, | |
| "num_tokens": 5782097.0, | |
| "reward": 0.5859375, | |
| "reward_std": 0.09522313810884953, | |
| "rewards/format_reward": 0.5859375, | |
| "step": 44 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 583.5546875, | |
| "epoch": 0.7563025210084033, | |
| "grad_norm": 0.30939722061157227, | |
| "kl": 0.00948333740234375, | |
| "learning_rate": 1.238814417272025e-07, | |
| "loss": 0.0011, | |
| "num_tokens": 5915848.0, | |
| "reward": 0.484375, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.484375, | |
| "step": 45 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 493.0, | |
| "epoch": 0.773109243697479, | |
| "grad_norm": 0.36379459500312805, | |
| "kl": 0.0174407958984375, | |
| "learning_rate": 1.1533337816991931e-07, | |
| "loss": 0.0097, | |
| "num_tokens": 6035144.0, | |
| "reward": 0.609375, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.609375, | |
| "step": 46 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 586.8125, | |
| "epoch": 0.7899159663865546, | |
| "grad_norm": 0.3710026741027832, | |
| "kl": 0.00640106201171875, | |
| "learning_rate": 1.086466238185463e-07, | |
| "loss": 0.0173, | |
| "num_tokens": 6172192.0, | |
| "reward": 0.359375, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.359375, | |
| "step": 47 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 575.15625, | |
| "epoch": 0.8067226890756303, | |
| "grad_norm": 0.010671989060938358, | |
| "kl": 0.00979757308959961, | |
| "learning_rate": 1.0384981238178533e-07, | |
| "loss": 0.0001, | |
| "num_tokens": 6307060.0, | |
| "reward": 0.4375, | |
| "reward_std": 0.0, | |
| "rewards/format_reward": 0.4375, | |
| "step": 48 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 570.890625, | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 0.40951627492904663, | |
| "kl": 0.011271476745605469, | |
| "learning_rate": 1.0096348454262843e-07, | |
| "loss": 0.0422, | |
| "num_tokens": 6441918.0, | |
| "reward": 0.421875, | |
| "reward_std": 0.04419417306780815, | |
| "rewards/format_reward": 0.421875, | |
| "step": 49 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 576.8515625, | |
| "epoch": 0.8403361344537815, | |
| "grad_norm": 0.013534413650631905, | |
| "kl": 0.0068531036376953125, | |
| "learning_rate": 1e-07, | |
| "loss": 0.0001, | |
| "num_tokens": 6575251.0, | |
| "reward": 0.3125, | |
| "reward_std": 0.0, | |
| "rewards/format_reward": 0.3125, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8403361344537815, | |
| "step": 50, | |
| "total_flos": 0.0, | |
| "train_loss": 0.008462187110126251, | |
| "train_runtime": 7668.932, | |
| "train_samples_per_second": 0.835, | |
| "train_steps_per_second": 0.007 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |