Qwen3-4b-BCF-400 / trainer_state.json
ozayezerceli's picture
Upload folder using huggingface_hub
dfdef1d verified
Raw
History Blame Contribute Delete
321 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.07108583614714768,
"eval_steps": 500,
"global_step": 400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"completion_length": 427.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 460.0,
"completions/max_terminated_length": 460.0,
"completions/mean_length": 427.0,
"completions/mean_terminated_length": 427.0,
"completions/min_length": 386.0,
"completions/min_terminated_length": 386.0,
"epoch": 0.00017771459036786921,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6094411611557007,
"kl": 1.0498238056898117,
"learning_rate": 0.0,
"loss": 0.001,
"num_tokens": 2836.0,
"reward": 0.6189357042312622,
"reward_std": 0.16637657582759857,
"rewards/check_answer/mean": 0.6189357042312622,
"rewards/check_answer/std": 0.16637657582759857,
"step": 1
},
{
"completion_length": 541.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 559.0,
"completions/max_terminated_length": 559.0,
"completions/mean_length": 541.75,
"completions/mean_terminated_length": 541.75,
"completions/min_length": 526.0,
"completions/min_terminated_length": 526.0,
"epoch": 0.00035542918073573843,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3810062408447266,
"kl": 1.1163947582244873,
"learning_rate": 7.142857142857144e-08,
"loss": 0.0011,
"num_tokens": 6171.0,
"reward": 0.6918560862541199,
"reward_std": 0.06921806931495667,
"rewards/check_answer/mean": 0.6918560862541199,
"rewards/check_answer/std": 0.06921806931495667,
"step": 2
},
{
"completion_length": 599.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 836.0,
"completions/max_terminated_length": 836.0,
"completions/mean_length": 599.0,
"completions/mean_terminated_length": 599.0,
"completions/min_length": 419.0,
"completions/min_terminated_length": 419.0,
"epoch": 0.0005331437711036076,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.256874442100525,
"kl": 0.801612138748169,
"learning_rate": 1.4285714285714287e-07,
"loss": 0.0008,
"num_tokens": 9807.0,
"reward": 0.6589475274085999,
"reward_std": 0.10396059602499008,
"rewards/check_answer/mean": 0.6589475274085999,
"rewards/check_answer/std": 0.10396059602499008,
"step": 3
},
{
"completion_length": 466.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 491.0,
"completions/max_terminated_length": 491.0,
"completions/mean_length": 466.75,
"completions/mean_terminated_length": 466.75,
"completions/min_length": 435.0,
"completions/min_terminated_length": 435.0,
"epoch": 0.0007108583614714769,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8016468286514282,
"kl": 1.3025178015232086,
"learning_rate": 2.142857142857143e-07,
"loss": 0.0013,
"num_tokens": 12814.0,
"reward": 0.4898241460323334,
"reward_std": 0.0683608204126358,
"rewards/check_answer/mean": 0.4898241460323334,
"rewards/check_answer/std": 0.06836080551147461,
"step": 4
},
{
"completion_length": 596.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 665.0,
"completions/max_terminated_length": 665.0,
"completions/mean_length": 596.25,
"completions/mean_terminated_length": 596.25,
"completions/min_length": 504.0,
"completions/min_terminated_length": 504.0,
"epoch": 0.000888572951839346,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4106318950653076,
"kl": 1.153193473815918,
"learning_rate": 2.8571428571428575e-07,
"loss": 0.0012,
"num_tokens": 16407.0,
"reward": 0.589160144329071,
"reward_std": 0.1366165727376938,
"rewards/check_answer/mean": 0.589160144329071,
"rewards/check_answer/std": 0.1366165727376938,
"step": 5
},
{
"completion_length": 521.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 566.0,
"completions/max_terminated_length": 566.0,
"completions/mean_length": 521.75,
"completions/mean_terminated_length": 521.75,
"completions/min_length": 491.0,
"completions/min_terminated_length": 491.0,
"epoch": 0.0010662875422072153,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6702299118041992,
"kl": 1.1629396080970764,
"learning_rate": 3.5714285714285716e-07,
"loss": 0.0012,
"num_tokens": 19658.0,
"reward": 0.6594047546386719,
"reward_std": 0.03622956573963165,
"rewards/check_answer/mean": 0.6594047546386719,
"rewards/check_answer/std": 0.03622956946492195,
"step": 6
},
{
"completion_length": 557.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 574.0,
"completions/max_terminated_length": 574.0,
"completions/mean_length": 557.75,
"completions/mean_terminated_length": 557.75,
"completions/min_length": 537.0,
"completions/min_terminated_length": 537.0,
"epoch": 0.0012440021325750845,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4264965057373047,
"kl": 1.3246853649616241,
"learning_rate": 4.285714285714286e-07,
"loss": 0.0013,
"num_tokens": 23093.0,
"reward": 0.5148710012435913,
"reward_std": 0.030068516731262207,
"rewards/check_answer/mean": 0.5148710012435913,
"rewards/check_answer/std": 0.030068522319197655,
"step": 7
},
{
"completion_length": 411.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 474.0,
"completions/max_terminated_length": 474.0,
"completions/mean_length": 411.5,
"completions/mean_terminated_length": 411.5,
"completions/min_length": 368.0,
"completions/min_terminated_length": 368.0,
"epoch": 0.0014217167229429537,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6491725444793701,
"kl": 1.1441383361816406,
"learning_rate": 5.000000000000001e-07,
"loss": 0.0011,
"num_tokens": 25891.0,
"reward": 0.39803165197372437,
"reward_std": 0.1641732007265091,
"rewards/check_answer/mean": 0.39803165197372437,
"rewards/check_answer/std": 0.1641732156276703,
"step": 8
},
{
"completion_length": 432.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 499.0,
"completions/max_terminated_length": 499.0,
"completions/mean_length": 432.0,
"completions/mean_terminated_length": 432.0,
"completions/min_length": 385.0,
"completions/min_terminated_length": 385.0,
"epoch": 0.0015994313133108227,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.763498067855835,
"kl": 1.447030395269394,
"learning_rate": 5.714285714285715e-07,
"loss": 0.0014,
"num_tokens": 28743.0,
"reward": 0.6316161155700684,
"reward_std": 0.08715175837278366,
"rewards/check_answer/mean": 0.6316161155700684,
"rewards/check_answer/std": 0.08715175092220306,
"step": 9
},
{
"completion_length": 467.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 486.0,
"completions/max_terminated_length": 486.0,
"completions/mean_length": 467.0,
"completions/mean_terminated_length": 467.0,
"completions/min_length": 442.0,
"completions/min_terminated_length": 442.0,
"epoch": 0.001777145903678692,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7400954961776733,
"kl": 2.312670022249222,
"learning_rate": 6.428571428571428e-07,
"loss": 0.0023,
"num_tokens": 31759.0,
"reward": 0.7970674633979797,
"reward_std": 0.18348881602287292,
"rewards/check_answer/mean": 0.7970674633979797,
"rewards/check_answer/std": 0.18348881602287292,
"step": 10
},
{
"completion_length": 405.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 495.0,
"completions/max_terminated_length": 495.0,
"completions/mean_length": 405.75,
"completions/mean_terminated_length": 405.75,
"completions/min_length": 339.0,
"completions/min_terminated_length": 339.0,
"epoch": 0.001954860494046561,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9999518394470215,
"kl": 1.1593981087207794,
"learning_rate": 7.142857142857143e-07,
"loss": 0.0012,
"num_tokens": 34514.0,
"reward": 0.3714173436164856,
"reward_std": 0.06423555314540863,
"rewards/check_answer/mean": 0.3714173436164856,
"rewards/check_answer/std": 0.06423555314540863,
"step": 11
},
{
"completion_length": 434.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 493.0,
"completions/max_terminated_length": 493.0,
"completions/mean_length": 434.0,
"completions/mean_terminated_length": 434.0,
"completions/min_length": 354.0,
"completions/min_terminated_length": 354.0,
"epoch": 0.0021325750844144306,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7830874919891357,
"kl": 2.1814075708389282,
"learning_rate": 7.857142857142857e-07,
"loss": 0.0022,
"num_tokens": 37398.0,
"reward": 0.6386286616325378,
"reward_std": 0.05019339174032211,
"rewards/check_answer/mean": 0.6386286616325378,
"rewards/check_answer/std": 0.05019339174032211,
"step": 12
},
{
"completion_length": 537.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 566.0,
"completions/max_terminated_length": 566.0,
"completions/mean_length": 537.5,
"completions/mean_terminated_length": 537.5,
"completions/min_length": 510.0,
"completions/min_terminated_length": 510.0,
"epoch": 0.0023102896747822996,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.486297369003296,
"kl": 1.0660419911146164,
"learning_rate": 8.571428571428572e-07,
"loss": 0.0011,
"num_tokens": 40724.0,
"reward": 0.35336199402809143,
"reward_std": 0.037462785840034485,
"rewards/check_answer/mean": 0.35336199402809143,
"rewards/check_answer/std": 0.03746278956532478,
"step": 13
},
{
"completion_length": 376.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 452.0,
"completions/max_terminated_length": 452.0,
"completions/mean_length": 376.0,
"completions/mean_terminated_length": 376.0,
"completions/min_length": 282.0,
"completions/min_terminated_length": 282.0,
"epoch": 0.002488004265150169,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.73249351978302,
"kl": 1.2243637442588806,
"learning_rate": 9.285714285714287e-07,
"loss": 0.0012,
"num_tokens": 43336.0,
"reward": 0.4773079752922058,
"reward_std": 0.18512390553951263,
"rewards/check_answer/mean": 0.4773079752922058,
"rewards/check_answer/std": 0.18512390553951263,
"step": 14
},
{
"completion_length": 410.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 440.0,
"completions/max_terminated_length": 440.0,
"completions/mean_length": 410.25,
"completions/mean_terminated_length": 410.25,
"completions/min_length": 348.0,
"completions/min_terminated_length": 348.0,
"epoch": 0.002665718855518038,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.818801999092102,
"kl": 1.239543467760086,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.0012,
"num_tokens": 46113.0,
"reward": 0.8123418092727661,
"reward_std": 0.2969806492328644,
"rewards/check_answer/mean": 0.8123418092727661,
"rewards/check_answer/std": 0.2969806492328644,
"step": 15
},
{
"completion_length": 585.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 634.0,
"completions/max_terminated_length": 634.0,
"completions/mean_length": 585.25,
"completions/mean_terminated_length": 585.25,
"completions/min_length": 523.0,
"completions/min_terminated_length": 523.0,
"epoch": 0.0028434334458859074,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3664265871047974,
"kl": 1.164329156279564,
"learning_rate": 1.0714285714285714e-06,
"loss": 0.0012,
"num_tokens": 49654.0,
"reward": 0.7955901622772217,
"reward_std": 0.16037997603416443,
"rewards/check_answer/mean": 0.7955901622772217,
"rewards/check_answer/std": 0.16037999093532562,
"step": 16
},
{
"completion_length": 409.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 498.0,
"completions/max_terminated_length": 498.0,
"completions/mean_length": 409.0,
"completions/mean_terminated_length": 409.0,
"completions/min_length": 307.0,
"completions/min_terminated_length": 307.0,
"epoch": 0.0030211480362537764,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7888686656951904,
"kl": 1.2454279661178589,
"learning_rate": 1.142857142857143e-06,
"loss": 0.0012,
"num_tokens": 52454.0,
"reward": 0.3565758168697357,
"reward_std": 0.19015806913375854,
"rewards/check_answer/mean": 0.3565758168697357,
"rewards/check_answer/std": 0.19015805423259735,
"step": 17
},
{
"completion_length": 492.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 615.0,
"completions/max_terminated_length": 615.0,
"completions/mean_length": 492.0,
"completions/mean_terminated_length": 492.0,
"completions/min_length": 426.0,
"completions/min_terminated_length": 426.0,
"epoch": 0.0031988626266216454,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6943333148956299,
"kl": 1.0921899378299713,
"learning_rate": 1.2142857142857144e-06,
"loss": 0.0011,
"num_tokens": 55558.0,
"reward": 1.0070428848266602,
"reward_std": 0.08447255939245224,
"rewards/check_answer/mean": 1.0070428848266602,
"rewards/check_answer/std": 0.08447252959012985,
"step": 18
},
{
"completion_length": 574.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 628.0,
"completions/max_terminated_length": 628.0,
"completions/mean_length": 574.0,
"completions/mean_terminated_length": 574.0,
"completions/min_length": 524.0,
"completions/min_terminated_length": 524.0,
"epoch": 0.003376577216989515,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5357227325439453,
"kl": 1.0007081925868988,
"learning_rate": 1.2857142857142856e-06,
"loss": 0.001,
"num_tokens": 59062.0,
"reward": 0.5813010931015015,
"reward_std": 0.17026397585868835,
"rewards/check_answer/mean": 0.5813010931015015,
"rewards/check_answer/std": 0.17026399075984955,
"step": 19
},
{
"completion_length": 422.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 441.0,
"completions/max_terminated_length": 441.0,
"completions/mean_length": 422.75,
"completions/mean_terminated_length": 422.75,
"completions/min_length": 408.0,
"completions/min_terminated_length": 408.0,
"epoch": 0.003554291807357384,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.447060227394104,
"kl": 1.0916816592216492,
"learning_rate": 1.3571428571428572e-06,
"loss": 0.0011,
"num_tokens": 61925.0,
"reward": 0.639136552810669,
"reward_std": 0.14528952538967133,
"rewards/check_answer/mean": 0.639136552810669,
"rewards/check_answer/std": 0.14528952538967133,
"step": 20
},
{
"completion_length": 491.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 507.0,
"completions/max_terminated_length": 507.0,
"completions/mean_length": 491.5,
"completions/mean_terminated_length": 491.5,
"completions/min_length": 463.0,
"completions/min_terminated_length": 463.0,
"epoch": 0.0037320063977252533,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4165836572647095,
"kl": 0.9325059801340103,
"learning_rate": 1.4285714285714286e-06,
"loss": 0.0009,
"num_tokens": 65127.0,
"reward": 0.8968478441238403,
"reward_std": 0.07308956235647202,
"rewards/check_answer/mean": 0.8968478441238403,
"rewards/check_answer/std": 0.07308956235647202,
"step": 21
},
{
"completion_length": 308.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 337.0,
"completions/max_terminated_length": 337.0,
"completions/mean_length": 308.75,
"completions/mean_terminated_length": 308.75,
"completions/min_length": 251.0,
"completions/min_terminated_length": 251.0,
"epoch": 0.003909720988093122,
"frac_reward_zero_std": 1.0,
"grad_norm": 0.006172445602715015,
"kl": 1.3253843784332275,
"learning_rate": 1.5e-06,
"loss": 0.0013,
"num_tokens": 67458.0,
"reward": 1.5,
"reward_std": 0.0,
"rewards/check_answer/mean": 1.5,
"rewards/check_answer/std": 0.0,
"step": 22
},
{
"completion_length": 457.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 521.0,
"completions/max_terminated_length": 521.0,
"completions/mean_length": 457.75,
"completions/mean_terminated_length": 457.75,
"completions/min_length": 405.0,
"completions/min_terminated_length": 405.0,
"epoch": 0.004087435578460991,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.757681965827942,
"kl": 1.0600132942199707,
"learning_rate": 1.5714285714285714e-06,
"loss": 0.0011,
"num_tokens": 70461.0,
"reward": 0.4985591769218445,
"reward_std": 0.18019117414951324,
"rewards/check_answer/mean": 0.4985591769218445,
"rewards/check_answer/std": 0.18019118905067444,
"step": 23
},
{
"completion_length": 409.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 423.0,
"completions/max_terminated_length": 423.0,
"completions/mean_length": 409.25,
"completions/mean_terminated_length": 409.25,
"completions/min_length": 383.0,
"completions/min_terminated_length": 383.0,
"epoch": 0.004265150168828861,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7924888134002686,
"kl": 1.278127372264862,
"learning_rate": 1.642857142857143e-06,
"loss": 0.0013,
"num_tokens": 73210.0,
"reward": 0.7848971486091614,
"reward_std": 0.2511295974254608,
"rewards/check_answer/mean": 0.7848971486091614,
"rewards/check_answer/std": 0.2511295974254608,
"step": 24
},
{
"completion_length": 418.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 503.0,
"completions/max_terminated_length": 503.0,
"completions/mean_length": 418.0,
"completions/mean_terminated_length": 418.0,
"completions/min_length": 333.0,
"completions/min_terminated_length": 333.0,
"epoch": 0.00444286475919673,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.607702612876892,
"kl": 1.2083918154239655,
"learning_rate": 1.7142857142857145e-06,
"loss": 0.0012,
"num_tokens": 75998.0,
"reward": 0.9702531099319458,
"reward_std": 0.14307360351085663,
"rewards/check_answer/mean": 0.9702531099319458,
"rewards/check_answer/std": 0.14307361841201782,
"step": 25
},
{
"completion_length": 488.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 609.0,
"completions/max_terminated_length": 609.0,
"completions/mean_length": 488.25,
"completions/mean_terminated_length": 488.25,
"completions/min_length": 331.0,
"completions/min_terminated_length": 331.0,
"epoch": 0.004620579349564599,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5748683214187622,
"kl": 1.0910266041755676,
"learning_rate": 1.7857142857142859e-06,
"loss": 0.0011,
"num_tokens": 79119.0,
"reward": 0.6022087335586548,
"reward_std": 0.10064248740673065,
"rewards/check_answer/mean": 0.6022087335586548,
"rewards/check_answer/std": 0.10064248740673065,
"step": 26
},
{
"completion_length": 632.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 674.0,
"completions/max_terminated_length": 674.0,
"completions/mean_length": 632.0,
"completions/mean_terminated_length": 632.0,
"completions/min_length": 600.0,
"completions/min_terminated_length": 600.0,
"epoch": 0.004798293939932468,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4055097103118896,
"kl": 0.9649907499551773,
"learning_rate": 1.8571428571428573e-06,
"loss": 0.001,
"num_tokens": 82855.0,
"reward": 0.5038251876831055,
"reward_std": 0.07822742313146591,
"rewards/check_answer/mean": 0.5038251876831055,
"rewards/check_answer/std": 0.07822742313146591,
"step": 27
},
{
"completion_length": 555.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 578.0,
"completions/max_terminated_length": 578.0,
"completions/mean_length": 555.0,
"completions/mean_terminated_length": 555.0,
"completions/min_length": 519.0,
"completions/min_terminated_length": 519.0,
"epoch": 0.004976008530300338,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5128551721572876,
"kl": 1.3766323626041412,
"learning_rate": 1.928571428571429e-06,
"loss": 0.0014,
"num_tokens": 86255.0,
"reward": 0.5897955894470215,
"reward_std": 0.14429643750190735,
"rewards/check_answer/mean": 0.5897955894470215,
"rewards/check_answer/std": 0.14429643750190735,
"step": 28
},
{
"completion_length": 351.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 398.0,
"completions/max_terminated_length": 398.0,
"completions/mean_length": 351.5,
"completions/mean_terminated_length": 351.5,
"completions/min_length": 269.0,
"completions/min_terminated_length": 269.0,
"epoch": 0.005153723120668207,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9017515182495117,
"kl": 1.2128655910491943,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.0012,
"num_tokens": 88761.0,
"reward": 0.24229314923286438,
"reward_std": 0.09404895454645157,
"rewards/check_answer/mean": 0.24229314923286438,
"rewards/check_answer/std": 0.09404894709587097,
"step": 29
},
{
"completion_length": 430.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 487.0,
"completions/max_terminated_length": 487.0,
"completions/mean_length": 430.75,
"completions/mean_terminated_length": 430.75,
"completions/min_length": 310.0,
"completions/min_terminated_length": 310.0,
"epoch": 0.005331437711036076,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.466556191444397,
"kl": 0.9134228527545929,
"learning_rate": 2.0714285714285717e-06,
"loss": 0.0009,
"num_tokens": 91664.0,
"reward": 0.7716047763824463,
"reward_std": 0.11445575952529907,
"rewards/check_answer/mean": 0.7716047763824463,
"rewards/check_answer/std": 0.11445576697587967,
"step": 30
},
{
"completion_length": 502.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 589.0,
"completions/max_terminated_length": 589.0,
"completions/mean_length": 502.25,
"completions/mean_terminated_length": 502.25,
"completions/min_length": 453.0,
"completions/min_terminated_length": 453.0,
"epoch": 0.005509152301403945,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8003754615783691,
"kl": 1.0646594911813736,
"learning_rate": 2.1428571428571427e-06,
"loss": 0.0011,
"num_tokens": 94865.0,
"reward": 0.9116252064704895,
"reward_std": 0.06001972407102585,
"rewards/check_answer/mean": 0.9116252064704895,
"rewards/check_answer/std": 0.060019705444574356,
"step": 31
},
{
"completion_length": 454.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 549.0,
"completions/max_terminated_length": 549.0,
"completions/mean_length": 454.25,
"completions/mean_terminated_length": 454.25,
"completions/min_length": 416.0,
"completions/min_terminated_length": 416.0,
"epoch": 0.005686866891771815,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.75742506980896,
"kl": 1.6527438461780548,
"learning_rate": 2.2142857142857146e-06,
"loss": 0.0017,
"num_tokens": 97818.0,
"reward": 0.5714230537414551,
"reward_std": 0.1710004061460495,
"rewards/check_answer/mean": 0.5714230537414551,
"rewards/check_answer/std": 0.1710004061460495,
"step": 32
},
{
"completion_length": 282.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 313.0,
"completions/max_terminated_length": 313.0,
"completions/mean_length": 282.75,
"completions/mean_terminated_length": 282.75,
"completions/min_length": 255.0,
"completions/min_terminated_length": 255.0,
"epoch": 0.005864581482139684,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9505561590194702,
"kl": 1.6063998937606812,
"learning_rate": 2.285714285714286e-06,
"loss": 0.0016,
"num_tokens": 100041.0,
"reward": 0.49875420331954956,
"reward_std": 0.2340739220380783,
"rewards/check_answer/mean": 0.49875420331954956,
"rewards/check_answer/std": 0.2340739369392395,
"step": 33
},
{
"completion_length": 376.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 416.0,
"completions/max_terminated_length": 416.0,
"completions/mean_length": 376.5,
"completions/mean_terminated_length": 376.5,
"completions/min_length": 345.0,
"completions/min_terminated_length": 345.0,
"epoch": 0.006042296072507553,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.888424277305603,
"kl": 1.435717135667801,
"learning_rate": 2.3571428571428574e-06,
"loss": 0.0014,
"num_tokens": 102667.0,
"reward": 0.9620379209518433,
"reward_std": 0.15259642899036407,
"rewards/check_answer/mean": 0.9620379209518433,
"rewards/check_answer/std": 0.15259642899036407,
"step": 34
},
{
"completion_length": 458.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 524.0,
"completions/max_terminated_length": 524.0,
"completions/mean_length": 458.0,
"completions/mean_terminated_length": 458.0,
"completions/min_length": 336.0,
"completions/min_terminated_length": 336.0,
"epoch": 0.006220010662875422,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9206902980804443,
"kl": 2.231161266565323,
"learning_rate": 2.428571428571429e-06,
"loss": 0.0022,
"num_tokens": 105659.0,
"reward": 0.653532862663269,
"reward_std": 0.1333208829164505,
"rewards/check_answer/mean": 0.653532862663269,
"rewards/check_answer/std": 0.1333208829164505,
"step": 35
},
{
"completion_length": 402.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 440.0,
"completions/max_terminated_length": 440.0,
"completions/mean_length": 402.75,
"completions/mean_terminated_length": 402.75,
"completions/min_length": 333.0,
"completions/min_terminated_length": 333.0,
"epoch": 0.006397725253243291,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7704061269760132,
"kl": 1.304360181093216,
"learning_rate": 2.5e-06,
"loss": 0.0013,
"num_tokens": 108382.0,
"reward": 0.6283889412879944,
"reward_std": 0.18885570764541626,
"rewards/check_answer/mean": 0.6283889412879944,
"rewards/check_answer/std": 0.18885570764541626,
"step": 36
},
{
"completion_length": 422.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 447.0,
"completions/max_terminated_length": 447.0,
"completions/mean_length": 422.5,
"completions/mean_terminated_length": 422.5,
"completions/min_length": 381.0,
"completions/min_terminated_length": 381.0,
"epoch": 0.006575439843611161,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7185007333755493,
"kl": 1.1553759276866913,
"learning_rate": 2.571428571428571e-06,
"loss": 0.0012,
"num_tokens": 111188.0,
"reward": 0.7625912427902222,
"reward_std": 0.3398703336715698,
"rewards/check_answer/mean": 0.7625912427902222,
"rewards/check_answer/std": 0.3398703336715698,
"step": 37
},
{
"completion_length": 395.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 417.0,
"completions/max_terminated_length": 417.0,
"completions/mean_length": 395.0,
"completions/mean_terminated_length": 395.0,
"completions/min_length": 376.0,
"completions/min_terminated_length": 376.0,
"epoch": 0.00675315443397903,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.813489556312561,
"kl": 1.176555871963501,
"learning_rate": 2.642857142857143e-06,
"loss": 0.0012,
"num_tokens": 113896.0,
"reward": 0.4037095606327057,
"reward_std": 0.14651015400886536,
"rewards/check_answer/mean": 0.4037095606327057,
"rewards/check_answer/std": 0.14651015400886536,
"step": 38
},
{
"completion_length": 518.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 557.0,
"completions/max_terminated_length": 557.0,
"completions/mean_length": 518.25,
"completions/mean_terminated_length": 518.25,
"completions/min_length": 475.0,
"completions/min_terminated_length": 475.0,
"epoch": 0.006930869024346899,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5206884145736694,
"kl": 1.066824734210968,
"learning_rate": 2.7142857142857144e-06,
"loss": 0.0011,
"num_tokens": 117121.0,
"reward": 0.6004906892776489,
"reward_std": 0.06248007342219353,
"rewards/check_answer/mean": 0.6004906892776489,
"rewards/check_answer/std": 0.06248008832335472,
"step": 39
},
{
"completion_length": 586.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 633.0,
"completions/max_terminated_length": 633.0,
"completions/mean_length": 586.5,
"completions/mean_terminated_length": 586.5,
"completions/min_length": 515.0,
"completions/min_terminated_length": 515.0,
"epoch": 0.007108583614714768,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2672127485275269,
"kl": 1.0179399698972702,
"learning_rate": 2.785714285714286e-06,
"loss": 0.001,
"num_tokens": 120671.0,
"reward": 0.7895050048828125,
"reward_std": 0.058173682540655136,
"rewards/check_answer/mean": 0.7895050048828125,
"rewards/check_answer/std": 0.058173682540655136,
"step": 40
},
{
"completion_length": 474.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 567.0,
"completions/max_terminated_length": 567.0,
"completions/mean_length": 474.5,
"completions/mean_terminated_length": 474.5,
"completions/min_length": 424.0,
"completions/min_terminated_length": 424.0,
"epoch": 0.007286298205082638,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.59950852394104,
"kl": 1.0878016501665115,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.0011,
"num_tokens": 123717.0,
"reward": 0.6761389374732971,
"reward_std": 0.04523153230547905,
"rewards/check_answer/mean": 0.6761389374732971,
"rewards/check_answer/std": 0.04523153603076935,
"step": 41
},
{
"completion_length": 589.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 659.0,
"completions/max_terminated_length": 659.0,
"completions/mean_length": 589.25,
"completions/mean_terminated_length": 589.25,
"completions/min_length": 549.0,
"completions/min_terminated_length": 549.0,
"epoch": 0.007464012795450507,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.543569564819336,
"kl": 1.1034990847110748,
"learning_rate": 2.928571428571429e-06,
"loss": 0.0011,
"num_tokens": 127262.0,
"reward": 0.5081104636192322,
"reward_std": 0.012905855663120747,
"rewards/check_answer/mean": 0.5081104636192322,
"rewards/check_answer/std": 0.012905837036669254,
"step": 42
},
{
"completion_length": 294.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 352.0,
"completions/max_terminated_length": 352.0,
"completions/mean_length": 294.25,
"completions/mean_terminated_length": 294.25,
"completions/min_length": 267.0,
"completions/min_terminated_length": 267.0,
"epoch": 0.007641727385818376,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7369331121444702,
"kl": 1.1378049850463867,
"learning_rate": 3e-06,
"loss": 0.0011,
"num_tokens": 129555.0,
"reward": 1.392117977142334,
"reward_std": 0.21576404571533203,
"rewards/check_answer/mean": 1.392117977142334,
"rewards/check_answer/std": 0.21576404571533203,
"step": 43
},
{
"completion_length": 610.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 714.0,
"completions/max_terminated_length": 714.0,
"completions/mean_length": 610.0,
"completions/mean_terminated_length": 610.0,
"completions/min_length": 521.0,
"completions/min_terminated_length": 521.0,
"epoch": 0.007819441976186245,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3297312259674072,
"kl": 0.7992035746574402,
"learning_rate": 3.071428571428572e-06,
"loss": 0.0008,
"num_tokens": 133243.0,
"reward": 0.379222571849823,
"reward_std": 0.07391934096813202,
"rewards/check_answer/mean": 0.379222571849823,
"rewards/check_answer/std": 0.07391934841871262,
"step": 44
},
{
"completion_length": 526.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 602.0,
"completions/max_terminated_length": 602.0,
"completions/mean_length": 526.0,
"completions/mean_terminated_length": 526.0,
"completions/min_length": 445.0,
"completions/min_terminated_length": 445.0,
"epoch": 0.007997156566554114,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3624135255813599,
"kl": 1.055791214108467,
"learning_rate": 3.142857142857143e-06,
"loss": 0.0011,
"num_tokens": 136531.0,
"reward": 0.5060752630233765,
"reward_std": 0.07115290313959122,
"rewards/check_answer/mean": 0.5060752630233765,
"rewards/check_answer/std": 0.07115291059017181,
"step": 45
},
{
"completion_length": 347.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 435.0,
"completions/max_terminated_length": 435.0,
"completions/mean_length": 347.75,
"completions/mean_terminated_length": 347.75,
"completions/min_length": 205.0,
"completions/min_terminated_length": 205.0,
"epoch": 0.008174871156921983,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5305379629135132,
"kl": 1.2677884995937347,
"learning_rate": 3.2142857142857147e-06,
"loss": 0.0013,
"num_tokens": 139030.0,
"reward": 0.5792946815490723,
"reward_std": 0.07226314395666122,
"rewards/check_answer/mean": 0.5792946815490723,
"rewards/check_answer/std": 0.07226314395666122,
"step": 46
},
{
"completion_length": 469.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 476.0,
"completions/max_terminated_length": 476.0,
"completions/mean_length": 469.75,
"completions/mean_terminated_length": 469.75,
"completions/min_length": 465.0,
"completions/min_terminated_length": 465.0,
"epoch": 0.008352585747289852,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5030322074890137,
"kl": 1.171961322426796,
"learning_rate": 3.285714285714286e-06,
"loss": 0.0012,
"num_tokens": 142061.0,
"reward": 1.0066816806793213,
"reward_std": 0.16484001278877258,
"rewards/check_answer/mean": 1.0066816806793213,
"rewards/check_answer/std": 0.16484001278877258,
"step": 47
},
{
"completion_length": 340.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 382.0,
"completions/max_terminated_length": 382.0,
"completions/mean_length": 340.5,
"completions/mean_terminated_length": 340.5,
"completions/min_length": 312.0,
"completions/min_terminated_length": 312.0,
"epoch": 0.008530300337657722,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6775590181350708,
"kl": 1.4609983712434769,
"learning_rate": 3.357142857142857e-06,
"loss": 0.0015,
"num_tokens": 144583.0,
"reward": 1.2281973361968994,
"reward_std": 0.18847818672657013,
"rewards/check_answer/mean": 1.2281973361968994,
"rewards/check_answer/std": 0.18847818672657013,
"step": 48
},
{
"completion_length": 298.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 315.0,
"completions/max_terminated_length": 315.0,
"completions/mean_length": 298.75,
"completions/mean_terminated_length": 298.75,
"completions/min_length": 290.0,
"completions/min_terminated_length": 290.0,
"epoch": 0.00870801492802559,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6502373218536377,
"kl": 1.2542754113674164,
"learning_rate": 3.428571428571429e-06,
"loss": 0.0013,
"num_tokens": 146890.0,
"reward": 0.5080112814903259,
"reward_std": 0.03593429550528526,
"rewards/check_answer/mean": 0.5080112814903259,
"rewards/check_answer/std": 0.03593427687883377,
"step": 49
},
{
"completion_length": 541.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 612.0,
"completions/max_terminated_length": 612.0,
"completions/mean_length": 541.25,
"completions/mean_terminated_length": 541.25,
"completions/min_length": 496.0,
"completions/min_terminated_length": 496.0,
"epoch": 0.00888572951839346,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5773475170135498,
"kl": 1.1764096915721893,
"learning_rate": 3.5e-06,
"loss": 0.0012,
"num_tokens": 150239.0,
"reward": 0.5298885107040405,
"reward_std": 0.062375664710998535,
"rewards/check_answer/mean": 0.5298885107040405,
"rewards/check_answer/std": 0.06237566098570824,
"step": 50
},
{
"completion_length": 473.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 491.0,
"completions/max_terminated_length": 491.0,
"completions/mean_length": 473.25,
"completions/mean_terminated_length": 473.25,
"completions/min_length": 445.0,
"completions/min_terminated_length": 445.0,
"epoch": 0.00906344410876133,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5712735652923584,
"kl": 1.2196315228939056,
"learning_rate": 3.5714285714285718e-06,
"loss": 0.0012,
"num_tokens": 153272.0,
"reward": 0.9013558626174927,
"reward_std": 0.12637704610824585,
"rewards/check_answer/mean": 0.9013558626174927,
"rewards/check_answer/std": 0.12637704610824585,
"step": 51
},
{
"completion_length": 430.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 444.0,
"completions/max_terminated_length": 444.0,
"completions/mean_length": 430.75,
"completions/mean_terminated_length": 430.75,
"completions/min_length": 408.0,
"completions/min_terminated_length": 408.0,
"epoch": 0.009241158699129198,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8517780303955078,
"kl": 1.4242485463619232,
"learning_rate": 3.642857142857143e-06,
"loss": 0.0014,
"num_tokens": 156111.0,
"reward": 0.48627614974975586,
"reward_std": 0.10064783692359924,
"rewards/check_answer/mean": 0.48627614974975586,
"rewards/check_answer/std": 0.10064783692359924,
"step": 52
},
{
"completion_length": 289.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 320.0,
"completions/max_terminated_length": 320.0,
"completions/mean_length": 289.5,
"completions/mean_terminated_length": 289.5,
"completions/min_length": 253.0,
"completions/min_terminated_length": 253.0,
"epoch": 0.009418873289497068,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9996415376663208,
"kl": 1.5143893957138062,
"learning_rate": 3.7142857142857146e-06,
"loss": 0.0015,
"num_tokens": 158361.0,
"reward": 0.5390114188194275,
"reward_std": 0.6577240824699402,
"rewards/check_answer/mean": 0.5390114188194275,
"rewards/check_answer/std": 0.657724142074585,
"step": 53
},
{
"completion_length": 443.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 520.0,
"completions/max_terminated_length": 520.0,
"completions/mean_length": 443.25,
"completions/mean_terminated_length": 443.25,
"completions/min_length": 368.0,
"completions/min_terminated_length": 368.0,
"epoch": 0.009596587879864936,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3870877027511597,
"kl": 1.062680721282959,
"learning_rate": 3.785714285714286e-06,
"loss": 0.0011,
"num_tokens": 161294.0,
"reward": 0.5653637647628784,
"reward_std": 0.06738435477018356,
"rewards/check_answer/mean": 0.5653637647628784,
"rewards/check_answer/std": 0.06738435477018356,
"step": 54
},
{
"completion_length": 391.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 449.0,
"completions/max_terminated_length": 449.0,
"completions/mean_length": 391.75,
"completions/mean_terminated_length": 391.75,
"completions/min_length": 333.0,
"completions/min_terminated_length": 333.0,
"epoch": 0.009774302470232806,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6596442461013794,
"kl": 0.9484704583883286,
"learning_rate": 3.857142857142858e-06,
"loss": 0.0009,
"num_tokens": 164025.0,
"reward": 0.7971336245536804,
"reward_std": 0.12026101350784302,
"rewards/check_answer/mean": 0.7971336245536804,
"rewards/check_answer/std": 0.12026101350784302,
"step": 55
},
{
"completion_length": 347.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 444.0,
"completions/max_terminated_length": 444.0,
"completions/mean_length": 347.0,
"completions/mean_terminated_length": 347.0,
"completions/min_length": 270.0,
"completions/min_terminated_length": 270.0,
"epoch": 0.009952017060600676,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.146622896194458,
"kl": 1.1937618106603622,
"learning_rate": 3.928571428571429e-06,
"loss": 0.0012,
"num_tokens": 166537.0,
"reward": 0.2618507444858551,
"reward_std": 0.054494522511959076,
"rewards/check_answer/mean": 0.2618507444858551,
"rewards/check_answer/std": 0.054494522511959076,
"step": 56
},
{
"completion_length": 520.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 584.0,
"completions/max_terminated_length": 584.0,
"completions/mean_length": 520.5,
"completions/mean_terminated_length": 520.5,
"completions/min_length": 436.0,
"completions/min_terminated_length": 436.0,
"epoch": 0.010129731650968544,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5688644647598267,
"kl": 1.048645719885826,
"learning_rate": 4.000000000000001e-06,
"loss": 0.001,
"num_tokens": 169803.0,
"reward": 0.9876464605331421,
"reward_std": 0.25668615102767944,
"rewards/check_answer/mean": 0.9876464605331421,
"rewards/check_answer/std": 0.25668615102767944,
"step": 57
},
{
"completion_length": 541.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 610.0,
"completions/max_terminated_length": 610.0,
"completions/mean_length": 541.0,
"completions/mean_terminated_length": 541.0,
"completions/min_length": 483.0,
"completions/min_terminated_length": 483.0,
"epoch": 0.010307446241336414,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2746442556381226,
"kl": 0.8855240046977997,
"learning_rate": 4.071428571428572e-06,
"loss": 0.0009,
"num_tokens": 173207.0,
"reward": 0.3755004405975342,
"reward_std": 0.049010019749403,
"rewards/check_answer/mean": 0.3755004405975342,
"rewards/check_answer/std": 0.04901003837585449,
"step": 58
},
{
"completion_length": 524.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 577.0,
"completions/max_terminated_length": 577.0,
"completions/mean_length": 524.0,
"completions/mean_terminated_length": 524.0,
"completions/min_length": 450.0,
"completions/min_terminated_length": 450.0,
"epoch": 0.010485160831704282,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5979560613632202,
"kl": 1.034047693014145,
"learning_rate": 4.1428571428571435e-06,
"loss": 0.001,
"num_tokens": 176475.0,
"reward": 0.8849939703941345,
"reward_std": 0.11704953014850616,
"rewards/check_answer/mean": 0.8849939703941345,
"rewards/check_answer/std": 0.11704951524734497,
"step": 59
},
{
"completion_length": 471.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 558.0,
"completions/max_terminated_length": 558.0,
"completions/mean_length": 471.25,
"completions/mean_terminated_length": 471.25,
"completions/min_length": 372.0,
"completions/min_terminated_length": 372.0,
"epoch": 0.010662875422072152,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5512830018997192,
"kl": 1.3353438973426819,
"learning_rate": 4.2142857142857145e-06,
"loss": 0.0013,
"num_tokens": 179544.0,
"reward": 0.4594007134437561,
"reward_std": 0.08352126181125641,
"rewards/check_answer/mean": 0.4594007134437561,
"rewards/check_answer/std": 0.08352126181125641,
"step": 60
},
{
"completion_length": 560.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 688.0,
"completions/max_terminated_length": 688.0,
"completions/mean_length": 560.5,
"completions/mean_terminated_length": 560.5,
"completions/min_length": 406.0,
"completions/min_terminated_length": 406.0,
"epoch": 0.010840590012440022,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7048546075820923,
"kl": 0.9816142320632935,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.001,
"num_tokens": 182994.0,
"reward": 0.38246601819992065,
"reward_std": 0.13571514189243317,
"rewards/check_answer/mean": 0.38246601819992065,
"rewards/check_answer/std": 0.13571515679359436,
"step": 61
},
{
"completion_length": 411.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 489.0,
"completions/max_terminated_length": 489.0,
"completions/mean_length": 411.25,
"completions/mean_terminated_length": 411.25,
"completions/min_length": 361.0,
"completions/min_terminated_length": 361.0,
"epoch": 0.01101830460280789,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.981577754020691,
"kl": 1.2702476382255554,
"learning_rate": 4.357142857142857e-06,
"loss": 0.0013,
"num_tokens": 185755.0,
"reward": 0.6234804391860962,
"reward_std": 0.2467672973871231,
"rewards/check_answer/mean": 0.6234804391860962,
"rewards/check_answer/std": 0.2467672973871231,
"step": 62
},
{
"completion_length": 511.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 589.0,
"completions/max_terminated_length": 589.0,
"completions/mean_length": 511.5,
"completions/mean_terminated_length": 511.5,
"completions/min_length": 388.0,
"completions/min_terminated_length": 388.0,
"epoch": 0.01119601919317576,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.365710735321045,
"kl": 0.9097413569688797,
"learning_rate": 4.428571428571429e-06,
"loss": 0.0009,
"num_tokens": 189005.0,
"reward": 0.4179466962814331,
"reward_std": 0.04861922189593315,
"rewards/check_answer/mean": 0.4179466962814331,
"rewards/check_answer/std": 0.04861922189593315,
"step": 63
},
{
"completion_length": 440.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 484.0,
"completions/max_terminated_length": 484.0,
"completions/mean_length": 440.0,
"completions/mean_terminated_length": 440.0,
"completions/min_length": 397.0,
"completions/min_terminated_length": 397.0,
"epoch": 0.01137373378354363,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8046718835830688,
"kl": 1.3415338695049286,
"learning_rate": 4.5e-06,
"loss": 0.0013,
"num_tokens": 191909.0,
"reward": 0.4736880362033844,
"reward_std": 0.09703974425792694,
"rewards/check_answer/mean": 0.4736880362033844,
"rewards/check_answer/std": 0.09703974425792694,
"step": 64
},
{
"completion_length": 411.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 509.0,
"completions/max_terminated_length": 509.0,
"completions/mean_length": 411.5,
"completions/mean_terminated_length": 411.5,
"completions/min_length": 299.0,
"completions/min_terminated_length": 299.0,
"epoch": 0.011551448373911498,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6717690229415894,
"kl": 1.1143593043088913,
"learning_rate": 4.571428571428572e-06,
"loss": 0.0011,
"num_tokens": 194675.0,
"reward": 0.7849711775779724,
"reward_std": 0.3310997784137726,
"rewards/check_answer/mean": 0.7849711775779724,
"rewards/check_answer/std": 0.3310997784137726,
"step": 65
},
{
"completion_length": 484.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 504.0,
"completions/max_terminated_length": 504.0,
"completions/mean_length": 484.0,
"completions/mean_terminated_length": 484.0,
"completions/min_length": 461.0,
"completions/min_terminated_length": 461.0,
"epoch": 0.011729162964279368,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3961551189422607,
"kl": 0.8092105090618134,
"learning_rate": 4.642857142857144e-06,
"loss": 0.0008,
"num_tokens": 197815.0,
"reward": 0.5569653511047363,
"reward_std": 0.11439482122659683,
"rewards/check_answer/mean": 0.5569653511047363,
"rewards/check_answer/std": 0.11439482867717743,
"step": 66
},
{
"completion_length": 580.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 598.0,
"completions/max_terminated_length": 598.0,
"completions/mean_length": 580.0,
"completions/mean_terminated_length": 580.0,
"completions/min_length": 553.0,
"completions/min_terminated_length": 553.0,
"epoch": 0.011906877554647236,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.20498788356781,
"kl": 0.9006475359201431,
"learning_rate": 4.714285714285715e-06,
"loss": 0.0009,
"num_tokens": 201331.0,
"reward": 0.5501281023025513,
"reward_std": 0.05327435955405235,
"rewards/check_answer/mean": 0.5501281023025513,
"rewards/check_answer/std": 0.053274352103471756,
"step": 67
},
{
"completion_length": 408.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 521.0,
"completions/max_terminated_length": 521.0,
"completions/mean_length": 408.0,
"completions/mean_terminated_length": 408.0,
"completions/min_length": 309.0,
"completions/min_terminated_length": 309.0,
"epoch": 0.012084592145015106,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.882188081741333,
"kl": 1.1671920120716095,
"learning_rate": 4.785714285714287e-06,
"loss": 0.0012,
"num_tokens": 204115.0,
"reward": 1.1435757875442505,
"reward_std": 0.22389693558216095,
"rewards/check_answer/mean": 1.1435757875442505,
"rewards/check_answer/std": 0.22389698028564453,
"step": 68
},
{
"completion_length": 481.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 590.0,
"completions/max_terminated_length": 590.0,
"completions/mean_length": 481.25,
"completions/mean_terminated_length": 481.25,
"completions/min_length": 363.0,
"completions/min_terminated_length": 363.0,
"epoch": 0.012262306735382976,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.533947229385376,
"kl": 1.1329405903816223,
"learning_rate": 4.857142857142858e-06,
"loss": 0.0011,
"num_tokens": 207200.0,
"reward": 0.44344037771224976,
"reward_std": 0.09549304842948914,
"rewards/check_answer/mean": 0.44344037771224976,
"rewards/check_answer/std": 0.09549304842948914,
"step": 69
},
{
"completion_length": 530.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 651.0,
"completions/max_terminated_length": 651.0,
"completions/mean_length": 530.0,
"completions/mean_terminated_length": 530.0,
"completions/min_length": 380.0,
"completions/min_terminated_length": 380.0,
"epoch": 0.012440021325750844,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.383154273033142,
"kl": 0.8603696227073669,
"learning_rate": 4.928571428571429e-06,
"loss": 0.0009,
"num_tokens": 210516.0,
"reward": 0.3628694415092468,
"reward_std": 0.04750685766339302,
"rewards/check_answer/mean": 0.3628694415092468,
"rewards/check_answer/std": 0.04750685766339302,
"step": 70
},
{
"completion_length": 467.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 501.0,
"completions/max_terminated_length": 501.0,
"completions/mean_length": 467.25,
"completions/mean_terminated_length": 467.25,
"completions/min_length": 428.0,
"completions/min_terminated_length": 428.0,
"epoch": 0.012617735916118714,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7030493021011353,
"kl": 1.256488174200058,
"learning_rate": 5e-06,
"loss": 0.0013,
"num_tokens": 213513.0,
"reward": 0.9228349924087524,
"reward_std": 0.2585166394710541,
"rewards/check_answer/mean": 0.9228349924087524,
"rewards/check_answer/std": 0.2585166394710541,
"step": 71
},
{
"completion_length": 380.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 463.0,
"completions/max_terminated_length": 463.0,
"completions/mean_length": 380.5,
"completions/mean_terminated_length": 380.5,
"completions/min_length": 317.0,
"completions/min_terminated_length": 317.0,
"epoch": 0.012795450506486582,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8185611963272095,
"kl": 1.088569387793541,
"learning_rate": 4.992063492063493e-06,
"loss": 0.0011,
"num_tokens": 216187.0,
"reward": 0.1358739137649536,
"reward_std": 0.03403108939528465,
"rewards/check_answer/mean": 0.1358739137649536,
"rewards/check_answer/std": 0.03403108939528465,
"step": 72
},
{
"completion_length": 391.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 472.0,
"completions/max_terminated_length": 472.0,
"completions/mean_length": 391.5,
"completions/mean_terminated_length": 391.5,
"completions/min_length": 352.0,
"completions/min_terminated_length": 352.0,
"epoch": 0.012973165096854452,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7923240661621094,
"kl": 1.187307983636856,
"learning_rate": 4.9841269841269845e-06,
"loss": 0.0012,
"num_tokens": 218881.0,
"reward": 0.4668271541595459,
"reward_std": 0.28230786323547363,
"rewards/check_answer/mean": 0.4668271541595459,
"rewards/check_answer/std": 0.28230786323547363,
"step": 73
},
{
"completion_length": 459.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 560.0,
"completions/max_terminated_length": 560.0,
"completions/mean_length": 459.5,
"completions/mean_terminated_length": 459.5,
"completions/min_length": 314.0,
"completions/min_terminated_length": 314.0,
"epoch": 0.013150879687222321,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5349751710891724,
"kl": 1.160512775182724,
"learning_rate": 4.976190476190477e-06,
"loss": 0.0012,
"num_tokens": 221875.0,
"reward": 0.5796365737915039,
"reward_std": 0.03070678748190403,
"rewards/check_answer/mean": 0.5796365737915039,
"rewards/check_answer/std": 0.03070676326751709,
"step": 74
},
{
"completion_length": 275.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 379.0,
"completions/max_terminated_length": 379.0,
"completions/mean_length": 275.0,
"completions/mean_terminated_length": 275.0,
"completions/min_length": 194.0,
"completions/min_terminated_length": 194.0,
"epoch": 0.01332859427759019,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9937189817428589,
"kl": 1.2936433553695679,
"learning_rate": 4.968253968253969e-06,
"loss": 0.0013,
"num_tokens": 224087.0,
"reward": 0.618080735206604,
"reward_std": 0.028470057994127274,
"rewards/check_answer/mean": 0.618080735206604,
"rewards/check_answer/std": 0.028470059856772423,
"step": 75
},
{
"completion_length": 561.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 883.0,
"completions/max_terminated_length": 883.0,
"completions/mean_length": 561.5,
"completions/mean_terminated_length": 561.5,
"completions/min_length": 382.0,
"completions/min_terminated_length": 382.0,
"epoch": 0.01350630886795806,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2865067720413208,
"kl": 1.0802096128463745,
"learning_rate": 4.960317460317461e-06,
"loss": 0.0011,
"num_tokens": 227501.0,
"reward": 0.45027637481689453,
"reward_std": 0.3139820694923401,
"rewards/check_answer/mean": 0.45027637481689453,
"rewards/check_answer/std": 0.3139820694923401,
"step": 76
},
{
"completion_length": 482.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 548.0,
"completions/max_terminated_length": 548.0,
"completions/mean_length": 482.0,
"completions/mean_terminated_length": 482.0,
"completions/min_length": 378.0,
"completions/min_terminated_length": 378.0,
"epoch": 0.01368402345832593,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5967481136322021,
"kl": 2.116203159093857,
"learning_rate": 4.952380952380953e-06,
"loss": 0.0021,
"num_tokens": 230589.0,
"reward": 0.8557864427566528,
"reward_std": 0.07187496870756149,
"rewards/check_answer/mean": 0.8557864427566528,
"rewards/check_answer/std": 0.0718749538064003,
"step": 77
},
{
"completion_length": 435.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 450.0,
"completions/max_terminated_length": 450.0,
"completions/mean_length": 435.0,
"completions/mean_terminated_length": 435.0,
"completions/min_length": 410.0,
"completions/min_terminated_length": 410.0,
"epoch": 0.013861738048693797,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7964273691177368,
"kl": 1.6678560972213745,
"learning_rate": 4.944444444444445e-06,
"loss": 0.0017,
"num_tokens": 233473.0,
"reward": 0.7173020839691162,
"reward_std": 0.1675439327955246,
"rewards/check_answer/mean": 0.7173020839691162,
"rewards/check_answer/std": 0.1675439327955246,
"step": 78
},
{
"completion_length": 329.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 370.0,
"completions/max_terminated_length": 370.0,
"completions/mean_length": 329.5,
"completions/mean_terminated_length": 329.5,
"completions/min_length": 297.0,
"completions/min_terminated_length": 297.0,
"epoch": 0.014039452639061667,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.785151481628418,
"kl": 1.2658391296863556,
"learning_rate": 4.936507936507937e-06,
"loss": 0.0013,
"num_tokens": 235919.0,
"reward": 0.3208128809928894,
"reward_std": 0.5496598482131958,
"rewards/check_answer/mean": 0.3208128809928894,
"rewards/check_answer/std": 0.5496599078178406,
"step": 79
},
{
"completion_length": 487.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 535.0,
"completions/max_terminated_length": 535.0,
"completions/mean_length": 487.25,
"completions/mean_terminated_length": 487.25,
"completions/min_length": 438.0,
"completions/min_terminated_length": 438.0,
"epoch": 0.014217167229429535,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4646213054656982,
"kl": 1.082797721028328,
"learning_rate": 4.928571428571429e-06,
"loss": 0.0011,
"num_tokens": 239072.0,
"reward": 0.47449803352355957,
"reward_std": 0.05666935071349144,
"rewards/check_answer/mean": 0.47449803352355957,
"rewards/check_answer/std": 0.05666935071349144,
"step": 80
},
{
"completion_length": 512.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 579.0,
"completions/max_terminated_length": 579.0,
"completions/mean_length": 512.0,
"completions/mean_terminated_length": 512.0,
"completions/min_length": 441.0,
"completions/min_terminated_length": 441.0,
"epoch": 0.014394881819797405,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.374875545501709,
"kl": 0.866473376750946,
"learning_rate": 4.920634920634921e-06,
"loss": 0.0009,
"num_tokens": 242296.0,
"reward": 0.9352073073387146,
"reward_std": 0.054083820432424545,
"rewards/check_answer/mean": 0.9352073073387146,
"rewards/check_answer/std": 0.054083820432424545,
"step": 81
},
{
"completion_length": 455.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 562.0,
"completions/max_terminated_length": 562.0,
"completions/mean_length": 455.5,
"completions/mean_terminated_length": 455.5,
"completions/min_length": 386.0,
"completions/min_terminated_length": 386.0,
"epoch": 0.014572596410165275,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3765476942062378,
"kl": 0.9197855442762375,
"learning_rate": 4.912698412698413e-06,
"loss": 0.0009,
"num_tokens": 245302.0,
"reward": 1.100656270980835,
"reward_std": 0.04409698769450188,
"rewards/check_answer/mean": 1.100656270980835,
"rewards/check_answer/std": 0.04409695789217949,
"step": 82
},
{
"completion_length": 396.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 412.0,
"completions/max_terminated_length": 412.0,
"completions/mean_length": 396.0,
"completions/mean_terminated_length": 396.0,
"completions/min_length": 377.0,
"completions/min_terminated_length": 377.0,
"epoch": 0.014750311000533143,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.643187165260315,
"kl": 1.173012375831604,
"learning_rate": 4.904761904761905e-06,
"loss": 0.0012,
"num_tokens": 248034.0,
"reward": 1.026364803314209,
"reward_std": 0.2264280915260315,
"rewards/check_answer/mean": 1.026364803314209,
"rewards/check_answer/std": 0.2264280915260315,
"step": 83
},
{
"completion_length": 508.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 557.0,
"completions/max_terminated_length": 557.0,
"completions/mean_length": 508.25,
"completions/mean_terminated_length": 508.25,
"completions/min_length": 484.0,
"completions/min_terminated_length": 484.0,
"epoch": 0.014928025590901013,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.608371376991272,
"kl": 1.112755537033081,
"learning_rate": 4.896825396825397e-06,
"loss": 0.0011,
"num_tokens": 251223.0,
"reward": 0.9159030914306641,
"reward_std": 0.09535390138626099,
"rewards/check_answer/mean": 0.9159030914306641,
"rewards/check_answer/std": 0.09535388648509979,
"step": 84
},
{
"completion_length": 377.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 478.0,
"completions/max_terminated_length": 478.0,
"completions/mean_length": 377.25,
"completions/mean_terminated_length": 377.25,
"completions/min_length": 260.0,
"completions/min_terminated_length": 260.0,
"epoch": 0.015105740181268883,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8007338047027588,
"kl": 1.063372939825058,
"learning_rate": 4.888888888888889e-06,
"loss": 0.0011,
"num_tokens": 253880.0,
"reward": 0.5686249732971191,
"reward_std": 0.1699865609407425,
"rewards/check_answer/mean": 0.5686249732971191,
"rewards/check_answer/std": 0.1699865609407425,
"step": 85
},
{
"completion_length": 625.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 722.0,
"completions/max_terminated_length": 722.0,
"completions/mean_length": 625.0,
"completions/mean_terminated_length": 625.0,
"completions/min_length": 504.0,
"completions/min_terminated_length": 504.0,
"epoch": 0.015283454771636751,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1691299676895142,
"kl": 0.7268833667039871,
"learning_rate": 4.880952380952381e-06,
"loss": 0.0007,
"num_tokens": 257636.0,
"reward": 0.6261911392211914,
"reward_std": 0.09556934982538223,
"rewards/check_answer/mean": 0.6261911392211914,
"rewards/check_answer/std": 0.09556934982538223,
"step": 86
},
{
"completion_length": 362.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 411.0,
"completions/max_terminated_length": 411.0,
"completions/mean_length": 362.25,
"completions/mean_terminated_length": 362.25,
"completions/min_length": 238.0,
"completions/min_terminated_length": 238.0,
"epoch": 0.015461169362004621,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9808728694915771,
"kl": 1.2715768963098526,
"learning_rate": 4.8730158730158735e-06,
"loss": 0.0013,
"num_tokens": 260217.0,
"reward": 0.6537027359008789,
"reward_std": 0.23676364123821259,
"rewards/check_answer/mean": 0.6537027359008789,
"rewards/check_answer/std": 0.23676365613937378,
"step": 87
},
{
"completion_length": 342.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 396.0,
"completions/max_terminated_length": 396.0,
"completions/mean_length": 342.5,
"completions/mean_terminated_length": 342.5,
"completions/min_length": 252.0,
"completions/min_terminated_length": 252.0,
"epoch": 0.01563888395237249,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8982717990875244,
"kl": 1.0811645984649658,
"learning_rate": 4.865079365079365e-06,
"loss": 0.0011,
"num_tokens": 262711.0,
"reward": 0.8981786966323853,
"reward_std": 0.07228042185306549,
"rewards/check_answer/mean": 0.8981786966323853,
"rewards/check_answer/std": 0.0722803920507431,
"step": 88
},
{
"completion_length": 317.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 469.0,
"completions/max_terminated_length": 469.0,
"completions/mean_length": 317.25,
"completions/mean_terminated_length": 317.25,
"completions/min_length": 227.0,
"completions/min_terminated_length": 227.0,
"epoch": 0.015816598542740357,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.1990370750427246,
"kl": 1.3140311539173126,
"learning_rate": 4.857142857142858e-06,
"loss": 0.0013,
"num_tokens": 265088.0,
"reward": 0.32814788818359375,
"reward_std": 0.029493404552340508,
"rewards/check_answer/mean": 0.32814788818359375,
"rewards/check_answer/std": 0.029493410140275955,
"step": 89
},
{
"completion_length": 323.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 406.0,
"completions/max_terminated_length": 406.0,
"completions/mean_length": 323.5,
"completions/mean_terminated_length": 323.5,
"completions/min_length": 250.0,
"completions/min_terminated_length": 250.0,
"epoch": 0.01599431313310823,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9783636331558228,
"kl": 2.316634923219681,
"learning_rate": 4.849206349206349e-06,
"loss": 0.0023,
"num_tokens": 267474.0,
"reward": 1.1331236362457275,
"reward_std": 0.43865159153938293,
"rewards/check_answer/mean": 1.1331236362457275,
"rewards/check_answer/std": 0.4386516213417053,
"step": 90
},
{
"completion_length": 530.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 547.0,
"completions/max_terminated_length": 547.0,
"completions/mean_length": 530.5,
"completions/mean_terminated_length": 530.5,
"completions/min_length": 487.0,
"completions/min_terminated_length": 487.0,
"epoch": 0.016172027723476097,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.401001214981079,
"kl": 0.944896399974823,
"learning_rate": 4.841269841269842e-06,
"loss": 0.0009,
"num_tokens": 270800.0,
"reward": 0.6363397240638733,
"reward_std": 0.08994114398956299,
"rewards/check_answer/mean": 0.6363397240638733,
"rewards/check_answer/std": 0.08994113653898239,
"step": 91
},
{
"completion_length": 500.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 569.0,
"completions/max_terminated_length": 569.0,
"completions/mean_length": 500.5,
"completions/mean_terminated_length": 500.5,
"completions/min_length": 432.0,
"completions/min_terminated_length": 432.0,
"epoch": 0.016349742313843965,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4314804077148438,
"kl": 0.9742838144302368,
"learning_rate": 4.833333333333333e-06,
"loss": 0.001,
"num_tokens": 273942.0,
"reward": 0.42001938819885254,
"reward_std": 0.056012097746133804,
"rewards/check_answer/mean": 0.42001938819885254,
"rewards/check_answer/std": 0.056012097746133804,
"step": 92
},
{
"completion_length": 535.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 596.0,
"completions/max_terminated_length": 596.0,
"completions/mean_length": 535.5,
"completions/mean_terminated_length": 535.5,
"completions/min_length": 408.0,
"completions/min_terminated_length": 408.0,
"epoch": 0.016527456904211837,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5192902088165283,
"kl": 1.0595851242542267,
"learning_rate": 4.825396825396826e-06,
"loss": 0.0011,
"num_tokens": 277252.0,
"reward": 0.4132889211177826,
"reward_std": 0.030004048720002174,
"rewards/check_answer/mean": 0.4132889211177826,
"rewards/check_answer/std": 0.030004046857357025,
"step": 93
},
{
"completion_length": 585.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 723.0,
"completions/max_terminated_length": 723.0,
"completions/mean_length": 585.25,
"completions/mean_terminated_length": 585.25,
"completions/min_length": 510.0,
"completions/min_terminated_length": 510.0,
"epoch": 0.016705171494579705,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1913037300109863,
"kl": 0.7135125249624252,
"learning_rate": 4.817460317460318e-06,
"loss": 0.0007,
"num_tokens": 280825.0,
"reward": 0.7782343626022339,
"reward_std": 0.09686165302991867,
"rewards/check_answer/mean": 0.7782343626022339,
"rewards/check_answer/std": 0.09686163812875748,
"step": 94
},
{
"completion_length": 390.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 437.0,
"completions/max_terminated_length": 437.0,
"completions/mean_length": 390.75,
"completions/mean_terminated_length": 390.75,
"completions/min_length": 344.0,
"completions/min_terminated_length": 344.0,
"epoch": 0.016882886084947573,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8841480016708374,
"kl": 0.9209953248500824,
"learning_rate": 4.80952380952381e-06,
"loss": 0.0009,
"num_tokens": 283544.0,
"reward": 0.5185092687606812,
"reward_std": 0.04221953824162483,
"rewards/check_answer/mean": 0.5185092687606812,
"rewards/check_answer/std": 0.04221952706575394,
"step": 95
},
{
"completion_length": 441.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 476.0,
"completions/max_terminated_length": 476.0,
"completions/mean_length": 441.75,
"completions/mean_terminated_length": 441.75,
"completions/min_length": 415.0,
"completions/min_terminated_length": 415.0,
"epoch": 0.017060600675315445,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6601427793502808,
"kl": 0.9485969841480255,
"learning_rate": 4.8015873015873025e-06,
"loss": 0.0009,
"num_tokens": 286475.0,
"reward": 0.5999540686607361,
"reward_std": 0.1883406788110733,
"rewards/check_answer/mean": 0.5999540686607361,
"rewards/check_answer/std": 0.1883406639099121,
"step": 96
},
{
"completion_length": 429.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 567.0,
"completions/max_terminated_length": 567.0,
"completions/mean_length": 429.75,
"completions/mean_terminated_length": 429.75,
"completions/min_length": 245.0,
"completions/min_terminated_length": 245.0,
"epoch": 0.017238315265683313,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4426722526550293,
"kl": 1.0458677113056183,
"learning_rate": 4.793650793650794e-06,
"loss": 0.001,
"num_tokens": 289326.0,
"reward": 0.33017024397850037,
"reward_std": 0.11357025057077408,
"rewards/check_answer/mean": 0.33017024397850037,
"rewards/check_answer/std": 0.11357025802135468,
"step": 97
},
{
"completion_length": 481.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 531.0,
"completions/max_terminated_length": 531.0,
"completions/mean_length": 481.75,
"completions/mean_terminated_length": 481.75,
"completions/min_length": 434.0,
"completions/min_terminated_length": 434.0,
"epoch": 0.01741602985605118,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5350953340530396,
"kl": 1.3473403006792068,
"learning_rate": 4.785714285714287e-06,
"loss": 0.0013,
"num_tokens": 292441.0,
"reward": 0.41716164350509644,
"reward_std": 0.06578143686056137,
"rewards/check_answer/mean": 0.41716164350509644,
"rewards/check_answer/std": 0.06578142195940018,
"step": 98
},
{
"completion_length": 493.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 555.0,
"completions/max_terminated_length": 555.0,
"completions/mean_length": 493.25,
"completions/mean_terminated_length": 493.25,
"completions/min_length": 416.0,
"completions/min_terminated_length": 416.0,
"epoch": 0.017593744446419052,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5714409351348877,
"kl": 1.004945993423462,
"learning_rate": 4.777777777777778e-06,
"loss": 0.001,
"num_tokens": 295578.0,
"reward": 0.6212315559387207,
"reward_std": 0.050040777772665024,
"rewards/check_answer/mean": 0.6212315559387207,
"rewards/check_answer/std": 0.05004078894853592,
"step": 99
},
{
"completion_length": 310.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 426.0,
"completions/max_terminated_length": 426.0,
"completions/mean_length": 310.75,
"completions/mean_terminated_length": 310.75,
"completions/min_length": 232.0,
"completions/min_terminated_length": 232.0,
"epoch": 0.01777145903678692,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5875662565231323,
"kl": 1.0317108184099197,
"learning_rate": 4.769841269841271e-06,
"loss": 0.001,
"num_tokens": 297969.0,
"reward": 0.5778508186340332,
"reward_std": 0.10223124176263809,
"rewards/check_answer/mean": 0.5778508186340332,
"rewards/check_answer/std": 0.10223124176263809,
"step": 100
},
{
"completion_length": 262.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 320.0,
"completions/max_terminated_length": 320.0,
"completions/mean_length": 262.25,
"completions/mean_terminated_length": 262.25,
"completions/min_length": 208.0,
"completions/min_terminated_length": 208.0,
"epoch": 0.01794917362715479,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.406073570251465,
"kl": 1.4292311072349548,
"learning_rate": 4.761904761904762e-06,
"loss": 0.0014,
"num_tokens": 300114.0,
"reward": 0.8254406452178955,
"reward_std": 0.555376410484314,
"rewards/check_answer/mean": 0.8254406452178955,
"rewards/check_answer/std": 0.555376410484314,
"step": 101
},
{
"completion_length": 281.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 407.0,
"completions/max_terminated_length": 407.0,
"completions/mean_length": 281.0,
"completions/mean_terminated_length": 281.0,
"completions/min_length": 218.0,
"completions/min_terminated_length": 218.0,
"epoch": 0.01812688821752266,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7285429239273071,
"kl": 1.1764685809612274,
"learning_rate": 4.753968253968254e-06,
"loss": 0.0012,
"num_tokens": 302354.0,
"reward": 0.9168250560760498,
"reward_std": 0.23327459394931793,
"rewards/check_answer/mean": 0.9168250560760498,
"rewards/check_answer/std": 0.23327459394931793,
"step": 102
},
{
"completion_length": 454.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 471.0,
"completions/max_terminated_length": 471.0,
"completions/mean_length": 454.25,
"completions/mean_terminated_length": 454.25,
"completions/min_length": 434.0,
"completions/min_terminated_length": 434.0,
"epoch": 0.01830460280789053,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7941714525222778,
"kl": 1.0800625532865524,
"learning_rate": 4.7460317460317465e-06,
"loss": 0.0011,
"num_tokens": 305319.0,
"reward": 0.644266664981842,
"reward_std": 0.14616478979587555,
"rewards/check_answer/mean": 0.644266664981842,
"rewards/check_answer/std": 0.14616480469703674,
"step": 103
},
{
"completion_length": 509.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 557.0,
"completions/max_terminated_length": 557.0,
"completions/mean_length": 509.75,
"completions/mean_terminated_length": 509.75,
"completions/min_length": 437.0,
"completions/min_terminated_length": 437.0,
"epoch": 0.018482317398258397,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3640716075897217,
"kl": 0.9189227521419525,
"learning_rate": 4.738095238095238e-06,
"loss": 0.0009,
"num_tokens": 308550.0,
"reward": 0.8741264939308167,
"reward_std": 0.15295472741127014,
"rewards/check_answer/mean": 0.8741264939308167,
"rewards/check_answer/std": 0.15295472741127014,
"step": 104
},
{
"completion_length": 343.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 467.0,
"completions/max_terminated_length": 467.0,
"completions/mean_length": 343.0,
"completions/mean_terminated_length": 343.0,
"completions/min_length": 265.0,
"completions/min_terminated_length": 265.0,
"epoch": 0.018660031988626265,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.171943187713623,
"kl": 1.0402073562145233,
"learning_rate": 4.730158730158731e-06,
"loss": 0.001,
"num_tokens": 311062.0,
"reward": 0.9081037044525146,
"reward_std": 0.18349121510982513,
"rewards/check_answer/mean": 0.9081037044525146,
"rewards/check_answer/std": 0.18349123001098633,
"step": 105
},
{
"completion_length": 411.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 511.0,
"completions/max_terminated_length": 511.0,
"completions/mean_length": 411.25,
"completions/mean_terminated_length": 411.25,
"completions/min_length": 228.0,
"completions/min_terminated_length": 228.0,
"epoch": 0.018837746578994136,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5505229234695435,
"kl": 0.8422607332468033,
"learning_rate": 4.722222222222222e-06,
"loss": 0.0008,
"num_tokens": 313911.0,
"reward": 0.8211601972579956,
"reward_std": 0.10855837911367416,
"rewards/check_answer/mean": 0.8211601972579956,
"rewards/check_answer/std": 0.10855837911367416,
"step": 106
},
{
"completion_length": 345.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 484.0,
"completions/max_terminated_length": 484.0,
"completions/mean_length": 345.75,
"completions/mean_terminated_length": 345.75,
"completions/min_length": 278.0,
"completions/min_terminated_length": 278.0,
"epoch": 0.019015461169362004,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7689751386642456,
"kl": 1.3255447894334793,
"learning_rate": 4.714285714285715e-06,
"loss": 0.0013,
"num_tokens": 316422.0,
"reward": 0.3924521505832672,
"reward_std": 0.11489391326904297,
"rewards/check_answer/mean": 0.3924521505832672,
"rewards/check_answer/std": 0.11489391326904297,
"step": 107
},
{
"completion_length": 577.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 633.0,
"completions/max_terminated_length": 633.0,
"completions/mean_length": 577.0,
"completions/mean_terminated_length": 577.0,
"completions/min_length": 520.0,
"completions/min_terminated_length": 520.0,
"epoch": 0.019193175759729873,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4045451879501343,
"kl": 0.948733925819397,
"learning_rate": 4.706349206349206e-06,
"loss": 0.0009,
"num_tokens": 319942.0,
"reward": 0.7583677768707275,
"reward_std": 0.09759147465229034,
"rewards/check_answer/mean": 0.7583677768707275,
"rewards/check_answer/std": 0.09759148210287094,
"step": 108
},
{
"completion_length": 488.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 522.0,
"completions/max_terminated_length": 522.0,
"completions/mean_length": 488.0,
"completions/mean_terminated_length": 488.0,
"completions/min_length": 408.0,
"completions/min_terminated_length": 408.0,
"epoch": 0.019370890350097744,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5084129571914673,
"kl": 0.9847242683172226,
"learning_rate": 4.698412698412699e-06,
"loss": 0.001,
"num_tokens": 323078.0,
"reward": 0.5234397649765015,
"reward_std": 0.09028327465057373,
"rewards/check_answer/mean": 0.5234397649765015,
"rewards/check_answer/std": 0.09028327465057373,
"step": 109
},
{
"completion_length": 471.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 501.0,
"completions/max_terminated_length": 501.0,
"completions/mean_length": 471.75,
"completions/mean_terminated_length": 471.75,
"completions/min_length": 422.0,
"completions/min_terminated_length": 422.0,
"epoch": 0.019548604940465612,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8055229187011719,
"kl": 1.0391763895750046,
"learning_rate": 4.6904761904761905e-06,
"loss": 0.001,
"num_tokens": 326121.0,
"reward": 0.2687339186668396,
"reward_std": 0.06397005170583725,
"rewards/check_answer/mean": 0.2687339186668396,
"rewards/check_answer/std": 0.06397005915641785,
"step": 110
},
{
"completion_length": 353.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 481.0,
"completions/max_terminated_length": 481.0,
"completions/mean_length": 353.5,
"completions/mean_terminated_length": 353.5,
"completions/min_length": 287.0,
"completions/min_terminated_length": 287.0,
"epoch": 0.01972631953083348,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7900550365447998,
"kl": 1.1994563341140747,
"learning_rate": 4.682539682539683e-06,
"loss": 0.0012,
"num_tokens": 328703.0,
"reward": 0.7745100259780884,
"reward_std": 0.07118824124336243,
"rewards/check_answer/mean": 0.7745100259780884,
"rewards/check_answer/std": 0.07118824124336243,
"step": 111
},
{
"completion_length": 314.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 395.0,
"completions/max_terminated_length": 395.0,
"completions/mean_length": 314.0,
"completions/mean_terminated_length": 314.0,
"completions/min_length": 221.0,
"completions/min_terminated_length": 221.0,
"epoch": 0.019904034121201352,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7322038412094116,
"kl": 1.1014239341020584,
"learning_rate": 4.674603174603175e-06,
"loss": 0.0011,
"num_tokens": 331067.0,
"reward": 0.712478518486023,
"reward_std": 0.2585718333721161,
"rewards/check_answer/mean": 0.712478518486023,
"rewards/check_answer/std": 0.2585718333721161,
"step": 112
},
{
"completion_length": 440.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 488.0,
"completions/max_terminated_length": 488.0,
"completions/mean_length": 440.75,
"completions/mean_terminated_length": 440.75,
"completions/min_length": 353.0,
"completions/min_terminated_length": 353.0,
"epoch": 0.02008174871156922,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.374124526977539,
"kl": 1.009152129292488,
"learning_rate": 4.666666666666667e-06,
"loss": 0.001,
"num_tokens": 334014.0,
"reward": 0.6918644905090332,
"reward_std": 0.28659263253211975,
"rewards/check_answer/mean": 0.6918644905090332,
"rewards/check_answer/std": 0.28659263253211975,
"step": 113
},
{
"completion_length": 325.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 361.0,
"completions/max_terminated_length": 361.0,
"completions/mean_length": 325.25,
"completions/mean_terminated_length": 325.25,
"completions/min_length": 298.0,
"completions/min_terminated_length": 298.0,
"epoch": 0.02025946330193709,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.688226580619812,
"kl": 0.8765446245670319,
"learning_rate": 4.658730158730159e-06,
"loss": 0.0009,
"num_tokens": 336467.0,
"reward": 0.7640917301177979,
"reward_std": 0.07848616689443588,
"rewards/check_answer/mean": 0.7640917301177979,
"rewards/check_answer/std": 0.07848618924617767,
"step": 114
},
{
"completion_length": 225.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 252.0,
"completions/max_terminated_length": 252.0,
"completions/mean_length": 225.75,
"completions/mean_terminated_length": 225.75,
"completions/min_length": 211.0,
"completions/min_terminated_length": 211.0,
"epoch": 0.02043717789230496,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8820006847381592,
"kl": 1.1795692443847656,
"learning_rate": 4.650793650793651e-06,
"loss": 0.0012,
"num_tokens": 338490.0,
"reward": 1.0758620500564575,
"reward_std": 0.014311790466308594,
"rewards/check_answer/mean": 1.0758620500564575,
"rewards/check_answer/std": 0.014311791397631168,
"step": 115
},
{
"completion_length": 326.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 360.0,
"completions/max_terminated_length": 360.0,
"completions/mean_length": 326.5,
"completions/mean_terminated_length": 326.5,
"completions/min_length": 247.0,
"completions/min_terminated_length": 247.0,
"epoch": 0.020614892482672828,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.018541097640991,
"kl": 1.2944844365119934,
"learning_rate": 4.642857142857144e-06,
"loss": 0.0013,
"num_tokens": 340904.0,
"reward": 0.8421782851219177,
"reward_std": 0.5801041722297668,
"rewards/check_answer/mean": 0.8421782851219177,
"rewards/check_answer/std": 0.5801041722297668,
"step": 116
},
{
"completion_length": 850.75,
"completions/clipped_ratio": 0.25,
"completions/max_length": 1733.0,
"completions/max_terminated_length": 584.0,
"completions/mean_length": 850.75,
"completions/mean_terminated_length": 556.6666870117188,
"completions/min_length": 539.0,
"completions/min_terminated_length": 539.0,
"epoch": 0.020792607073040696,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.0379717350006104,
"kl": 1.2277846336364746,
"learning_rate": 4.634920634920635e-06,
"loss": 0.0012,
"num_tokens": 345531.0,
"reward": 0.8824595808982849,
"reward_std": 0.3745229244232178,
"rewards/check_answer/mean": 0.8824595808982849,
"rewards/check_answer/std": 0.3745229244232178,
"step": 117
},
{
"completion_length": 429.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 585.0,
"completions/max_terminated_length": 585.0,
"completions/mean_length": 429.75,
"completions/mean_terminated_length": 429.75,
"completions/min_length": 282.0,
"completions/min_terminated_length": 282.0,
"epoch": 0.020970321663408564,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5808711051940918,
"kl": 1.2089886367321014,
"learning_rate": 4.626984126984128e-06,
"loss": 0.0012,
"num_tokens": 348426.0,
"reward": 0.49039697647094727,
"reward_std": 0.13966494798660278,
"rewards/check_answer/mean": 0.49039697647094727,
"rewards/check_answer/std": 0.1396649330854416,
"step": 118
},
{
"completion_length": 332.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 408.0,
"completions/max_terminated_length": 408.0,
"completions/mean_length": 332.5,
"completions/mean_terminated_length": 332.5,
"completions/min_length": 267.0,
"completions/min_terminated_length": 267.0,
"epoch": 0.021148036253776436,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.896164894104004,
"kl": 1.2581031918525696,
"learning_rate": 4.6190476190476196e-06,
"loss": 0.0013,
"num_tokens": 350868.0,
"reward": 0.5684536695480347,
"reward_std": 0.17921964824199677,
"rewards/check_answer/mean": 0.5684536695480347,
"rewards/check_answer/std": 0.17921964824199677,
"step": 119
},
{
"completion_length": 419.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 510.0,
"completions/max_terminated_length": 510.0,
"completions/mean_length": 419.75,
"completions/mean_terminated_length": 419.75,
"completions/min_length": 318.0,
"completions/min_terminated_length": 318.0,
"epoch": 0.021325750844144304,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3449409008026123,
"kl": 0.9334018975496292,
"learning_rate": 4.611111111111112e-06,
"loss": 0.0009,
"num_tokens": 353731.0,
"reward": 0.9145520329475403,
"reward_std": 0.16757537424564362,
"rewards/check_answer/mean": 0.9145520329475403,
"rewards/check_answer/std": 0.16757537424564362,
"step": 120
},
{
"completion_length": 644.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 843.0,
"completions/max_terminated_length": 843.0,
"completions/mean_length": 644.0,
"completions/mean_terminated_length": 644.0,
"completions/min_length": 508.0,
"completions/min_terminated_length": 508.0,
"epoch": 0.021503465434512172,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2333385944366455,
"kl": 1.0253196954727173,
"learning_rate": 4.603174603174604e-06,
"loss": 0.001,
"num_tokens": 357515.0,
"reward": 0.5978646874427795,
"reward_std": 0.23829804360866547,
"rewards/check_answer/mean": 0.5978646874427795,
"rewards/check_answer/std": 0.23829805850982666,
"step": 121
},
{
"completion_length": 589.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 744.0,
"completions/max_terminated_length": 744.0,
"completions/mean_length": 589.25,
"completions/mean_terminated_length": 589.25,
"completions/min_length": 368.0,
"completions/min_terminated_length": 368.0,
"epoch": 0.021681180024880044,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2892918586730957,
"kl": 0.852889820933342,
"learning_rate": 4.595238095238095e-06,
"loss": 0.0009,
"num_tokens": 361076.0,
"reward": 0.32557737827301025,
"reward_std": 0.07417508214712143,
"rewards/check_answer/mean": 0.32557737827301025,
"rewards/check_answer/std": 0.07417508214712143,
"step": 122
},
{
"completion_length": 238.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 379.0,
"completions/max_terminated_length": 379.0,
"completions/mean_length": 238.0,
"completions/mean_terminated_length": 238.0,
"completions/min_length": 166.0,
"completions/min_terminated_length": 166.0,
"epoch": 0.021858894615247912,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9720977544784546,
"kl": 1.2939183413982391,
"learning_rate": 4.587301587301588e-06,
"loss": 0.0013,
"num_tokens": 363136.0,
"reward": 1.2107616662979126,
"reward_std": 0.3746259808540344,
"rewards/check_answer/mean": 1.2107616662979126,
"rewards/check_answer/std": 0.3746259808540344,
"step": 123
},
{
"completion_length": 520.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 610.0,
"completions/max_terminated_length": 610.0,
"completions/mean_length": 520.5,
"completions/mean_terminated_length": 520.5,
"completions/min_length": 442.0,
"completions/min_terminated_length": 442.0,
"epoch": 0.02203660920561578,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2540651559829712,
"kl": 0.7220829427242279,
"learning_rate": 4.5793650793650795e-06,
"loss": 0.0007,
"num_tokens": 366438.0,
"reward": 0.4332646131515503,
"reward_std": 0.06124412640929222,
"rewards/check_answer/mean": 0.4332646131515503,
"rewards/check_answer/std": 0.06124412640929222,
"step": 124
},
{
"completion_length": 424.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 611.0,
"completions/max_terminated_length": 611.0,
"completions/mean_length": 424.25,
"completions/mean_terminated_length": 424.25,
"completions/min_length": 299.0,
"completions/min_terminated_length": 299.0,
"epoch": 0.02221432379598365,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7561167478561401,
"kl": 0.9426107108592987,
"learning_rate": 4.571428571428572e-06,
"loss": 0.0009,
"num_tokens": 369263.0,
"reward": 0.47949886322021484,
"reward_std": 0.014681631699204445,
"rewards/check_answer/mean": 0.47949886322021484,
"rewards/check_answer/std": 0.014681617729365826,
"step": 125
},
{
"completion_length": 447.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 513.0,
"completions/max_terminated_length": 513.0,
"completions/mean_length": 447.25,
"completions/mean_terminated_length": 447.25,
"completions/min_length": 354.0,
"completions/min_terminated_length": 354.0,
"epoch": 0.02239203838635152,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5910497903823853,
"kl": 0.9867564290761948,
"learning_rate": 4.563492063492064e-06,
"loss": 0.001,
"num_tokens": 372228.0,
"reward": 0.41330093145370483,
"reward_std": 0.061453063040971756,
"rewards/check_answer/mean": 0.41330093145370483,
"rewards/check_answer/std": 0.061453066766262054,
"step": 126
},
{
"completion_length": 385.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 450.0,
"completions/max_terminated_length": 450.0,
"completions/mean_length": 385.5,
"completions/mean_terminated_length": 385.5,
"completions/min_length": 241.0,
"completions/min_terminated_length": 241.0,
"epoch": 0.022569752976719388,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7781590223312378,
"kl": 1.2027154564857483,
"learning_rate": 4.555555555555556e-06,
"loss": 0.0012,
"num_tokens": 374906.0,
"reward": 1.0481634140014648,
"reward_std": 0.38712385296821594,
"rewards/check_answer/mean": 1.0481634140014648,
"rewards/check_answer/std": 0.38712385296821594,
"step": 127
},
{
"completion_length": 479.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 570.0,
"completions/max_terminated_length": 570.0,
"completions/mean_length": 479.75,
"completions/mean_terminated_length": 479.75,
"completions/min_length": 417.0,
"completions/min_terminated_length": 417.0,
"epoch": 0.02274746756708726,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4399559497833252,
"kl": 1.1581873446702957,
"learning_rate": 4.547619047619048e-06,
"loss": 0.0012,
"num_tokens": 378005.0,
"reward": 0.5218023061752319,
"reward_std": 0.10244939476251602,
"rewards/check_answer/mean": 0.5218023061752319,
"rewards/check_answer/std": 0.10244939476251602,
"step": 128
},
{
"completion_length": 314.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 376.0,
"completions/max_terminated_length": 376.0,
"completions/mean_length": 314.0,
"completions/mean_terminated_length": 314.0,
"completions/min_length": 214.0,
"completions/min_terminated_length": 214.0,
"epoch": 0.022925182157455128,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.065781831741333,
"kl": 1.0792067795991898,
"learning_rate": 4.53968253968254e-06,
"loss": 0.0011,
"num_tokens": 380365.0,
"reward": 0.14184531569480896,
"reward_std": 0.11691629886627197,
"rewards/check_answer/mean": 0.14184531569480896,
"rewards/check_answer/std": 0.11691630631685257,
"step": 129
},
{
"completion_length": 394.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 459.0,
"completions/max_terminated_length": 459.0,
"completions/mean_length": 394.5,
"completions/mean_terminated_length": 394.5,
"completions/min_length": 305.0,
"completions/min_terminated_length": 305.0,
"epoch": 0.023102896747822996,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7899744510650635,
"kl": 1.202794149518013,
"learning_rate": 4.531746031746032e-06,
"loss": 0.0012,
"num_tokens": 383103.0,
"reward": 0.9240409731864929,
"reward_std": 0.19024822115898132,
"rewards/check_answer/mean": 0.9240409731864929,
"rewards/check_answer/std": 0.19024820625782013,
"step": 130
},
{
"completion_length": 525.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 636.0,
"completions/max_terminated_length": 636.0,
"completions/mean_length": 525.75,
"completions/mean_terminated_length": 525.75,
"completions/min_length": 425.0,
"completions/min_terminated_length": 425.0,
"epoch": 0.023280611338190864,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5444499254226685,
"kl": 1.0851371437311172,
"learning_rate": 4.523809523809524e-06,
"loss": 0.0011,
"num_tokens": 386370.0,
"reward": 0.5510776042938232,
"reward_std": 0.14302389323711395,
"rewards/check_answer/mean": 0.5510776042938232,
"rewards/check_answer/std": 0.14302392303943634,
"step": 131
},
{
"completion_length": 391.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 467.0,
"completions/max_terminated_length": 467.0,
"completions/mean_length": 391.5,
"completions/mean_terminated_length": 391.5,
"completions/min_length": 324.0,
"completions/min_terminated_length": 324.0,
"epoch": 0.023458325928558735,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.977134346961975,
"kl": 1.2055960595607758,
"learning_rate": 4.515873015873016e-06,
"loss": 0.0012,
"num_tokens": 389068.0,
"reward": 0.630324125289917,
"reward_std": 0.14643849432468414,
"rewards/check_answer/mean": 0.630324125289917,
"rewards/check_answer/std": 0.14643850922584534,
"step": 132
},
{
"completion_length": 365.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 426.0,
"completions/max_terminated_length": 426.0,
"completions/mean_length": 365.75,
"completions/mean_terminated_length": 365.75,
"completions/min_length": 282.0,
"completions/min_terminated_length": 282.0,
"epoch": 0.023636040518926604,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7458113431930542,
"kl": 1.1333343386650085,
"learning_rate": 4.5079365079365085e-06,
"loss": 0.0011,
"num_tokens": 391643.0,
"reward": 0.3864520192146301,
"reward_std": 0.10917352885007858,
"rewards/check_answer/mean": 0.3864520192146301,
"rewards/check_answer/std": 0.10917352885007858,
"step": 133
},
{
"completion_length": 313.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 367.0,
"completions/max_terminated_length": 367.0,
"completions/mean_length": 313.75,
"completions/mean_terminated_length": 313.75,
"completions/min_length": 215.0,
"completions/min_terminated_length": 215.0,
"epoch": 0.02381375510929447,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8480489253997803,
"kl": 1.1135710626840591,
"learning_rate": 4.5e-06,
"loss": 0.0011,
"num_tokens": 394014.0,
"reward": 0.4650374948978424,
"reward_std": 0.16275101900100708,
"rewards/check_answer/mean": 0.4650374948978424,
"rewards/check_answer/std": 0.16275101900100708,
"step": 134
},
{
"completion_length": 202.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 253.0,
"completions/max_terminated_length": 253.0,
"completions/mean_length": 202.5,
"completions/mean_terminated_length": 202.5,
"completions/min_length": 172.0,
"completions/min_terminated_length": 172.0,
"epoch": 0.023991469699662343,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.3134193420410156,
"kl": 1.4941602051258087,
"learning_rate": 4.492063492063493e-06,
"loss": 0.0015,
"num_tokens": 395924.0,
"reward": 0.2674972414970398,
"reward_std": 0.05775442719459534,
"rewards/check_answer/mean": 0.2674972414970398,
"rewards/check_answer/std": 0.05775442719459534,
"step": 135
},
{
"completion_length": 546.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 603.0,
"completions/max_terminated_length": 603.0,
"completions/mean_length": 546.5,
"completions/mean_terminated_length": 546.5,
"completions/min_length": 477.0,
"completions/min_terminated_length": 477.0,
"epoch": 0.02416918429003021,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2359838485717773,
"kl": 0.8067422658205032,
"learning_rate": 4.484126984126984e-06,
"loss": 0.0008,
"num_tokens": 399342.0,
"reward": 0.5291463136672974,
"reward_std": 0.05505036190152168,
"rewards/check_answer/mean": 0.5291463136672974,
"rewards/check_answer/std": 0.05505036562681198,
"step": 136
},
{
"completion_length": 515.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 612.0,
"completions/max_terminated_length": 612.0,
"completions/mean_length": 515.5,
"completions/mean_terminated_length": 515.5,
"completions/min_length": 431.0,
"completions/min_terminated_length": 431.0,
"epoch": 0.02434689888039808,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.452023983001709,
"kl": 1.1316813081502914,
"learning_rate": 4.476190476190477e-06,
"loss": 0.0011,
"num_tokens": 402580.0,
"reward": 0.7063450813293457,
"reward_std": 0.05188734456896782,
"rewards/check_answer/mean": 0.7063450813293457,
"rewards/check_answer/std": 0.05188734456896782,
"step": 137
},
{
"completion_length": 164.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 180.0,
"completions/max_terminated_length": 180.0,
"completions/mean_length": 164.25,
"completions/mean_terminated_length": 164.25,
"completions/min_length": 148.0,
"completions/min_terminated_length": 148.0,
"epoch": 0.02452461347076595,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.583338975906372,
"kl": 1.4289655685424805,
"learning_rate": 4.468253968253969e-06,
"loss": 0.0014,
"num_tokens": 404341.0,
"reward": 0.7204106450080872,
"reward_std": 0.08577322959899902,
"rewards/check_answer/mean": 0.7204106450080872,
"rewards/check_answer/std": 0.08577322959899902,
"step": 138
},
{
"completion_length": 324.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 381.0,
"completions/max_terminated_length": 381.0,
"completions/mean_length": 324.0,
"completions/mean_terminated_length": 324.0,
"completions/min_length": 295.0,
"completions/min_terminated_length": 295.0,
"epoch": 0.02470232806113382,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7330161333084106,
"kl": 0.9842696934938431,
"learning_rate": 4.460317460317461e-06,
"loss": 0.001,
"num_tokens": 406773.0,
"reward": 0.48003125190734863,
"reward_std": 0.07409851998090744,
"rewards/check_answer/mean": 0.48003125190734863,
"rewards/check_answer/std": 0.07409852743148804,
"step": 139
},
{
"completion_length": 252.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 284.0,
"completions/max_terminated_length": 284.0,
"completions/mean_length": 252.0,
"completions/mean_terminated_length": 252.0,
"completions/min_length": 207.0,
"completions/min_terminated_length": 207.0,
"epoch": 0.024880042651501687,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.1670877933502197,
"kl": 1.2738738059997559,
"learning_rate": 4.4523809523809525e-06,
"loss": 0.0013,
"num_tokens": 408881.0,
"reward": 0.5589589476585388,
"reward_std": 0.3076104521751404,
"rewards/check_answer/mean": 0.5589589476585388,
"rewards/check_answer/std": 0.30761048197746277,
"step": 140
},
{
"completion_length": 485.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 536.0,
"completions/max_terminated_length": 536.0,
"completions/mean_length": 485.75,
"completions/mean_terminated_length": 485.75,
"completions/min_length": 424.0,
"completions/min_terminated_length": 424.0,
"epoch": 0.02505775724186956,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4860774278640747,
"kl": 1.0862714499235153,
"learning_rate": 4.444444444444444e-06,
"loss": 0.0011,
"num_tokens": 412028.0,
"reward": 1.1310088634490967,
"reward_std": 0.0607450008392334,
"rewards/check_answer/mean": 1.1310088634490967,
"rewards/check_answer/std": 0.0607449971139431,
"step": 141
},
{
"completion_length": 357.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 442.0,
"completions/max_terminated_length": 442.0,
"completions/mean_length": 357.5,
"completions/mean_terminated_length": 357.5,
"completions/min_length": 279.0,
"completions/min_terminated_length": 279.0,
"epoch": 0.025235471832237427,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.857495665550232,
"kl": 0.9955440014600754,
"learning_rate": 4.436507936507937e-06,
"loss": 0.001,
"num_tokens": 414582.0,
"reward": 0.371147096157074,
"reward_std": 0.03809575363993645,
"rewards/check_answer/mean": 0.371147096157074,
"rewards/check_answer/std": 0.03809575363993645,
"step": 142
},
{
"completion_length": 512.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 556.0,
"completions/max_terminated_length": 556.0,
"completions/mean_length": 512.5,
"completions/mean_terminated_length": 512.5,
"completions/min_length": 478.0,
"completions/min_terminated_length": 478.0,
"epoch": 0.025413186422605295,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3737552165985107,
"kl": 0.8718635737895966,
"learning_rate": 4.428571428571429e-06,
"loss": 0.0009,
"num_tokens": 417828.0,
"reward": 0.7945088148117065,
"reward_std": 0.10661814361810684,
"rewards/check_answer/mean": 0.7945088148117065,
"rewards/check_answer/std": 0.10661814361810684,
"step": 143
},
{
"completion_length": 400.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 514.0,
"completions/max_terminated_length": 514.0,
"completions/mean_length": 400.5,
"completions/mean_terminated_length": 400.5,
"completions/min_length": 279.0,
"completions/min_terminated_length": 279.0,
"epoch": 0.025590901012973163,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7630726099014282,
"kl": 1.0710228085517883,
"learning_rate": 4.420634920634921e-06,
"loss": 0.0011,
"num_tokens": 420578.0,
"reward": 0.47816359996795654,
"reward_std": 0.0742780864238739,
"rewards/check_answer/mean": 0.47816359996795654,
"rewards/check_answer/std": 0.0742780864238739,
"step": 144
},
{
"completion_length": 344.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 401.0,
"completions/max_terminated_length": 401.0,
"completions/mean_length": 344.0,
"completions/mean_terminated_length": 344.0,
"completions/min_length": 289.0,
"completions/min_terminated_length": 289.0,
"epoch": 0.025768615603341035,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.808454155921936,
"kl": 1.1346302181482315,
"learning_rate": 4.412698412698413e-06,
"loss": 0.0011,
"num_tokens": 423110.0,
"reward": 1.2456703186035156,
"reward_std": 0.18238836526870728,
"rewards/check_answer/mean": 1.2456703186035156,
"rewards/check_answer/std": 0.18238838016986847,
"step": 145
},
{
"completion_length": 563.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 614.0,
"completions/max_terminated_length": 614.0,
"completions/mean_length": 563.25,
"completions/mean_terminated_length": 563.25,
"completions/min_length": 447.0,
"completions/min_terminated_length": 447.0,
"epoch": 0.025946330193708903,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4395103454589844,
"kl": 0.9940735101699829,
"learning_rate": 4.404761904761905e-06,
"loss": 0.001,
"num_tokens": 426535.0,
"reward": 0.4605726897716522,
"reward_std": 0.018929272890090942,
"rewards/check_answer/mean": 0.4605726897716522,
"rewards/check_answer/std": 0.018929272890090942,
"step": 146
},
{
"completion_length": 404.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 463.0,
"completions/max_terminated_length": 463.0,
"completions/mean_length": 404.75,
"completions/mean_terminated_length": 404.75,
"completions/min_length": 320.0,
"completions/min_terminated_length": 320.0,
"epoch": 0.02612404478407677,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.565238118171692,
"kl": 0.8450873494148254,
"learning_rate": 4.396825396825397e-06,
"loss": 0.0008,
"num_tokens": 429318.0,
"reward": 0.655875027179718,
"reward_std": 0.012029829435050488,
"rewards/check_answer/mean": 0.655875027179718,
"rewards/check_answer/std": 0.012029829435050488,
"step": 147
},
{
"completion_length": 401.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 442.0,
"completions/max_terminated_length": 442.0,
"completions/mean_length": 401.25,
"completions/mean_terminated_length": 401.25,
"completions/min_length": 358.0,
"completions/min_terminated_length": 358.0,
"epoch": 0.026301759374444643,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7787821292877197,
"kl": 1.0799409300088882,
"learning_rate": 4.388888888888889e-06,
"loss": 0.0011,
"num_tokens": 432043.0,
"reward": 0.3119197487831116,
"reward_std": 0.10698849707841873,
"rewards/check_answer/mean": 0.3119197487831116,
"rewards/check_answer/std": 0.10698850452899933,
"step": 148
},
{
"completion_length": 631.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 713.0,
"completions/max_terminated_length": 713.0,
"completions/mean_length": 631.0,
"completions/mean_terminated_length": 631.0,
"completions/min_length": 548.0,
"completions/min_terminated_length": 548.0,
"epoch": 0.02647947396481251,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.177681565284729,
"kl": 1.2260465174913406,
"learning_rate": 4.3809523809523815e-06,
"loss": 0.0012,
"num_tokens": 435819.0,
"reward": 0.6100305318832397,
"reward_std": 0.04862495884299278,
"rewards/check_answer/mean": 0.6100305318832397,
"rewards/check_answer/std": 0.048624955117702484,
"step": 149
},
{
"completion_length": 546.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 620.0,
"completions/max_terminated_length": 620.0,
"completions/mean_length": 546.5,
"completions/mean_terminated_length": 546.5,
"completions/min_length": 404.0,
"completions/min_terminated_length": 404.0,
"epoch": 0.02665718855518038,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6101106405258179,
"kl": 0.8679392486810684,
"learning_rate": 4.373015873015873e-06,
"loss": 0.0009,
"num_tokens": 439197.0,
"reward": 0.461988627910614,
"reward_std": 0.08495597541332245,
"rewards/check_answer/mean": 0.461988627910614,
"rewards/check_answer/std": 0.08495597541332245,
"step": 150
},
{
"completion_length": 471.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 567.0,
"completions/max_terminated_length": 567.0,
"completions/mean_length": 471.5,
"completions/mean_terminated_length": 471.5,
"completions/min_length": 432.0,
"completions/min_terminated_length": 432.0,
"epoch": 0.02683490314554825,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6391887664794922,
"kl": 0.7990295141935349,
"learning_rate": 4.365079365079366e-06,
"loss": 0.0008,
"num_tokens": 442279.0,
"reward": 0.4876329302787781,
"reward_std": 0.05170879140496254,
"rewards/check_answer/mean": 0.4876329302787781,
"rewards/check_answer/std": 0.05170879140496254,
"step": 151
},
{
"completion_length": 340.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 390.0,
"completions/max_terminated_length": 390.0,
"completions/mean_length": 340.0,
"completions/mean_terminated_length": 340.0,
"completions/min_length": 277.0,
"completions/min_terminated_length": 277.0,
"epoch": 0.02701261773591612,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9967378377914429,
"kl": 1.0871484875679016,
"learning_rate": 4.357142857142857e-06,
"loss": 0.0011,
"num_tokens": 444783.0,
"reward": 0.7888412475585938,
"reward_std": 0.09264621883630753,
"rewards/check_answer/mean": 0.7888412475585938,
"rewards/check_answer/std": 0.09264620393514633,
"step": 152
},
{
"completion_length": 443.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 535.0,
"completions/max_terminated_length": 535.0,
"completions/mean_length": 443.0,
"completions/mean_terminated_length": 443.0,
"completions/min_length": 377.0,
"completions/min_terminated_length": 377.0,
"epoch": 0.027190332326283987,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5320669412612915,
"kl": 1.216535747051239,
"learning_rate": 4.34920634920635e-06,
"loss": 0.0012,
"num_tokens": 447747.0,
"reward": 0.4390728175640106,
"reward_std": 0.09861871600151062,
"rewards/check_answer/mean": 0.4390728175640106,
"rewards/check_answer/std": 0.09861870110034943,
"step": 153
},
{
"completion_length": 496.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 552.0,
"completions/max_terminated_length": 552.0,
"completions/mean_length": 496.5,
"completions/mean_terminated_length": 496.5,
"completions/min_length": 459.0,
"completions/min_terminated_length": 459.0,
"epoch": 0.02736804691665186,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5672584772109985,
"kl": 0.9466253966093063,
"learning_rate": 4.3412698412698414e-06,
"loss": 0.0009,
"num_tokens": 450913.0,
"reward": 0.4815061390399933,
"reward_std": 0.04056670889258385,
"rewards/check_answer/mean": 0.4815061390399933,
"rewards/check_answer/std": 0.04056670516729355,
"step": 154
},
{
"completion_length": 432.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 448.0,
"completions/max_terminated_length": 448.0,
"completions/mean_length": 432.25,
"completions/mean_terminated_length": 432.25,
"completions/min_length": 400.0,
"completions/min_terminated_length": 400.0,
"epoch": 0.027545761507019727,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9660674333572388,
"kl": 0.9921428710222244,
"learning_rate": 4.333333333333334e-06,
"loss": 0.001,
"num_tokens": 453814.0,
"reward": 0.32661956548690796,
"reward_std": 0.08909963071346283,
"rewards/check_answer/mean": 0.32661956548690796,
"rewards/check_answer/std": 0.08909963071346283,
"step": 155
},
{
"completion_length": 394.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 474.0,
"completions/max_terminated_length": 474.0,
"completions/mean_length": 394.25,
"completions/mean_terminated_length": 394.25,
"completions/min_length": 334.0,
"completions/min_terminated_length": 334.0,
"epoch": 0.027723476097387595,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7710676193237305,
"kl": 1.1431185454130173,
"learning_rate": 4.3253968253968256e-06,
"loss": 0.0011,
"num_tokens": 456531.0,
"reward": 0.5441449880599976,
"reward_std": 0.27911967039108276,
"rewards/check_answer/mean": 0.5441449880599976,
"rewards/check_answer/std": 0.27911967039108276,
"step": 156
},
{
"completion_length": 351.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 457.0,
"completions/max_terminated_length": 457.0,
"completions/mean_length": 351.75,
"completions/mean_terminated_length": 351.75,
"completions/min_length": 224.0,
"completions/min_terminated_length": 224.0,
"epoch": 0.027901190687755463,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7665750980377197,
"kl": 0.970624715089798,
"learning_rate": 4.317460317460318e-06,
"loss": 0.001,
"num_tokens": 459094.0,
"reward": 0.7966454029083252,
"reward_std": 0.2309737652540207,
"rewards/check_answer/mean": 0.7966454029083252,
"rewards/check_answer/std": 0.2309737652540207,
"step": 157
},
{
"completion_length": 237.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 340.0,
"completions/max_terminated_length": 340.0,
"completions/mean_length": 237.0,
"completions/mean_terminated_length": 237.0,
"completions/min_length": 183.0,
"completions/min_terminated_length": 183.0,
"epoch": 0.028078905278123335,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.2502260208129883,
"kl": 1.3823933005332947,
"learning_rate": 4.30952380952381e-06,
"loss": 0.0014,
"num_tokens": 461150.0,
"reward": 1.2577112913131714,
"reward_std": 0.3310515582561493,
"rewards/check_answer/mean": 1.2577112913131714,
"rewards/check_answer/std": 0.3310515582561493,
"step": 158
},
{
"completion_length": 435.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 502.0,
"completions/max_terminated_length": 502.0,
"completions/mean_length": 435.25,
"completions/mean_terminated_length": 435.25,
"completions/min_length": 334.0,
"completions/min_terminated_length": 334.0,
"epoch": 0.028256619868491203,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5836832523345947,
"kl": 1.281766653060913,
"learning_rate": 4.301587301587302e-06,
"loss": 0.0013,
"num_tokens": 464043.0,
"reward": 0.7175251841545105,
"reward_std": 0.11627113074064255,
"rewards/check_answer/mean": 0.7175251841545105,
"rewards/check_answer/std": 0.11627113074064255,
"step": 159
},
{
"completion_length": 478.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 559.0,
"completions/max_terminated_length": 559.0,
"completions/mean_length": 478.25,
"completions/mean_terminated_length": 478.25,
"completions/min_length": 378.0,
"completions/min_terminated_length": 378.0,
"epoch": 0.02843433445885907,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5781432390213013,
"kl": 2.0126918852329254,
"learning_rate": 4.293650793650794e-06,
"loss": 0.002,
"num_tokens": 467112.0,
"reward": 1.0377893447875977,
"reward_std": 0.17348872125148773,
"rewards/check_answer/mean": 1.0377893447875977,
"rewards/check_answer/std": 0.17348873615264893,
"step": 160
},
{
"completion_length": 353.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 534.0,
"completions/max_terminated_length": 534.0,
"completions/mean_length": 353.0,
"completions/mean_terminated_length": 353.0,
"completions/min_length": 245.0,
"completions/min_terminated_length": 245.0,
"epoch": 0.028612049049226942,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.5405149459838867,
"kl": 1.1099863350391388,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.0011,
"num_tokens": 469672.0,
"reward": 0.7634687423706055,
"reward_std": 0.16830436885356903,
"rewards/check_answer/mean": 0.7634687423706055,
"rewards/check_answer/std": 0.16830436885356903,
"step": 161
},
{
"completion_length": 463.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 582.0,
"completions/max_terminated_length": 582.0,
"completions/mean_length": 463.25,
"completions/mean_terminated_length": 463.25,
"completions/min_length": 376.0,
"completions/min_terminated_length": 376.0,
"epoch": 0.02878976363959481,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3183034658432007,
"kl": 1.1439854502677917,
"learning_rate": 4.277777777777778e-06,
"loss": 0.0011,
"num_tokens": 472685.0,
"reward": 0.5173400640487671,
"reward_std": 0.17256943881511688,
"rewards/check_answer/mean": 0.5173400640487671,
"rewards/check_answer/std": 0.1725694090127945,
"step": 162
},
{
"completion_length": 531.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 546.0,
"completions/max_terminated_length": 546.0,
"completions/mean_length": 531.5,
"completions/mean_terminated_length": 531.5,
"completions/min_length": 511.0,
"completions/min_terminated_length": 511.0,
"epoch": 0.02896747822996268,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.347712755203247,
"kl": 0.7350674420595169,
"learning_rate": 4.26984126984127e-06,
"loss": 0.0007,
"num_tokens": 476047.0,
"reward": 0.6685494184494019,
"reward_std": 0.031588803976774216,
"rewards/check_answer/mean": 0.6685494184494019,
"rewards/check_answer/std": 0.031588807702064514,
"step": 163
},
{
"completion_length": 556.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 631.0,
"completions/max_terminated_length": 631.0,
"completions/mean_length": 556.75,
"completions/mean_terminated_length": 556.75,
"completions/min_length": 470.0,
"completions/min_terminated_length": 470.0,
"epoch": 0.02914519282033055,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4339076280593872,
"kl": 0.8411607146263123,
"learning_rate": 4.261904761904762e-06,
"loss": 0.0008,
"num_tokens": 479498.0,
"reward": 0.6374480724334717,
"reward_std": 0.13143180310726166,
"rewards/check_answer/mean": 0.6374480724334717,
"rewards/check_answer/std": 0.13143180310726166,
"step": 164
},
{
"completion_length": 331.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 445.0,
"completions/max_terminated_length": 445.0,
"completions/mean_length": 331.75,
"completions/mean_terminated_length": 331.75,
"completions/min_length": 241.0,
"completions/min_terminated_length": 241.0,
"epoch": 0.02932290741069842,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8252785205841064,
"kl": 1.183883011341095,
"learning_rate": 4.2539682539682546e-06,
"loss": 0.0012,
"num_tokens": 481929.0,
"reward": 0.7619638442993164,
"reward_std": 0.16213203966617584,
"rewards/check_answer/mean": 0.7619638442993164,
"rewards/check_answer/std": 0.16213203966617584,
"step": 165
},
{
"completion_length": 503.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 649.0,
"completions/max_terminated_length": 649.0,
"completions/mean_length": 503.5,
"completions/mean_terminated_length": 503.5,
"completions/min_length": 391.0,
"completions/min_terminated_length": 391.0,
"epoch": 0.029500622001066287,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3157219886779785,
"kl": 0.7668842077255249,
"learning_rate": 4.246031746031746e-06,
"loss": 0.0008,
"num_tokens": 485155.0,
"reward": 0.7842249870300293,
"reward_std": 0.1028033196926117,
"rewards/check_answer/mean": 0.7842249870300293,
"rewards/check_answer/std": 0.1028033196926117,
"step": 166
},
{
"completion_length": 480.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 685.0,
"completions/max_terminated_length": 685.0,
"completions/mean_length": 480.5,
"completions/mean_terminated_length": 480.5,
"completions/min_length": 305.0,
"completions/min_terminated_length": 305.0,
"epoch": 0.029678336591434158,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7447102069854736,
"kl": 1.1376843452453613,
"learning_rate": 4.238095238095239e-06,
"loss": 0.0011,
"num_tokens": 488257.0,
"reward": 0.468927800655365,
"reward_std": 0.09710308164358139,
"rewards/check_answer/mean": 0.468927800655365,
"rewards/check_answer/std": 0.09710308164358139,
"step": 167
},
{
"completion_length": 690.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 811.0,
"completions/max_terminated_length": 811.0,
"completions/mean_length": 690.0,
"completions/mean_terminated_length": 690.0,
"completions/min_length": 599.0,
"completions/min_terminated_length": 599.0,
"epoch": 0.029856051181802026,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.0533298254013062,
"kl": 0.6091443598270416,
"learning_rate": 4.23015873015873e-06,
"loss": 0.0006,
"num_tokens": 492269.0,
"reward": 0.565061628818512,
"reward_std": 0.0684511587023735,
"rewards/check_answer/mean": 0.565061628818512,
"rewards/check_answer/std": 0.0684511587023735,
"step": 168
},
{
"completion_length": 280.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 387.0,
"completions/max_terminated_length": 387.0,
"completions/mean_length": 280.25,
"completions/mean_terminated_length": 280.25,
"completions/min_length": 167.0,
"completions/min_terminated_length": 167.0,
"epoch": 0.030033765772169894,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.046386480331421,
"kl": 1.14199760556221,
"learning_rate": 4.222222222222223e-06,
"loss": 0.0011,
"num_tokens": 494498.0,
"reward": 0.30173179507255554,
"reward_std": 0.22188961505889893,
"rewards/check_answer/mean": 0.30173179507255554,
"rewards/check_answer/std": 0.22188962996006012,
"step": 169
},
{
"completion_length": 378.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 448.0,
"completions/max_terminated_length": 448.0,
"completions/mean_length": 378.75,
"completions/mean_terminated_length": 378.75,
"completions/min_length": 234.0,
"completions/min_terminated_length": 234.0,
"epoch": 0.030211480362537766,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9076892137527466,
"kl": 1.1338582932949066,
"learning_rate": 4.2142857142857145e-06,
"loss": 0.0011,
"num_tokens": 497121.0,
"reward": 0.22424927353858948,
"reward_std": 0.06156895309686661,
"rewards/check_answer/mean": 0.22424927353858948,
"rewards/check_answer/std": 0.06156895309686661,
"step": 170
},
{
"completion_length": 424.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 444.0,
"completions/max_terminated_length": 444.0,
"completions/mean_length": 424.75,
"completions/mean_terminated_length": 424.75,
"completions/min_length": 386.0,
"completions/min_terminated_length": 386.0,
"epoch": 0.030389194952905634,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9031684398651123,
"kl": 1.1977380812168121,
"learning_rate": 4.206349206349207e-06,
"loss": 0.0012,
"num_tokens": 499936.0,
"reward": 0.8130784034729004,
"reward_std": 0.12363550066947937,
"rewards/check_answer/mean": 0.8130784034729004,
"rewards/check_answer/std": 0.12363548576831818,
"step": 171
},
{
"completion_length": 510.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 552.0,
"completions/max_terminated_length": 552.0,
"completions/mean_length": 510.75,
"completions/mean_terminated_length": 510.75,
"completions/min_length": 470.0,
"completions/min_terminated_length": 470.0,
"epoch": 0.030566909543273502,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3281104564666748,
"kl": 1.2302511632442474,
"learning_rate": 4.198412698412699e-06,
"loss": 0.0012,
"num_tokens": 503151.0,
"reward": 0.8297035694122314,
"reward_std": 0.059518009424209595,
"rewards/check_answer/mean": 0.8297035694122314,
"rewards/check_answer/std": 0.059518001973629,
"step": 172
},
{
"completion_length": 467.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 603.0,
"completions/max_terminated_length": 603.0,
"completions/mean_length": 467.25,
"completions/mean_terminated_length": 467.25,
"completions/min_length": 321.0,
"completions/min_terminated_length": 321.0,
"epoch": 0.03074462413364137,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6283422708511353,
"kl": 1.2639419585466385,
"learning_rate": 4.190476190476191e-06,
"loss": 0.0013,
"num_tokens": 506156.0,
"reward": 0.8152176737785339,
"reward_std": 0.14423207938671112,
"rewards/check_answer/mean": 0.8152176737785339,
"rewards/check_answer/std": 0.14423207938671112,
"step": 173
},
{
"completion_length": 297.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 405.0,
"completions/max_terminated_length": 405.0,
"completions/mean_length": 297.75,
"completions/mean_terminated_length": 297.75,
"completions/min_length": 168.0,
"completions/min_terminated_length": 168.0,
"epoch": 0.030922338724009242,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8512938022613525,
"kl": 1.276896983385086,
"learning_rate": 4.182539682539683e-06,
"loss": 0.0013,
"num_tokens": 508459.0,
"reward": 0.8983228206634521,
"reward_std": 0.23461507260799408,
"rewards/check_answer/mean": 0.8983228206634521,
"rewards/check_answer/std": 0.23461505770683289,
"step": 174
},
{
"completion_length": 326.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 411.0,
"completions/max_terminated_length": 411.0,
"completions/mean_length": 326.25,
"completions/mean_terminated_length": 326.25,
"completions/min_length": 282.0,
"completions/min_terminated_length": 282.0,
"epoch": 0.03110005331437711,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8056871891021729,
"kl": 1.1819824874401093,
"learning_rate": 4.174603174603175e-06,
"loss": 0.0012,
"num_tokens": 510888.0,
"reward": 0.6271681785583496,
"reward_std": 0.08832673728466034,
"rewards/check_answer/mean": 0.6271681785583496,
"rewards/check_answer/std": 0.08832673728466034,
"step": 175
},
{
"completion_length": 531.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 577.0,
"completions/max_terminated_length": 577.0,
"completions/mean_length": 531.5,
"completions/mean_terminated_length": 531.5,
"completions/min_length": 408.0,
"completions/min_terminated_length": 408.0,
"epoch": 0.03127776790474498,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4564107656478882,
"kl": 1.3183227330446243,
"learning_rate": 4.166666666666667e-06,
"loss": 0.0013,
"num_tokens": 514166.0,
"reward": 0.6827693581581116,
"reward_std": 0.08819714188575745,
"rewards/check_answer/mean": 0.6827693581581116,
"rewards/check_answer/std": 0.08819711953401566,
"step": 176
},
{
"completion_length": 314.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 395.0,
"completions/max_terminated_length": 395.0,
"completions/mean_length": 314.0,
"completions/mean_terminated_length": 314.0,
"completions/min_length": 233.0,
"completions/min_terminated_length": 233.0,
"epoch": 0.031455482495112846,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.986071228981018,
"kl": 1.2774607837200165,
"learning_rate": 4.158730158730159e-06,
"loss": 0.0013,
"num_tokens": 516530.0,
"reward": 0.16138049960136414,
"reward_std": 0.04136866331100464,
"rewards/check_answer/mean": 0.16138049960136414,
"rewards/check_answer/std": 0.04136866703629494,
"step": 177
},
{
"completion_length": 636.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 869.0,
"completions/max_terminated_length": 869.0,
"completions/mean_length": 636.25,
"completions/mean_terminated_length": 636.25,
"completions/min_length": 491.0,
"completions/min_terminated_length": 491.0,
"epoch": 0.031633197085480715,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3712565898895264,
"kl": 0.7792616635560989,
"learning_rate": 4.150793650793651e-06,
"loss": 0.0008,
"num_tokens": 520303.0,
"reward": 0.5512293577194214,
"reward_std": 0.06619084626436234,
"rewards/check_answer/mean": 0.5512293577194214,
"rewards/check_answer/std": 0.06619084626436234,
"step": 178
},
{
"completion_length": 293.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 492.0,
"completions/max_terminated_length": 492.0,
"completions/mean_length": 293.5,
"completions/mean_terminated_length": 293.5,
"completions/min_length": 222.0,
"completions/min_terminated_length": 222.0,
"epoch": 0.03181091167584859,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8168277740478516,
"kl": 1.2853178530931473,
"learning_rate": 4.1428571428571435e-06,
"loss": 0.0013,
"num_tokens": 522605.0,
"reward": 0.7641668319702148,
"reward_std": 0.2912050485610962,
"rewards/check_answer/mean": 0.7641668319702148,
"rewards/check_answer/std": 0.2912050187587738,
"step": 179
},
{
"completion_length": 455.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 530.0,
"completions/max_terminated_length": 530.0,
"completions/mean_length": 455.25,
"completions/mean_terminated_length": 455.25,
"completions/min_length": 354.0,
"completions/min_terminated_length": 354.0,
"epoch": 0.03198862626621646,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6546411514282227,
"kl": 0.9170921891927719,
"learning_rate": 4.134920634920635e-06,
"loss": 0.0009,
"num_tokens": 525582.0,
"reward": 0.5996488928794861,
"reward_std": 0.16385574638843536,
"rewards/check_answer/mean": 0.5996488928794861,
"rewards/check_answer/std": 0.16385574638843536,
"step": 180
},
{
"completion_length": 413.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 532.0,
"completions/max_terminated_length": 532.0,
"completions/mean_length": 413.0,
"completions/mean_terminated_length": 413.0,
"completions/min_length": 357.0,
"completions/min_terminated_length": 357.0,
"epoch": 0.032166340856584326,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8733916282653809,
"kl": 1.0652090162038803,
"learning_rate": 4.126984126984127e-06,
"loss": 0.0011,
"num_tokens": 528350.0,
"reward": 0.8629907369613647,
"reward_std": 0.19258101284503937,
"rewards/check_answer/mean": 0.8629907369613647,
"rewards/check_answer/std": 0.19258098304271698,
"step": 181
},
{
"completion_length": 315.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 341.0,
"completions/max_terminated_length": 341.0,
"completions/mean_length": 315.75,
"completions/mean_terminated_length": 315.75,
"completions/min_length": 253.0,
"completions/min_terminated_length": 253.0,
"epoch": 0.032344055446952194,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.031298875808716,
"kl": 1.131786733865738,
"learning_rate": 4.119047619047619e-06,
"loss": 0.0011,
"num_tokens": 530741.0,
"reward": 0.45975661277770996,
"reward_std": 0.13406918942928314,
"rewards/check_answer/mean": 0.45975661277770996,
"rewards/check_answer/std": 0.13406920433044434,
"step": 182
},
{
"completion_length": 321.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 331.0,
"completions/max_terminated_length": 331.0,
"completions/mean_length": 321.0,
"completions/mean_terminated_length": 321.0,
"completions/min_length": 305.0,
"completions/min_terminated_length": 305.0,
"epoch": 0.03252177003732006,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.075031042098999,
"kl": 1.3818604350090027,
"learning_rate": 4.111111111111111e-06,
"loss": 0.0014,
"num_tokens": 533125.0,
"reward": 0.633269190788269,
"reward_std": 0.10433237999677658,
"rewards/check_answer/mean": 0.633269190788269,
"rewards/check_answer/std": 0.10433237999677658,
"step": 183
},
{
"completion_length": 331.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 389.0,
"completions/max_terminated_length": 389.0,
"completions/mean_length": 331.0,
"completions/mean_terminated_length": 331.0,
"completions/min_length": 212.0,
"completions/min_terminated_length": 212.0,
"epoch": 0.03269948462768793,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.916518211364746,
"kl": 1.230439841747284,
"learning_rate": 4.103174603174603e-06,
"loss": 0.0012,
"num_tokens": 535565.0,
"reward": 0.5834702253341675,
"reward_std": 0.1800461858510971,
"rewards/check_answer/mean": 0.5834702253341675,
"rewards/check_answer/std": 0.1800461858510971,
"step": 184
},
{
"completion_length": 484.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 616.0,
"completions/max_terminated_length": 616.0,
"completions/mean_length": 484.75,
"completions/mean_terminated_length": 484.75,
"completions/min_length": 284.0,
"completions/min_terminated_length": 284.0,
"epoch": 0.032877199218055805,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5022894144058228,
"kl": 0.8969959318637848,
"learning_rate": 4.095238095238096e-06,
"loss": 0.0009,
"num_tokens": 538676.0,
"reward": 0.4728669822216034,
"reward_std": 0.3648776710033417,
"rewards/check_answer/mean": 0.4728669822216034,
"rewards/check_answer/std": 0.3648776710033417,
"step": 185
},
{
"completion_length": 520.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 580.0,
"completions/max_terminated_length": 580.0,
"completions/mean_length": 520.0,
"completions/mean_terminated_length": 520.0,
"completions/min_length": 443.0,
"completions/min_terminated_length": 443.0,
"epoch": 0.033054913808423673,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.616834044456482,
"kl": 0.8219860643148422,
"learning_rate": 4.0873015873015875e-06,
"loss": 0.0008,
"num_tokens": 541912.0,
"reward": 0.5659242868423462,
"reward_std": 0.0763397291302681,
"rewards/check_answer/mean": 0.5659242868423462,
"rewards/check_answer/std": 0.0763397291302681,
"step": 186
},
{
"completion_length": 630.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 716.0,
"completions/max_terminated_length": 716.0,
"completions/mean_length": 630.25,
"completions/mean_terminated_length": 630.25,
"completions/min_length": 514.0,
"completions/min_terminated_length": 514.0,
"epoch": 0.03323262839879154,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3261500597000122,
"kl": 0.6475310772657394,
"learning_rate": 4.07936507936508e-06,
"loss": 0.0006,
"num_tokens": 545637.0,
"reward": 0.5103482604026794,
"reward_std": 0.04495608061552048,
"rewards/check_answer/mean": 0.5103482604026794,
"rewards/check_answer/std": 0.04495606571435928,
"step": 187
},
{
"completion_length": 656.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 983.0,
"completions/max_terminated_length": 983.0,
"completions/mean_length": 656.75,
"completions/mean_terminated_length": 656.75,
"completions/min_length": 469.0,
"completions/min_terminated_length": 469.0,
"epoch": 0.03341034298915941,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2920455932617188,
"kl": 0.8283561021089554,
"learning_rate": 4.071428571428572e-06,
"loss": 0.0008,
"num_tokens": 549436.0,
"reward": 0.5403710603713989,
"reward_std": 0.24990840256214142,
"rewards/check_answer/mean": 0.5403710603713989,
"rewards/check_answer/std": 0.2499084323644638,
"step": 188
},
{
"completion_length": 410.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 520.0,
"completions/max_terminated_length": 520.0,
"completions/mean_length": 410.5,
"completions/mean_terminated_length": 410.5,
"completions/min_length": 275.0,
"completions/min_terminated_length": 275.0,
"epoch": 0.03358805757952728,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.776705026626587,
"kl": 0.9838277697563171,
"learning_rate": 4.063492063492064e-06,
"loss": 0.001,
"num_tokens": 552222.0,
"reward": 0.6380422115325928,
"reward_std": 0.1814417541027069,
"rewards/check_answer/mean": 0.6380422115325928,
"rewards/check_answer/std": 0.1814417541027069,
"step": 189
},
{
"completion_length": 551.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 637.0,
"completions/max_terminated_length": 637.0,
"completions/mean_length": 551.5,
"completions/mean_terminated_length": 551.5,
"completions/min_length": 469.0,
"completions/min_terminated_length": 469.0,
"epoch": 0.033765772169895146,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4350589513778687,
"kl": 0.8946151733398438,
"learning_rate": 4.055555555555556e-06,
"loss": 0.0009,
"num_tokens": 555628.0,
"reward": 0.6631824374198914,
"reward_std": 0.0526728481054306,
"rewards/check_answer/mean": 0.6631824374198914,
"rewards/check_answer/std": 0.05267287418246269,
"step": 190
},
{
"completion_length": 335.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 372.0,
"completions/max_terminated_length": 372.0,
"completions/mean_length": 335.25,
"completions/mean_terminated_length": 335.25,
"completions/min_length": 290.0,
"completions/min_terminated_length": 290.0,
"epoch": 0.033943486760263014,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6080844402313232,
"kl": 1.0406917929649353,
"learning_rate": 4.047619047619048e-06,
"loss": 0.001,
"num_tokens": 558101.0,
"reward": 0.8645050525665283,
"reward_std": 0.16412971913814545,
"rewards/check_answer/mean": 0.8645050525665283,
"rewards/check_answer/std": 0.16412973403930664,
"step": 191
},
{
"completion_length": 295.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 321.0,
"completions/max_terminated_length": 321.0,
"completions/mean_length": 295.5,
"completions/mean_terminated_length": 295.5,
"completions/min_length": 279.0,
"completions/min_terminated_length": 279.0,
"epoch": 0.03412120135063089,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0053203105926514,
"kl": 1.3715750873088837,
"learning_rate": 4.03968253968254e-06,
"loss": 0.0014,
"num_tokens": 560415.0,
"reward": 0.37511610984802246,
"reward_std": 0.12241348624229431,
"rewards/check_answer/mean": 0.37511610984802246,
"rewards/check_answer/std": 0.12241347879171371,
"step": 192
},
{
"completion_length": 351.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 426.0,
"completions/max_terminated_length": 426.0,
"completions/mean_length": 351.0,
"completions/mean_terminated_length": 351.0,
"completions/min_length": 254.0,
"completions/min_terminated_length": 254.0,
"epoch": 0.03429891594099876,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6852611303329468,
"kl": 0.9062083661556244,
"learning_rate": 4.031746031746032e-06,
"loss": 0.0009,
"num_tokens": 562943.0,
"reward": 0.7322611808776855,
"reward_std": 0.10273779183626175,
"rewards/check_answer/mean": 0.7322611808776855,
"rewards/check_answer/std": 0.10273779183626175,
"step": 193
},
{
"completion_length": 387.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 441.0,
"completions/max_terminated_length": 441.0,
"completions/mean_length": 387.5,
"completions/mean_terminated_length": 387.5,
"completions/min_length": 258.0,
"completions/min_terminated_length": 258.0,
"epoch": 0.034476630531366625,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.495898962020874,
"kl": 0.9667275995016098,
"learning_rate": 4.023809523809524e-06,
"loss": 0.001,
"num_tokens": 565661.0,
"reward": 1.0379855632781982,
"reward_std": 0.013653420843183994,
"rewards/check_answer/mean": 1.0379855632781982,
"rewards/check_answer/std": 0.013653420843183994,
"step": 194
},
{
"completion_length": 516.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 655.0,
"completions/max_terminated_length": 655.0,
"completions/mean_length": 516.25,
"completions/mean_terminated_length": 516.25,
"completions/min_length": 363.0,
"completions/min_terminated_length": 363.0,
"epoch": 0.034654345121734494,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8335789442062378,
"kl": 2.6260423213243484,
"learning_rate": 4.0158730158730165e-06,
"loss": 0.0026,
"num_tokens": 568954.0,
"reward": 0.4649602472782135,
"reward_std": 0.05998750403523445,
"rewards/check_answer/mean": 0.4649602472782135,
"rewards/check_answer/std": 0.05998750776052475,
"step": 195
},
{
"completion_length": 486.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 562.0,
"completions/max_terminated_length": 562.0,
"completions/mean_length": 486.0,
"completions/mean_terminated_length": 486.0,
"completions/min_length": 445.0,
"completions/min_terminated_length": 445.0,
"epoch": 0.03483205971210236,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7766040563583374,
"kl": 1.0821466594934464,
"learning_rate": 4.007936507936508e-06,
"loss": 0.0011,
"num_tokens": 572042.0,
"reward": 0.5732154846191406,
"reward_std": 0.08260545134544373,
"rewards/check_answer/mean": 0.5732154846191406,
"rewards/check_answer/std": 0.08260544389486313,
"step": 196
},
{
"completion_length": 681.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 823.0,
"completions/max_terminated_length": 823.0,
"completions/mean_length": 681.0,
"completions/mean_terminated_length": 681.0,
"completions/min_length": 593.0,
"completions/min_terminated_length": 593.0,
"epoch": 0.03500977430247023,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1446372270584106,
"kl": 0.685902014374733,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0007,
"num_tokens": 576022.0,
"reward": 0.714977502822876,
"reward_std": 0.0700133666396141,
"rewards/check_answer/mean": 0.714977502822876,
"rewards/check_answer/std": 0.0700133740901947,
"step": 197
},
{
"completion_length": 267.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 280.0,
"completions/max_terminated_length": 280.0,
"completions/mean_length": 267.0,
"completions/mean_terminated_length": 267.0,
"completions/min_length": 238.0,
"completions/min_terminated_length": 238.0,
"epoch": 0.035187488892838105,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.380545139312744,
"kl": 1.7545035779476166,
"learning_rate": 3.992063492063492e-06,
"loss": 0.0018,
"num_tokens": 578194.0,
"reward": 0.5900921821594238,
"reward_std": 0.35790684819221497,
"rewards/check_answer/mean": 0.5900921821594238,
"rewards/check_answer/std": 0.3579068183898926,
"step": 198
},
{
"completion_length": 466.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 545.0,
"completions/max_terminated_length": 545.0,
"completions/mean_length": 466.5,
"completions/mean_terminated_length": 466.5,
"completions/min_length": 384.0,
"completions/min_terminated_length": 384.0,
"epoch": 0.03536520348320597,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8136112689971924,
"kl": 0.944290354847908,
"learning_rate": 3.984126984126984e-06,
"loss": 0.0009,
"num_tokens": 581228.0,
"reward": 0.5919873714447021,
"reward_std": 0.09145790338516235,
"rewards/check_answer/mean": 0.5919873714447021,
"rewards/check_answer/std": 0.09145790338516235,
"step": 199
},
{
"completion_length": 232.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 285.0,
"completions/max_terminated_length": 285.0,
"completions/mean_length": 232.75,
"completions/mean_terminated_length": 232.75,
"completions/min_length": 179.0,
"completions/min_terminated_length": 179.0,
"epoch": 0.03554291807357384,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.2700116634368896,
"kl": 1.3578459918498993,
"learning_rate": 3.9761904761904764e-06,
"loss": 0.0014,
"num_tokens": 583263.0,
"reward": 0.6666097640991211,
"reward_std": 0.313691109418869,
"rewards/check_answer/mean": 0.6666097640991211,
"rewards/check_answer/std": 0.3136911392211914,
"step": 200
},
{
"completion_length": 492.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 568.0,
"completions/max_terminated_length": 568.0,
"completions/mean_length": 492.75,
"completions/mean_terminated_length": 492.75,
"completions/min_length": 426.0,
"completions/min_terminated_length": 426.0,
"epoch": 0.03572063266394171,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5221089124679565,
"kl": 0.8189870119094849,
"learning_rate": 3.968253968253968e-06,
"loss": 0.0008,
"num_tokens": 586426.0,
"reward": 0.5162832736968994,
"reward_std": 0.054980941116809845,
"rewards/check_answer/mean": 0.5162832736968994,
"rewards/check_answer/std": 0.05498094484210014,
"step": 201
},
{
"completion_length": 405.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 463.0,
"completions/max_terminated_length": 463.0,
"completions/mean_length": 405.5,
"completions/mean_terminated_length": 405.5,
"completions/min_length": 347.0,
"completions/min_terminated_length": 347.0,
"epoch": 0.03589834725430958,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6586710214614868,
"kl": 1.192984789609909,
"learning_rate": 3.9603174603174606e-06,
"loss": 0.0012,
"num_tokens": 589196.0,
"reward": 0.8408722281455994,
"reward_std": 0.12209592759609222,
"rewards/check_answer/mean": 0.8408722281455994,
"rewards/check_answer/std": 0.12209593504667282,
"step": 202
},
{
"completion_length": 551.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 675.0,
"completions/max_terminated_length": 675.0,
"completions/mean_length": 551.0,
"completions/mean_terminated_length": 551.0,
"completions/min_length": 466.0,
"completions/min_terminated_length": 466.0,
"epoch": 0.036076061844677446,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3964484930038452,
"kl": 0.7525434195995331,
"learning_rate": 3.952380952380952e-06,
"loss": 0.0008,
"num_tokens": 592620.0,
"reward": 0.43263351917266846,
"reward_std": 0.058862458914518356,
"rewards/check_answer/mean": 0.43263351917266846,
"rewards/check_answer/std": 0.05886244773864746,
"step": 203
},
{
"completion_length": 329.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 385.0,
"completions/max_terminated_length": 385.0,
"completions/mean_length": 329.75,
"completions/mean_terminated_length": 329.75,
"completions/min_length": 294.0,
"completions/min_terminated_length": 294.0,
"epoch": 0.03625377643504532,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8947840929031372,
"kl": 1.2776507586240768,
"learning_rate": 3.944444444444445e-06,
"loss": 0.0013,
"num_tokens": 595067.0,
"reward": 0.8257237076759338,
"reward_std": 0.09456206113100052,
"rewards/check_answer/mean": 0.8257237076759338,
"rewards/check_answer/std": 0.09456204622983932,
"step": 204
},
{
"completion_length": 331.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 424.0,
"completions/max_terminated_length": 424.0,
"completions/mean_length": 331.0,
"completions/mean_terminated_length": 331.0,
"completions/min_length": 284.0,
"completions/min_terminated_length": 284.0,
"epoch": 0.03643149102541319,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9775497913360596,
"kl": 1.0914415419101715,
"learning_rate": 3.936507936507936e-06,
"loss": 0.0011,
"num_tokens": 597503.0,
"reward": 0.4486619532108307,
"reward_std": 0.20394665002822876,
"rewards/check_answer/mean": 0.4486619532108307,
"rewards/check_answer/std": 0.20394665002822876,
"step": 205
},
{
"completion_length": 525.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 612.0,
"completions/max_terminated_length": 612.0,
"completions/mean_length": 525.5,
"completions/mean_terminated_length": 525.5,
"completions/min_length": 480.0,
"completions/min_terminated_length": 480.0,
"epoch": 0.03660920561578106,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4579870700836182,
"kl": 0.8831206411123276,
"learning_rate": 3.928571428571429e-06,
"loss": 0.0009,
"num_tokens": 600773.0,
"reward": 0.5694155693054199,
"reward_std": 0.0706477239727974,
"rewards/check_answer/mean": 0.5694155693054199,
"rewards/check_answer/std": 0.0706477090716362,
"step": 206
},
{
"completion_length": 397.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 495.0,
"completions/max_terminated_length": 495.0,
"completions/mean_length": 397.0,
"completions/mean_terminated_length": 397.0,
"completions/min_length": 309.0,
"completions/min_terminated_length": 309.0,
"epoch": 0.036786920206148925,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9087179899215698,
"kl": 1.082411751151085,
"learning_rate": 3.920634920634921e-06,
"loss": 0.0011,
"num_tokens": 603501.0,
"reward": 0.6026151180267334,
"reward_std": 0.09127797186374664,
"rewards/check_answer/mean": 0.6026151180267334,
"rewards/check_answer/std": 0.09127800166606903,
"step": 207
},
{
"completion_length": 541.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 618.0,
"completions/max_terminated_length": 618.0,
"completions/mean_length": 541.5,
"completions/mean_terminated_length": 541.5,
"completions/min_length": 487.0,
"completions/min_terminated_length": 487.0,
"epoch": 0.03696463479651679,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5631201267242432,
"kl": 0.7632221281528473,
"learning_rate": 3.912698412698413e-06,
"loss": 0.0008,
"num_tokens": 606823.0,
"reward": 0.46171265840530396,
"reward_std": 0.0858674943447113,
"rewards/check_answer/mean": 0.46171265840530396,
"rewards/check_answer/std": 0.0858675092458725,
"step": 208
},
{
"completion_length": 299.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 358.0,
"completions/max_terminated_length": 358.0,
"completions/mean_length": 299.25,
"completions/mean_terminated_length": 299.25,
"completions/min_length": 217.0,
"completions/min_terminated_length": 217.0,
"epoch": 0.03714234938688466,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7001370191574097,
"kl": 1.1468262076377869,
"learning_rate": 3.9047619047619055e-06,
"loss": 0.0011,
"num_tokens": 609140.0,
"reward": 1.1788017749786377,
"reward_std": 0.23657526075839996,
"rewards/check_answer/mean": 1.1788017749786377,
"rewards/check_answer/std": 0.23657529056072235,
"step": 209
},
{
"completion_length": 321.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 377.0,
"completions/max_terminated_length": 377.0,
"completions/mean_length": 321.25,
"completions/mean_terminated_length": 321.25,
"completions/min_length": 252.0,
"completions/min_terminated_length": 252.0,
"epoch": 0.03732006397725253,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0018045902252197,
"kl": 1.2047775089740753,
"learning_rate": 3.896825396825397e-06,
"loss": 0.0012,
"num_tokens": 611549.0,
"reward": 0.4826478362083435,
"reward_std": 0.2732682228088379,
"rewards/check_answer/mean": 0.4826478362083435,
"rewards/check_answer/std": 0.2732682228088379,
"step": 210
},
{
"completion_length": 553.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 622.0,
"completions/max_terminated_length": 622.0,
"completions/mean_length": 553.5,
"completions/mean_terminated_length": 553.5,
"completions/min_length": 421.0,
"completions/min_terminated_length": 421.0,
"epoch": 0.037497778567620405,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4428868293762207,
"kl": 1.2566351145505905,
"learning_rate": 3.88888888888889e-06,
"loss": 0.0013,
"num_tokens": 614959.0,
"reward": 0.562415361404419,
"reward_std": 0.03217893838882446,
"rewards/check_answer/mean": 0.562415361404419,
"rewards/check_answer/std": 0.03217893838882446,
"step": 211
},
{
"completion_length": 533.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 616.0,
"completions/max_terminated_length": 616.0,
"completions/mean_length": 533.75,
"completions/mean_terminated_length": 533.75,
"completions/min_length": 441.0,
"completions/min_terminated_length": 441.0,
"epoch": 0.03767549315798827,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.474406361579895,
"kl": 0.8861031532287598,
"learning_rate": 3.880952380952381e-06,
"loss": 0.0009,
"num_tokens": 618286.0,
"reward": 0.716295599937439,
"reward_std": 0.08145526796579361,
"rewards/check_answer/mean": 0.716295599937439,
"rewards/check_answer/std": 0.08145526051521301,
"step": 212
},
{
"completion_length": 603.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 669.0,
"completions/max_terminated_length": 669.0,
"completions/mean_length": 603.75,
"completions/mean_terminated_length": 603.75,
"completions/min_length": 454.0,
"completions/min_terminated_length": 454.0,
"epoch": 0.03785320774835614,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3624635934829712,
"kl": 0.817508801817894,
"learning_rate": 3.873015873015874e-06,
"loss": 0.0008,
"num_tokens": 621881.0,
"reward": 0.3234153985977173,
"reward_std": 0.07267056405544281,
"rewards/check_answer/mean": 0.3234153985977173,
"rewards/check_answer/std": 0.07267057150602341,
"step": 213
},
{
"completion_length": 549.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 658.0,
"completions/max_terminated_length": 658.0,
"completions/mean_length": 549.5,
"completions/mean_terminated_length": 549.5,
"completions/min_length": 462.0,
"completions/min_terminated_length": 462.0,
"epoch": 0.03803092233872401,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5851331949234009,
"kl": 0.9887559115886688,
"learning_rate": 3.865079365079365e-06,
"loss": 0.001,
"num_tokens": 625263.0,
"reward": 0.7112226486206055,
"reward_std": 0.02053823135793209,
"rewards/check_answer/mean": 0.7112226486206055,
"rewards/check_answer/std": 0.02053823322057724,
"step": 214
},
{
"completion_length": 545.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 621.0,
"completions/max_terminated_length": 621.0,
"completions/mean_length": 545.25,
"completions/mean_terminated_length": 545.25,
"completions/min_length": 425.0,
"completions/min_terminated_length": 425.0,
"epoch": 0.03820863692909188,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.423057198524475,
"kl": 1.0859228074550629,
"learning_rate": 3.857142857142858e-06,
"loss": 0.0011,
"num_tokens": 628648.0,
"reward": 0.5066788196563721,
"reward_std": 0.026953205466270447,
"rewards/check_answer/mean": 0.5066788196563721,
"rewards/check_answer/std": 0.02695320174098015,
"step": 215
},
{
"completion_length": 483.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 613.0,
"completions/max_terminated_length": 613.0,
"completions/mean_length": 483.75,
"completions/mean_terminated_length": 483.75,
"completions/min_length": 411.0,
"completions/min_terminated_length": 411.0,
"epoch": 0.038386351519459745,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6435989141464233,
"kl": 0.8813229352235794,
"learning_rate": 3.8492063492063495e-06,
"loss": 0.0009,
"num_tokens": 631751.0,
"reward": 0.7219762802124023,
"reward_std": 0.17963886260986328,
"rewards/check_answer/mean": 0.7219762802124023,
"rewards/check_answer/std": 0.17963886260986328,
"step": 216
},
{
"completion_length": 537.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 697.0,
"completions/max_terminated_length": 697.0,
"completions/mean_length": 537.5,
"completions/mean_terminated_length": 537.5,
"completions/min_length": 435.0,
"completions/min_terminated_length": 435.0,
"epoch": 0.03856406610982762,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4791259765625,
"kl": 0.9588049501180649,
"learning_rate": 3.841269841269842e-06,
"loss": 0.001,
"num_tokens": 635053.0,
"reward": 0.5066652894020081,
"reward_std": 0.0735304206609726,
"rewards/check_answer/mean": 0.5066652894020081,
"rewards/check_answer/std": 0.07353042811155319,
"step": 217
},
{
"completion_length": 392.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 492.0,
"completions/max_terminated_length": 492.0,
"completions/mean_length": 392.75,
"completions/mean_terminated_length": 392.75,
"completions/min_length": 245.0,
"completions/min_terminated_length": 245.0,
"epoch": 0.03874178070019549,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9984493255615234,
"kl": 1.0702104568481445,
"learning_rate": 3.833333333333334e-06,
"loss": 0.0011,
"num_tokens": 637744.0,
"reward": 0.7809625267982483,
"reward_std": 0.5125808119773865,
"rewards/check_answer/mean": 0.7809625267982483,
"rewards/check_answer/std": 0.5125808119773865,
"step": 218
},
{
"completion_length": 447.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 499.0,
"completions/max_terminated_length": 499.0,
"completions/mean_length": 447.75,
"completions/mean_terminated_length": 447.75,
"completions/min_length": 360.0,
"completions/min_terminated_length": 360.0,
"epoch": 0.038919495290563356,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.494664192199707,
"kl": 0.837522029876709,
"learning_rate": 3.825396825396825e-06,
"loss": 0.0008,
"num_tokens": 640751.0,
"reward": 1.227266550064087,
"reward_std": 0.04486949369311333,
"rewards/check_answer/mean": 1.227266550064087,
"rewards/check_answer/std": 0.04486951231956482,
"step": 219
},
{
"completion_length": 407.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 558.0,
"completions/max_terminated_length": 558.0,
"completions/mean_length": 407.0,
"completions/mean_terminated_length": 407.0,
"completions/min_length": 277.0,
"completions/min_terminated_length": 277.0,
"epoch": 0.039097209880931225,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7991682291030884,
"kl": 1.061606004834175,
"learning_rate": 3.817460317460318e-06,
"loss": 0.0011,
"num_tokens": 643523.0,
"reward": 0.7400416135787964,
"reward_std": 0.11785461008548737,
"rewards/check_answer/mean": 0.7400416135787964,
"rewards/check_answer/std": 0.11785461753606796,
"step": 220
},
{
"completion_length": 523.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 602.0,
"completions/max_terminated_length": 602.0,
"completions/mean_length": 523.0,
"completions/mean_terminated_length": 523.0,
"completions/min_length": 450.0,
"completions/min_terminated_length": 450.0,
"epoch": 0.03927492447129909,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4204972982406616,
"kl": 1.000058576464653,
"learning_rate": 3.80952380952381e-06,
"loss": 0.001,
"num_tokens": 646795.0,
"reward": 0.7960029244422913,
"reward_std": 0.11042596399784088,
"rewards/check_answer/mean": 0.7960029244422913,
"rewards/check_answer/std": 0.11042595654726028,
"step": 221
},
{
"completion_length": 509.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 586.0,
"completions/max_terminated_length": 586.0,
"completions/mean_length": 509.75,
"completions/mean_terminated_length": 509.75,
"completions/min_length": 406.0,
"completions/min_terminated_length": 406.0,
"epoch": 0.03945263906166696,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4458086490631104,
"kl": 1.0492773056030273,
"learning_rate": 3.801587301587302e-06,
"loss": 0.001,
"num_tokens": 650042.0,
"reward": 0.8759781718254089,
"reward_std": 0.07902158796787262,
"rewards/check_answer/mean": 0.8759781718254089,
"rewards/check_answer/std": 0.07902161031961441,
"step": 222
},
{
"completion_length": 422.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 494.0,
"completions/max_terminated_length": 494.0,
"completions/mean_length": 422.25,
"completions/mean_terminated_length": 422.25,
"completions/min_length": 330.0,
"completions/min_terminated_length": 330.0,
"epoch": 0.03963035365203483,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7075461149215698,
"kl": 1.1013980060815811,
"learning_rate": 3.793650793650794e-06,
"loss": 0.0011,
"num_tokens": 652883.0,
"reward": 0.6839185953140259,
"reward_std": 0.16940036416053772,
"rewards/check_answer/mean": 0.6839185953140259,
"rewards/check_answer/std": 0.1694003790616989,
"step": 223
},
{
"completion_length": 410.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 459.0,
"completions/max_terminated_length": 459.0,
"completions/mean_length": 410.0,
"completions/mean_terminated_length": 410.0,
"completions/min_length": 368.0,
"completions/min_terminated_length": 368.0,
"epoch": 0.039808068242402704,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.579961895942688,
"kl": 1.1135330200195312,
"learning_rate": 3.785714285714286e-06,
"loss": 0.0011,
"num_tokens": 655659.0,
"reward": 0.7339705228805542,
"reward_std": 0.09850845485925674,
"rewards/check_answer/mean": 0.7339705228805542,
"rewards/check_answer/std": 0.09850846230983734,
"step": 224
},
{
"completion_length": 405.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 492.0,
"completions/max_terminated_length": 492.0,
"completions/mean_length": 405.75,
"completions/mean_terminated_length": 405.75,
"completions/min_length": 327.0,
"completions/min_terminated_length": 327.0,
"epoch": 0.03998578283277057,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5088939666748047,
"kl": 1.0870481431484222,
"learning_rate": 3.777777777777778e-06,
"loss": 0.0011,
"num_tokens": 658394.0,
"reward": 0.6195275783538818,
"reward_std": 0.32846853137016296,
"rewards/check_answer/mean": 0.6195275783538818,
"rewards/check_answer/std": 0.32846853137016296,
"step": 225
},
{
"completion_length": 425.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 504.0,
"completions/max_terminated_length": 504.0,
"completions/mean_length": 425.75,
"completions/mean_terminated_length": 425.75,
"completions/min_length": 352.0,
"completions/min_terminated_length": 352.0,
"epoch": 0.04016349742313844,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6888916492462158,
"kl": 1.0393297523260117,
"learning_rate": 3.76984126984127e-06,
"loss": 0.001,
"num_tokens": 661237.0,
"reward": 0.5425740480422974,
"reward_std": 0.14747239649295807,
"rewards/check_answer/mean": 0.5425740480422974,
"rewards/check_answer/std": 0.14747241139411926,
"step": 226
},
{
"completion_length": 430.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 551.0,
"completions/max_terminated_length": 551.0,
"completions/mean_length": 430.0,
"completions/mean_terminated_length": 430.0,
"completions/min_length": 325.0,
"completions/min_terminated_length": 325.0,
"epoch": 0.04034121201350631,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.497071385383606,
"kl": 0.8925988674163818,
"learning_rate": 3.761904761904762e-06,
"loss": 0.0009,
"num_tokens": 664129.0,
"reward": 0.5838330984115601,
"reward_std": 0.14572261273860931,
"rewards/check_answer/mean": 0.5838330984115601,
"rewards/check_answer/std": 0.14572261273860931,
"step": 227
},
{
"completion_length": 387.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 466.0,
"completions/max_terminated_length": 466.0,
"completions/mean_length": 387.75,
"completions/mean_terminated_length": 387.75,
"completions/min_length": 284.0,
"completions/min_terminated_length": 284.0,
"epoch": 0.04051892660387418,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.95894193649292,
"kl": 1.2442685663700104,
"learning_rate": 3.7539682539682543e-06,
"loss": 0.0012,
"num_tokens": 666784.0,
"reward": 0.3317446708679199,
"reward_std": 0.12045136839151382,
"rewards/check_answer/mean": 0.3317446708679199,
"rewards/check_answer/std": 0.12045136094093323,
"step": 228
},
{
"completion_length": 627.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 788.0,
"completions/max_terminated_length": 788.0,
"completions/mean_length": 627.25,
"completions/mean_terminated_length": 627.25,
"completions/min_length": 499.0,
"completions/min_terminated_length": 499.0,
"epoch": 0.040696641194242045,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2475444078445435,
"kl": 0.8898588418960571,
"learning_rate": 3.7460317460317463e-06,
"loss": 0.0009,
"num_tokens": 670541.0,
"reward": 0.5734164714813232,
"reward_std": 0.06777976453304291,
"rewards/check_answer/mean": 0.5734164714813232,
"rewards/check_answer/std": 0.0677797719836235,
"step": 229
},
{
"completion_length": 528.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 586.0,
"completions/max_terminated_length": 586.0,
"completions/mean_length": 528.25,
"completions/mean_terminated_length": 528.25,
"completions/min_length": 443.0,
"completions/min_terminated_length": 443.0,
"epoch": 0.04087435578460992,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5907459259033203,
"kl": 0.9156730771064758,
"learning_rate": 3.7380952380952384e-06,
"loss": 0.0009,
"num_tokens": 673814.0,
"reward": 0.6122113466262817,
"reward_std": 0.06687498092651367,
"rewards/check_answer/mean": 0.6122113466262817,
"rewards/check_answer/std": 0.06687495857477188,
"step": 230
},
{
"completion_length": 572.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 606.0,
"completions/max_terminated_length": 606.0,
"completions/mean_length": 572.25,
"completions/mean_terminated_length": 572.25,
"completions/min_length": 514.0,
"completions/min_terminated_length": 514.0,
"epoch": 0.04105207037497779,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.419886589050293,
"kl": 0.9749362170696259,
"learning_rate": 3.7301587301587305e-06,
"loss": 0.001,
"num_tokens": 677267.0,
"reward": 0.8499937653541565,
"reward_std": 0.16837330162525177,
"rewards/check_answer/mean": 0.8499937653541565,
"rewards/check_answer/std": 0.16837328672409058,
"step": 231
},
{
"completion_length": 461.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 542.0,
"completions/max_terminated_length": 542.0,
"completions/mean_length": 461.75,
"completions/mean_terminated_length": 461.75,
"completions/min_length": 377.0,
"completions/min_terminated_length": 377.0,
"epoch": 0.041229784965345656,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.645576000213623,
"kl": 0.988830104470253,
"learning_rate": 3.7222222222222225e-06,
"loss": 0.001,
"num_tokens": 680234.0,
"reward": 0.9341785311698914,
"reward_std": 0.23470927774906158,
"rewards/check_answer/mean": 0.9341785311698914,
"rewards/check_answer/std": 0.23470927774906158,
"step": 232
},
{
"completion_length": 610.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 754.0,
"completions/max_terminated_length": 754.0,
"completions/mean_length": 610.0,
"completions/mean_terminated_length": 610.0,
"completions/min_length": 448.0,
"completions/min_terminated_length": 448.0,
"epoch": 0.041407499555713524,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2507656812667847,
"kl": 0.7385053485631943,
"learning_rate": 3.7142857142857146e-06,
"loss": 0.0007,
"num_tokens": 683910.0,
"reward": 0.5352281332015991,
"reward_std": 0.04549167677760124,
"rewards/check_answer/mean": 0.5352281332015991,
"rewards/check_answer/std": 0.04549166187644005,
"step": 233
},
{
"completion_length": 609.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 648.0,
"completions/max_terminated_length": 648.0,
"completions/mean_length": 609.75,
"completions/mean_terminated_length": 609.75,
"completions/min_length": 558.0,
"completions/min_terminated_length": 558.0,
"epoch": 0.04158521414608139,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3870525360107422,
"kl": 0.7475785911083221,
"learning_rate": 3.7063492063492067e-06,
"loss": 0.0007,
"num_tokens": 687569.0,
"reward": 0.4529799520969391,
"reward_std": 0.045896902680397034,
"rewards/check_answer/mean": 0.4529799520969391,
"rewards/check_answer/std": 0.04589690640568733,
"step": 234
},
{
"completion_length": 511.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 583.0,
"completions/max_terminated_length": 583.0,
"completions/mean_length": 511.75,
"completions/mean_terminated_length": 511.75,
"completions/min_length": 420.0,
"completions/min_terminated_length": 420.0,
"epoch": 0.04176292873644926,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6288220882415771,
"kl": 0.9887201339006424,
"learning_rate": 3.6984126984126987e-06,
"loss": 0.001,
"num_tokens": 690784.0,
"reward": 0.705742359161377,
"reward_std": 0.1110423356294632,
"rewards/check_answer/mean": 0.705742359161377,
"rewards/check_answer/std": 0.11104235798120499,
"step": 235
},
{
"completion_length": 531.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 603.0,
"completions/max_terminated_length": 603.0,
"completions/mean_length": 531.5,
"completions/mean_terminated_length": 531.5,
"completions/min_length": 350.0,
"completions/min_terminated_length": 350.0,
"epoch": 0.04194064332681713,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2495639324188232,
"kl": 0.7813977897167206,
"learning_rate": 3.690476190476191e-06,
"loss": 0.0008,
"num_tokens": 694166.0,
"reward": 1.2415096759796143,
"reward_std": 0.19065117835998535,
"rewards/check_answer/mean": 1.2415096759796143,
"rewards/check_answer/std": 0.19065116345882416,
"step": 236
},
{
"completion_length": 403.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 450.0,
"completions/max_terminated_length": 450.0,
"completions/mean_length": 403.25,
"completions/mean_terminated_length": 403.25,
"completions/min_length": 345.0,
"completions/min_terminated_length": 345.0,
"epoch": 0.042118357917185004,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8500796556472778,
"kl": 1.0187488496303558,
"learning_rate": 3.6825396825396833e-06,
"loss": 0.001,
"num_tokens": 696903.0,
"reward": 0.6371432542800903,
"reward_std": 0.07777383923530579,
"rewards/check_answer/mean": 0.6371432542800903,
"rewards/check_answer/std": 0.07777383923530579,
"step": 237
},
{
"completion_length": 454.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 533.0,
"completions/max_terminated_length": 533.0,
"completions/mean_length": 454.5,
"completions/mean_terminated_length": 454.5,
"completions/min_length": 386.0,
"completions/min_terminated_length": 386.0,
"epoch": 0.04229607250755287,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4619567394256592,
"kl": 0.7681007236242294,
"learning_rate": 3.6746031746031754e-06,
"loss": 0.0008,
"num_tokens": 699897.0,
"reward": 0.6196759939193726,
"reward_std": 0.051524724811315536,
"rewards/check_answer/mean": 0.6196759939193726,
"rewards/check_answer/std": 0.051524724811315536,
"step": 238
},
{
"completion_length": 376.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 462.0,
"completions/max_terminated_length": 462.0,
"completions/mean_length": 376.0,
"completions/mean_terminated_length": 376.0,
"completions/min_length": 311.0,
"completions/min_terminated_length": 311.0,
"epoch": 0.04247378709792074,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5873010158538818,
"kl": 0.9846309572458267,
"learning_rate": 3.6666666666666666e-06,
"loss": 0.001,
"num_tokens": 702549.0,
"reward": 0.5979254245758057,
"reward_std": 0.08328143507242203,
"rewards/check_answer/mean": 0.5979254245758057,
"rewards/check_answer/std": 0.08328142762184143,
"step": 239
},
{
"completion_length": 417.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 492.0,
"completions/max_terminated_length": 492.0,
"completions/mean_length": 417.0,
"completions/mean_terminated_length": 417.0,
"completions/min_length": 343.0,
"completions/min_terminated_length": 343.0,
"epoch": 0.04265150168828861,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7484502792358398,
"kl": 1.022307112812996,
"learning_rate": 3.6587301587301586e-06,
"loss": 0.001,
"num_tokens": 705357.0,
"reward": 0.35202890634536743,
"reward_std": 0.06691336631774902,
"rewards/check_answer/mean": 0.35202890634536743,
"rewards/check_answer/std": 0.06691337376832962,
"step": 240
},
{
"completion_length": 423.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 551.0,
"completions/max_terminated_length": 551.0,
"completions/mean_length": 423.5,
"completions/mean_terminated_length": 423.5,
"completions/min_length": 363.0,
"completions/min_terminated_length": 363.0,
"epoch": 0.042829216278656476,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7925474643707275,
"kl": 1.7321273684501648,
"learning_rate": 3.6507936507936507e-06,
"loss": 0.0017,
"num_tokens": 708235.0,
"reward": 0.6433290243148804,
"reward_std": 0.17733269929885864,
"rewards/check_answer/mean": 0.6433290243148804,
"rewards/check_answer/std": 0.17733269929885864,
"step": 241
},
{
"completion_length": 344.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 444.0,
"completions/max_terminated_length": 444.0,
"completions/mean_length": 344.75,
"completions/mean_terminated_length": 344.75,
"completions/min_length": 244.0,
"completions/min_terminated_length": 244.0,
"epoch": 0.043006930869024344,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7820135354995728,
"kl": 1.646221399307251,
"learning_rate": 3.642857142857143e-06,
"loss": 0.0016,
"num_tokens": 710738.0,
"reward": 0.4637607932090759,
"reward_std": 0.2009269893169403,
"rewards/check_answer/mean": 0.4637607932090759,
"rewards/check_answer/std": 0.2009270042181015,
"step": 242
},
{
"completion_length": 279.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 336.0,
"completions/max_terminated_length": 336.0,
"completions/mean_length": 279.0,
"completions/mean_terminated_length": 279.0,
"completions/min_length": 233.0,
"completions/min_terminated_length": 233.0,
"epoch": 0.04318464545939222,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.884946346282959,
"kl": 1.4627968966960907,
"learning_rate": 3.6349206349206353e-06,
"loss": 0.0015,
"num_tokens": 712954.0,
"reward": 0.08365356922149658,
"reward_std": 0.048354603350162506,
"rewards/check_answer/mean": 0.08365356922149658,
"rewards/check_answer/std": 0.048354603350162506,
"step": 243
},
{
"completion_length": 521.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 646.0,
"completions/max_terminated_length": 646.0,
"completions/mean_length": 521.0,
"completions/mean_terminated_length": 521.0,
"completions/min_length": 433.0,
"completions/min_terminated_length": 433.0,
"epoch": 0.04336236004976009,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3385543823242188,
"kl": 0.8651070147752762,
"learning_rate": 3.6269841269841273e-06,
"loss": 0.0009,
"num_tokens": 716238.0,
"reward": 0.8709990382194519,
"reward_std": 0.025177743285894394,
"rewards/check_answer/mean": 0.8709990382194519,
"rewards/check_answer/std": 0.025177743285894394,
"step": 244
},
{
"completion_length": 534.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 588.0,
"completions/max_terminated_length": 588.0,
"completions/mean_length": 534.75,
"completions/mean_terminated_length": 534.75,
"completions/min_length": 391.0,
"completions/min_terminated_length": 391.0,
"epoch": 0.043540074640127956,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4057061672210693,
"kl": 1.1244578957557678,
"learning_rate": 3.6190476190476194e-06,
"loss": 0.0011,
"num_tokens": 719533.0,
"reward": 0.9694785475730896,
"reward_std": 0.1145370677113533,
"rewards/check_answer/mean": 0.9694785475730896,
"rewards/check_answer/std": 0.1145370677113533,
"step": 245
},
{
"completion_length": 387.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 416.0,
"completions/max_terminated_length": 416.0,
"completions/mean_length": 387.25,
"completions/mean_terminated_length": 387.25,
"completions/min_length": 325.0,
"completions/min_terminated_length": 325.0,
"epoch": 0.043717789230495824,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.820640206336975,
"kl": 1.5689309239387512,
"learning_rate": 3.6111111111111115e-06,
"loss": 0.0016,
"num_tokens": 722202.0,
"reward": 0.4090389609336853,
"reward_std": 0.09788350760936737,
"rewards/check_answer/mean": 0.4090389609336853,
"rewards/check_answer/std": 0.09788351505994797,
"step": 246
},
{
"completion_length": 524.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 615.0,
"completions/max_terminated_length": 615.0,
"completions/mean_length": 524.75,
"completions/mean_terminated_length": 524.75,
"completions/min_length": 416.0,
"completions/min_terminated_length": 416.0,
"epoch": 0.04389550382086369,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3757284879684448,
"kl": 0.7826510369777679,
"learning_rate": 3.6031746031746035e-06,
"loss": 0.0008,
"num_tokens": 725489.0,
"reward": 0.9417409896850586,
"reward_std": 0.39931824803352356,
"rewards/check_answer/mean": 0.9417409896850586,
"rewards/check_answer/std": 0.39931821823120117,
"step": 247
},
{
"completion_length": 692.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 981.0,
"completions/max_terminated_length": 981.0,
"completions/mean_length": 692.75,
"completions/mean_terminated_length": 692.75,
"completions/min_length": 486.0,
"completions/min_terminated_length": 486.0,
"epoch": 0.04407321841123156,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1299306154251099,
"kl": 0.7229564934968948,
"learning_rate": 3.5952380952380956e-06,
"loss": 0.0007,
"num_tokens": 729500.0,
"reward": 0.6503614783287048,
"reward_std": 0.045035649091005325,
"rewards/check_answer/mean": 0.6503614783287048,
"rewards/check_answer/std": 0.04503565654158592,
"step": 248
},
{
"completion_length": 480.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 591.0,
"completions/max_terminated_length": 591.0,
"completions/mean_length": 480.5,
"completions/mean_terminated_length": 480.5,
"completions/min_length": 351.0,
"completions/min_terminated_length": 351.0,
"epoch": 0.04425093300159943,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3678374290466309,
"kl": 0.8728310763835907,
"learning_rate": 3.5873015873015877e-06,
"loss": 0.0009,
"num_tokens": 732590.0,
"reward": 0.8919079303741455,
"reward_std": 0.10072972625494003,
"rewards/check_answer/mean": 0.8919079303741455,
"rewards/check_answer/std": 0.10072971135377884,
"step": 249
},
{
"completion_length": 466.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 505.0,
"completions/max_terminated_length": 505.0,
"completions/mean_length": 466.75,
"completions/mean_terminated_length": 466.75,
"completions/min_length": 437.0,
"completions/min_terminated_length": 437.0,
"epoch": 0.0444286475919673,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.570044994354248,
"kl": 0.9534753710031509,
"learning_rate": 3.5793650793650797e-06,
"loss": 0.001,
"num_tokens": 735681.0,
"reward": 0.6919865012168884,
"reward_std": 0.014157108031213284,
"rewards/check_answer/mean": 0.6919865012168884,
"rewards/check_answer/std": 0.014157108031213284,
"step": 250
},
{
"completion_length": 376.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 505.0,
"completions/max_terminated_length": 505.0,
"completions/mean_length": 376.5,
"completions/mean_terminated_length": 376.5,
"completions/min_length": 261.0,
"completions/min_terminated_length": 261.0,
"epoch": 0.04460636218233517,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9166706800460815,
"kl": 0.9647689759731293,
"learning_rate": 3.5714285714285718e-06,
"loss": 0.001,
"num_tokens": 738319.0,
"reward": 0.44414740800857544,
"reward_std": 0.13158369064331055,
"rewards/check_answer/mean": 0.44414740800857544,
"rewards/check_answer/std": 0.13158370554447174,
"step": 251
},
{
"completion_length": 319.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 406.0,
"completions/max_terminated_length": 406.0,
"completions/mean_length": 319.0,
"completions/mean_terminated_length": 319.0,
"completions/min_length": 231.0,
"completions/min_terminated_length": 231.0,
"epoch": 0.04478407677270304,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.1092257499694824,
"kl": 1.1556675136089325,
"learning_rate": 3.563492063492064e-06,
"loss": 0.0012,
"num_tokens": 740707.0,
"reward": 0.5653150081634521,
"reward_std": 0.1314965784549713,
"rewards/check_answer/mean": 0.5653150081634521,
"rewards/check_answer/std": 0.1314965784549713,
"step": 252
},
{
"completion_length": 321.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 339.0,
"completions/max_terminated_length": 339.0,
"completions/mean_length": 321.0,
"completions/mean_terminated_length": 321.0,
"completions/min_length": 305.0,
"completions/min_terminated_length": 305.0,
"epoch": 0.04496179136307091,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.977848768234253,
"kl": 1.1421026587486267,
"learning_rate": 3.555555555555556e-06,
"loss": 0.0011,
"num_tokens": 743111.0,
"reward": 1.2765793800354004,
"reward_std": 0.37045302987098694,
"rewards/check_answer/mean": 1.2765793800354004,
"rewards/check_answer/std": 0.3704530596733093,
"step": 253
},
{
"completion_length": 491.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 574.0,
"completions/max_terminated_length": 574.0,
"completions/mean_length": 491.5,
"completions/mean_terminated_length": 491.5,
"completions/min_length": 415.0,
"completions/min_terminated_length": 415.0,
"epoch": 0.045139505953438776,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4590531587600708,
"kl": 1.0585539489984512,
"learning_rate": 3.547619047619048e-06,
"loss": 0.0011,
"num_tokens": 746269.0,
"reward": 0.9823049306869507,
"reward_std": 0.07934407889842987,
"rewards/check_answer/mean": 0.9823049306869507,
"rewards/check_answer/std": 0.07934407889842987,
"step": 254
},
{
"completion_length": 429.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 472.0,
"completions/max_terminated_length": 472.0,
"completions/mean_length": 429.0,
"completions/mean_terminated_length": 429.0,
"completions/min_length": 353.0,
"completions/min_terminated_length": 353.0,
"epoch": 0.045317220543806644,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9310379028320312,
"kl": 1.1345547437667847,
"learning_rate": 3.53968253968254e-06,
"loss": 0.0011,
"num_tokens": 749121.0,
"reward": 0.45538854598999023,
"reward_std": 0.10226347297430038,
"rewards/check_answer/mean": 0.45538854598999023,
"rewards/check_answer/std": 0.10226348042488098,
"step": 255
},
{
"completion_length": 404.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 415.0,
"completions/max_terminated_length": 415.0,
"completions/mean_length": 404.5,
"completions/mean_terminated_length": 404.5,
"completions/min_length": 387.0,
"completions/min_terminated_length": 387.0,
"epoch": 0.04549493513417452,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4988094568252563,
"kl": 1.2277261018753052,
"learning_rate": 3.531746031746032e-06,
"loss": 0.0012,
"num_tokens": 751847.0,
"reward": 0.504880428314209,
"reward_std": 0.26580870151519775,
"rewards/check_answer/mean": 0.504880428314209,
"rewards/check_answer/std": 0.26580870151519775,
"step": 256
},
{
"completion_length": 606.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 704.0,
"completions/max_terminated_length": 704.0,
"completions/mean_length": 606.75,
"completions/mean_terminated_length": 606.75,
"completions/min_length": 545.0,
"completions/min_terminated_length": 545.0,
"epoch": 0.04567264972454239,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3845511674880981,
"kl": 0.7865335196256638,
"learning_rate": 3.523809523809524e-06,
"loss": 0.0008,
"num_tokens": 755498.0,
"reward": 0.4449111521244049,
"reward_std": 0.04709044471383095,
"rewards/check_answer/mean": 0.4449111521244049,
"rewards/check_answer/std": 0.04709044098854065,
"step": 257
},
{
"completion_length": 568.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 588.0,
"completions/max_terminated_length": 588.0,
"completions/mean_length": 568.25,
"completions/mean_terminated_length": 568.25,
"completions/min_length": 536.0,
"completions/min_terminated_length": 536.0,
"epoch": 0.045850364314910255,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5005916357040405,
"kl": 0.961561843752861,
"learning_rate": 3.5158730158730162e-06,
"loss": 0.001,
"num_tokens": 758955.0,
"reward": 0.9184130430221558,
"reward_std": 0.08506201952695847,
"rewards/check_answer/mean": 0.9184130430221558,
"rewards/check_answer/std": 0.08506201952695847,
"step": 258
},
{
"completion_length": 488.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 599.0,
"completions/max_terminated_length": 599.0,
"completions/mean_length": 488.25,
"completions/mean_terminated_length": 488.25,
"completions/min_length": 409.0,
"completions/min_terminated_length": 409.0,
"epoch": 0.04602807890527812,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.433339238166809,
"kl": 1.8387483060359955,
"learning_rate": 3.507936507936508e-06,
"loss": 0.0018,
"num_tokens": 762080.0,
"reward": 0.5739443302154541,
"reward_std": 0.04967578500509262,
"rewards/check_answer/mean": 0.5739443302154541,
"rewards/check_answer/std": 0.04967578500509262,
"step": 259
},
{
"completion_length": 665.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 779.0,
"completions/max_terminated_length": 779.0,
"completions/mean_length": 665.0,
"completions/mean_terminated_length": 665.0,
"completions/min_length": 594.0,
"completions/min_terminated_length": 594.0,
"epoch": 0.04620579349564599,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3149513006210327,
"kl": 0.5900047421455383,
"learning_rate": 3.5e-06,
"loss": 0.0006,
"num_tokens": 765968.0,
"reward": 0.5332039594650269,
"reward_std": 0.03528818115592003,
"rewards/check_answer/mean": 0.5332039594650269,
"rewards/check_answer/std": 0.03528818115592003,
"step": 260
},
{
"completion_length": 470.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 551.0,
"completions/max_terminated_length": 551.0,
"completions/mean_length": 470.0,
"completions/mean_terminated_length": 470.0,
"completions/min_length": 378.0,
"completions/min_terminated_length": 378.0,
"epoch": 0.04638350808601386,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5507720708847046,
"kl": 0.8859493583440781,
"learning_rate": 3.492063492063492e-06,
"loss": 0.0009,
"num_tokens": 769020.0,
"reward": 1.028480052947998,
"reward_std": 0.039744578301906586,
"rewards/check_answer/mean": 1.028480052947998,
"rewards/check_answer/std": 0.03974457457661629,
"step": 261
},
{
"completion_length": 562.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 594.0,
"completions/max_terminated_length": 594.0,
"completions/mean_length": 562.5,
"completions/mean_terminated_length": 562.5,
"completions/min_length": 518.0,
"completions/min_terminated_length": 518.0,
"epoch": 0.04656122267638173,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4958258867263794,
"kl": 0.9075725972652435,
"learning_rate": 3.484126984126984e-06,
"loss": 0.0009,
"num_tokens": 772478.0,
"reward": 0.5773640871047974,
"reward_std": 0.07390649616718292,
"rewards/check_answer/mean": 0.5773640871047974,
"rewards/check_answer/std": 0.07390650361776352,
"step": 262
},
{
"completion_length": 534.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 592.0,
"completions/max_terminated_length": 592.0,
"completions/mean_length": 534.5,
"completions/mean_terminated_length": 534.5,
"completions/min_length": 404.0,
"completions/min_terminated_length": 404.0,
"epoch": 0.0467389372667496,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.40358567237854,
"kl": 0.9262001812458038,
"learning_rate": 3.476190476190476e-06,
"loss": 0.0009,
"num_tokens": 775804.0,
"reward": 0.7191150784492493,
"reward_std": 0.18629220128059387,
"rewards/check_answer/mean": 0.7191150784492493,
"rewards/check_answer/std": 0.18629220128059387,
"step": 263
},
{
"completion_length": 729.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 931.0,
"completions/max_terminated_length": 931.0,
"completions/mean_length": 729.5,
"completions/mean_terminated_length": 729.5,
"completions/min_length": 573.0,
"completions/min_terminated_length": 573.0,
"epoch": 0.04691665185711747,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1973987817764282,
"kl": 0.7220532447099686,
"learning_rate": 3.4682539682539686e-06,
"loss": 0.0007,
"num_tokens": 779970.0,
"reward": 0.8442820310592651,
"reward_std": 0.10566776990890503,
"rewards/check_answer/mean": 0.8442820310592651,
"rewards/check_answer/std": 0.10566778481006622,
"step": 264
},
{
"completion_length": 417.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 596.0,
"completions/max_terminated_length": 596.0,
"completions/mean_length": 417.5,
"completions/mean_terminated_length": 417.5,
"completions/min_length": 335.0,
"completions/min_terminated_length": 335.0,
"epoch": 0.04709436644748534,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7977854013442993,
"kl": 1.008522853255272,
"learning_rate": 3.4603174603174607e-06,
"loss": 0.001,
"num_tokens": 782792.0,
"reward": 0.4289383888244629,
"reward_std": 0.09566648304462433,
"rewards/check_answer/mean": 0.4289383888244629,
"rewards/check_answer/std": 0.09566648304462433,
"step": 265
},
{
"completion_length": 324.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 376.0,
"completions/max_terminated_length": 376.0,
"completions/mean_length": 324.5,
"completions/mean_terminated_length": 324.5,
"completions/min_length": 201.0,
"completions/min_terminated_length": 201.0,
"epoch": 0.04727208103785321,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0535054206848145,
"kl": 1.2962539196014404,
"learning_rate": 3.4523809523809528e-06,
"loss": 0.0013,
"num_tokens": 785202.0,
"reward": 0.7310268878936768,
"reward_std": 0.12941241264343262,
"rewards/check_answer/mean": 0.7310268878936768,
"rewards/check_answer/std": 0.12941241264343262,
"step": 266
},
{
"completion_length": 397.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 506.0,
"completions/max_terminated_length": 506.0,
"completions/mean_length": 397.0,
"completions/mean_terminated_length": 397.0,
"completions/min_length": 305.0,
"completions/min_terminated_length": 305.0,
"epoch": 0.047449795628221075,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6865134239196777,
"kl": 1.0350788980722427,
"learning_rate": 3.444444444444445e-06,
"loss": 0.001,
"num_tokens": 787926.0,
"reward": 0.6125245094299316,
"reward_std": 0.042323432862758636,
"rewards/check_answer/mean": 0.6125245094299316,
"rewards/check_answer/std": 0.04232344776391983,
"step": 267
},
{
"completion_length": 452.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 496.0,
"completions/max_terminated_length": 496.0,
"completions/mean_length": 452.0,
"completions/mean_terminated_length": 452.0,
"completions/min_length": 372.0,
"completions/min_terminated_length": 372.0,
"epoch": 0.04762751021858894,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8140619993209839,
"kl": 1.1835395395755768,
"learning_rate": 3.436507936507937e-06,
"loss": 0.0012,
"num_tokens": 790898.0,
"reward": 0.7142941951751709,
"reward_std": 0.04723852872848511,
"rewards/check_answer/mean": 0.7142941951751709,
"rewards/check_answer/std": 0.04723851755261421,
"step": 268
},
{
"completion_length": 423.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 485.0,
"completions/max_terminated_length": 485.0,
"completions/mean_length": 423.0,
"completions/mean_terminated_length": 423.0,
"completions/min_length": 310.0,
"completions/min_terminated_length": 310.0,
"epoch": 0.04780522480895682,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.790672779083252,
"kl": 0.9827256351709366,
"learning_rate": 3.428571428571429e-06,
"loss": 0.001,
"num_tokens": 793762.0,
"reward": 0.8307985067367554,
"reward_std": 0.3221786618232727,
"rewards/check_answer/mean": 0.8307985067367554,
"rewards/check_answer/std": 0.3221786618232727,
"step": 269
},
{
"completion_length": 358.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 426.0,
"completions/max_terminated_length": 426.0,
"completions/mean_length": 358.25,
"completions/mean_terminated_length": 358.25,
"completions/min_length": 307.0,
"completions/min_terminated_length": 307.0,
"epoch": 0.04798293939932469,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.941380500793457,
"kl": 1.0881590843200684,
"learning_rate": 3.420634920634921e-06,
"loss": 0.0011,
"num_tokens": 796307.0,
"reward": 0.4849153161048889,
"reward_std": 0.29613256454467773,
"rewards/check_answer/mean": 0.4849153161048889,
"rewards/check_answer/std": 0.2961325943470001,
"step": 270
},
{
"completion_length": 326.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 385.0,
"completions/max_terminated_length": 385.0,
"completions/mean_length": 326.0,
"completions/mean_terminated_length": 326.0,
"completions/min_length": 215.0,
"completions/min_terminated_length": 215.0,
"epoch": 0.048160653989692555,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9679803848266602,
"kl": 1.3071230351924896,
"learning_rate": 3.412698412698413e-06,
"loss": 0.0013,
"num_tokens": 798719.0,
"reward": 0.22918394207954407,
"reward_std": 0.062074821442365646,
"rewards/check_answer/mean": 0.22918394207954407,
"rewards/check_answer/std": 0.062074825167655945,
"step": 271
},
{
"completion_length": 398.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 448.0,
"completions/max_terminated_length": 448.0,
"completions/mean_length": 398.0,
"completions/mean_terminated_length": 398.0,
"completions/min_length": 349.0,
"completions/min_terminated_length": 349.0,
"epoch": 0.04833836858006042,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.018653392791748,
"kl": 1.0570762902498245,
"learning_rate": 3.404761904761905e-06,
"loss": 0.0011,
"num_tokens": 801451.0,
"reward": 0.6168738603591919,
"reward_std": 0.2640104293823242,
"rewards/check_answer/mean": 0.6168738603591919,
"rewards/check_answer/std": 0.26401039958000183,
"step": 272
},
{
"completion_length": 683.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 757.0,
"completions/max_terminated_length": 757.0,
"completions/mean_length": 683.75,
"completions/mean_terminated_length": 683.75,
"completions/min_length": 593.0,
"completions/min_terminated_length": 593.0,
"epoch": 0.04851608317042829,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.255545973777771,
"kl": 0.656520888209343,
"learning_rate": 3.3968253968253972e-06,
"loss": 0.0007,
"num_tokens": 805418.0,
"reward": 0.600953221321106,
"reward_std": 0.053118567913770676,
"rewards/check_answer/mean": 0.600953221321106,
"rewards/check_answer/std": 0.053118567913770676,
"step": 273
},
{
"completion_length": 410.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 464.0,
"completions/max_terminated_length": 464.0,
"completions/mean_length": 410.25,
"completions/mean_terminated_length": 410.25,
"completions/min_length": 371.0,
"completions/min_terminated_length": 371.0,
"epoch": 0.04869379776079616,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6982554197311401,
"kl": 1.035290315747261,
"learning_rate": 3.3888888888888893e-06,
"loss": 0.001,
"num_tokens": 808167.0,
"reward": 0.4320813715457916,
"reward_std": 0.07345610857009888,
"rewards/check_answer/mean": 0.4320813715457916,
"rewards/check_answer/std": 0.07345611602067947,
"step": 274
},
{
"completion_length": 362.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 449.0,
"completions/max_terminated_length": 449.0,
"completions/mean_length": 362.25,
"completions/mean_terminated_length": 362.25,
"completions/min_length": 277.0,
"completions/min_terminated_length": 277.0,
"epoch": 0.04887151235116403,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.048283100128174,
"kl": 1.2046183943748474,
"learning_rate": 3.3809523809523814e-06,
"loss": 0.0012,
"num_tokens": 810744.0,
"reward": 0.3853052854537964,
"reward_std": 0.17090635001659393,
"rewards/check_answer/mean": 0.3853052854537964,
"rewards/check_answer/std": 0.17090635001659393,
"step": 275
},
{
"completion_length": 305.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 355.0,
"completions/max_terminated_length": 355.0,
"completions/mean_length": 305.25,
"completions/mean_terminated_length": 305.25,
"completions/min_length": 209.0,
"completions/min_terminated_length": 209.0,
"epoch": 0.0490492269415319,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6985806226730347,
"kl": 1.415896326303482,
"learning_rate": 3.3730158730158734e-06,
"loss": 0.0014,
"num_tokens": 813081.0,
"reward": 0.8497557640075684,
"reward_std": 0.18619711697101593,
"rewards/check_answer/mean": 0.8497557640075684,
"rewards/check_answer/std": 0.18619710206985474,
"step": 276
},
{
"completion_length": 553.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 627.0,
"completions/max_terminated_length": 627.0,
"completions/mean_length": 553.25,
"completions/mean_terminated_length": 553.25,
"completions/min_length": 489.0,
"completions/min_terminated_length": 489.0,
"epoch": 0.04922694153189977,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2501318454742432,
"kl": 0.8245724886655807,
"learning_rate": 3.3650793650793655e-06,
"loss": 0.0008,
"num_tokens": 816462.0,
"reward": 0.7029004096984863,
"reward_std": 0.06054462492465973,
"rewards/check_answer/mean": 0.7029004096984863,
"rewards/check_answer/std": 0.060544636100530624,
"step": 277
},
{
"completion_length": 511.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 577.0,
"completions/max_terminated_length": 577.0,
"completions/mean_length": 511.0,
"completions/mean_terminated_length": 511.0,
"completions/min_length": 341.0,
"completions/min_terminated_length": 341.0,
"epoch": 0.04940465612226764,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5694273710250854,
"kl": 0.9416662156581879,
"learning_rate": 3.357142857142857e-06,
"loss": 0.0009,
"num_tokens": 819678.0,
"reward": 0.6121678948402405,
"reward_std": 0.11947009712457657,
"rewards/check_answer/mean": 0.6121678948402405,
"rewards/check_answer/std": 0.11947011202573776,
"step": 278
},
{
"completion_length": 588.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 637.0,
"completions/max_terminated_length": 637.0,
"completions/mean_length": 588.25,
"completions/mean_terminated_length": 588.25,
"completions/min_length": 566.0,
"completions/min_terminated_length": 566.0,
"epoch": 0.04958237071263551,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.327501893043518,
"kl": 0.7060510069131851,
"learning_rate": 3.349206349206349e-06,
"loss": 0.0007,
"num_tokens": 823227.0,
"reward": 0.8100916147232056,
"reward_std": 0.054197296500205994,
"rewards/check_answer/mean": 0.8100916147232056,
"rewards/check_answer/std": 0.0541972815990448,
"step": 279
},
{
"completion_length": 620.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 661.0,
"completions/max_terminated_length": 661.0,
"completions/mean_length": 620.25,
"completions/mean_terminated_length": 620.25,
"completions/min_length": 580.0,
"completions/min_terminated_length": 580.0,
"epoch": 0.049760085303003375,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3387950658798218,
"kl": 0.7631412148475647,
"learning_rate": 3.3412698412698413e-06,
"loss": 0.0008,
"num_tokens": 826916.0,
"reward": 0.5462270975112915,
"reward_std": 0.060185160487890244,
"rewards/check_answer/mean": 0.5462270975112915,
"rewards/check_answer/std": 0.06018515303730965,
"step": 280
},
{
"completion_length": 428.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 487.0,
"completions/max_terminated_length": 487.0,
"completions/mean_length": 428.75,
"completions/mean_terminated_length": 428.75,
"completions/min_length": 378.0,
"completions/min_terminated_length": 378.0,
"epoch": 0.04993779989337124,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5694767236709595,
"kl": 1.0430829524993896,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.001,
"num_tokens": 829767.0,
"reward": 1.1746463775634766,
"reward_std": 0.15259021520614624,
"rewards/check_answer/mean": 1.1746463775634766,
"rewards/check_answer/std": 0.15259020030498505,
"step": 281
},
{
"completion_length": 363.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 405.0,
"completions/max_terminated_length": 405.0,
"completions/mean_length": 363.0,
"completions/mean_terminated_length": 363.0,
"completions/min_length": 301.0,
"completions/min_terminated_length": 301.0,
"epoch": 0.05011551448373912,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8097895383834839,
"kl": 1.2064782232046127,
"learning_rate": 3.3253968253968254e-06,
"loss": 0.0012,
"num_tokens": 832327.0,
"reward": 0.6995459794998169,
"reward_std": 0.19229966402053833,
"rewards/check_answer/mean": 0.6995459794998169,
"rewards/check_answer/std": 0.19229967892169952,
"step": 282
},
{
"completion_length": 564.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 601.0,
"completions/max_terminated_length": 601.0,
"completions/mean_length": 564.0,
"completions/mean_terminated_length": 564.0,
"completions/min_length": 499.0,
"completions/min_terminated_length": 499.0,
"epoch": 0.050293229074106986,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3872801065444946,
"kl": 0.9052465409040451,
"learning_rate": 3.3174603174603175e-06,
"loss": 0.0009,
"num_tokens": 835767.0,
"reward": 0.8402754068374634,
"reward_std": 0.1533222496509552,
"rewards/check_answer/mean": 0.8402754068374634,
"rewards/check_answer/std": 0.1533222496509552,
"step": 283
},
{
"completion_length": 567.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 637.0,
"completions/max_terminated_length": 637.0,
"completions/mean_length": 567.25,
"completions/mean_terminated_length": 567.25,
"completions/min_length": 483.0,
"completions/min_terminated_length": 483.0,
"epoch": 0.050470943664474854,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4436535835266113,
"kl": 0.6898186206817627,
"learning_rate": 3.3095238095238095e-06,
"loss": 0.0007,
"num_tokens": 839224.0,
"reward": 0.372622549533844,
"reward_std": 0.02591540291905403,
"rewards/check_answer/mean": 0.372622549533844,
"rewards/check_answer/std": 0.02591540291905403,
"step": 284
},
{
"completion_length": 371.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 427.0,
"completions/max_terminated_length": 427.0,
"completions/mean_length": 371.75,
"completions/mean_terminated_length": 371.75,
"completions/min_length": 307.0,
"completions/min_terminated_length": 307.0,
"epoch": 0.05064865825484272,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.879496693611145,
"kl": 1.3733893036842346,
"learning_rate": 3.3015873015873016e-06,
"loss": 0.0014,
"num_tokens": 841827.0,
"reward": 0.5645827651023865,
"reward_std": 0.12172205001115799,
"rewards/check_answer/mean": 0.5645827651023865,
"rewards/check_answer/std": 0.1217220351099968,
"step": 285
},
{
"completion_length": 399.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 494.0,
"completions/max_terminated_length": 494.0,
"completions/mean_length": 399.5,
"completions/mean_terminated_length": 399.5,
"completions/min_length": 210.0,
"completions/min_terminated_length": 210.0,
"epoch": 0.05082637284521059,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6859583854675293,
"kl": 1.535703033208847,
"learning_rate": 3.293650793650794e-06,
"loss": 0.0015,
"num_tokens": 844553.0,
"reward": 1.0446977615356445,
"reward_std": 0.1805015206336975,
"rewards/check_answer/mean": 1.0446977615356445,
"rewards/check_answer/std": 0.1805015206336975,
"step": 286
},
{
"completion_length": 450.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 504.0,
"completions/max_terminated_length": 504.0,
"completions/mean_length": 450.5,
"completions/mean_terminated_length": 450.5,
"completions/min_length": 394.0,
"completions/min_terminated_length": 394.0,
"epoch": 0.05100408743557846,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5410058498382568,
"kl": 0.8857669532299042,
"learning_rate": 3.285714285714286e-06,
"loss": 0.0009,
"num_tokens": 847511.0,
"reward": 0.3833571672439575,
"reward_std": 0.0725344717502594,
"rewards/check_answer/mean": 0.3833571672439575,
"rewards/check_answer/std": 0.0725344717502594,
"step": 287
},
{
"completion_length": 525.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 545.0,
"completions/max_terminated_length": 545.0,
"completions/mean_length": 525.5,
"completions/mean_terminated_length": 525.5,
"completions/min_length": 507.0,
"completions/min_terminated_length": 507.0,
"epoch": 0.05118180202594633,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.534903645515442,
"kl": 0.9601296484470367,
"learning_rate": 3.277777777777778e-06,
"loss": 0.001,
"num_tokens": 850789.0,
"reward": 0.9623134732246399,
"reward_std": 0.11412428319454193,
"rewards/check_answer/mean": 0.9623134732246399,
"rewards/check_answer/std": 0.11412427574396133,
"step": 288
},
{
"completion_length": 501.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 532.0,
"completions/max_terminated_length": 532.0,
"completions/mean_length": 501.0,
"completions/mean_terminated_length": 501.0,
"completions/min_length": 464.0,
"completions/min_terminated_length": 464.0,
"epoch": 0.0513595166163142,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.619199275970459,
"kl": 1.0911777317523956,
"learning_rate": 3.2698412698412703e-06,
"loss": 0.0011,
"num_tokens": 853957.0,
"reward": 0.3744562864303589,
"reward_std": 0.029305126518011093,
"rewards/check_answer/mean": 0.3744562864303589,
"rewards/check_answer/std": 0.029305126518011093,
"step": 289
},
{
"completion_length": 615.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 705.0,
"completions/max_terminated_length": 705.0,
"completions/mean_length": 615.5,
"completions/mean_terminated_length": 615.5,
"completions/min_length": 537.0,
"completions/min_terminated_length": 537.0,
"epoch": 0.05153723120668207,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3343628644943237,
"kl": 0.6739598959684372,
"learning_rate": 3.2619047619047623e-06,
"loss": 0.0007,
"num_tokens": 857651.0,
"reward": 1.0102436542510986,
"reward_std": 0.10770290344953537,
"rewards/check_answer/mean": 1.0102436542510986,
"rewards/check_answer/std": 0.10770291090011597,
"step": 290
},
{
"completion_length": 428.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 513.0,
"completions/max_terminated_length": 513.0,
"completions/mean_length": 428.75,
"completions/mean_terminated_length": 428.75,
"completions/min_length": 345.0,
"completions/min_terminated_length": 345.0,
"epoch": 0.05171494579704994,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6610342264175415,
"kl": 1.076801985502243,
"learning_rate": 3.2539682539682544e-06,
"loss": 0.0011,
"num_tokens": 860514.0,
"reward": 0.9736640453338623,
"reward_std": 0.10310596972703934,
"rewards/check_answer/mean": 0.9736640453338623,
"rewards/check_answer/std": 0.10310596972703934,
"step": 291
},
{
"completion_length": 526.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 622.0,
"completions/max_terminated_length": 622.0,
"completions/mean_length": 526.0,
"completions/mean_terminated_length": 526.0,
"completions/min_length": 438.0,
"completions/min_terminated_length": 438.0,
"epoch": 0.051892660387417806,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3932909965515137,
"kl": 0.8913145214319229,
"learning_rate": 3.2460317460317465e-06,
"loss": 0.0009,
"num_tokens": 863774.0,
"reward": 0.6207559704780579,
"reward_std": 0.0433323010802269,
"rewards/check_answer/mean": 0.6207559704780579,
"rewards/check_answer/std": 0.04333232715725899,
"step": 292
},
{
"completion_length": 550.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 607.0,
"completions/max_terminated_length": 607.0,
"completions/mean_length": 550.0,
"completions/mean_terminated_length": 550.0,
"completions/min_length": 488.0,
"completions/min_terminated_length": 488.0,
"epoch": 0.052070374977785674,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4882313013076782,
"kl": 0.8690508455038071,
"learning_rate": 3.2380952380952385e-06,
"loss": 0.0009,
"num_tokens": 867162.0,
"reward": 0.3711192011833191,
"reward_std": 0.09456292539834976,
"rewards/check_answer/mean": 0.3711192011833191,
"rewards/check_answer/std": 0.09456291794776917,
"step": 293
},
{
"completion_length": 310.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 377.0,
"completions/max_terminated_length": 377.0,
"completions/mean_length": 310.5,
"completions/mean_terminated_length": 310.5,
"completions/min_length": 241.0,
"completions/min_terminated_length": 241.0,
"epoch": 0.05224808956815354,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9652611017227173,
"kl": 2.249081164598465,
"learning_rate": 3.2301587301587306e-06,
"loss": 0.0022,
"num_tokens": 869524.0,
"reward": 0.4758301377296448,
"reward_std": 0.16247986257076263,
"rewards/check_answer/mean": 0.4758301377296448,
"rewards/check_answer/std": 0.16247986257076263,
"step": 294
},
{
"completion_length": 476.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 581.0,
"completions/max_terminated_length": 581.0,
"completions/mean_length": 476.0,
"completions/mean_terminated_length": 476.0,
"completions/min_length": 418.0,
"completions/min_terminated_length": 418.0,
"epoch": 0.05242580415852142,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3406497240066528,
"kl": 0.8857537657022476,
"learning_rate": 3.2222222222222227e-06,
"loss": 0.0009,
"num_tokens": 872568.0,
"reward": 0.6054399013519287,
"reward_std": 0.11192188411951065,
"rewards/check_answer/mean": 0.6054399013519287,
"rewards/check_answer/std": 0.11192188411951065,
"step": 295
},
{
"completion_length": 510.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 547.0,
"completions/max_terminated_length": 547.0,
"completions/mean_length": 510.75,
"completions/mean_terminated_length": 510.75,
"completions/min_length": 476.0,
"completions/min_terminated_length": 476.0,
"epoch": 0.052603518748889286,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6864010095596313,
"kl": 1.1410781145095825,
"learning_rate": 3.2142857142857147e-06,
"loss": 0.0011,
"num_tokens": 875763.0,
"reward": 0.5221328735351562,
"reward_std": 0.12563703954219818,
"rewards/check_answer/mean": 0.5221328735351562,
"rewards/check_answer/std": 0.12563705444335938,
"step": 296
},
{
"completion_length": 376.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 542.0,
"completions/max_terminated_length": 542.0,
"completions/mean_length": 376.25,
"completions/mean_terminated_length": 376.25,
"completions/min_length": 249.0,
"completions/min_terminated_length": 249.0,
"epoch": 0.052781233339257154,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.528670310974121,
"kl": 1.0501147210597992,
"learning_rate": 3.206349206349207e-06,
"loss": 0.0011,
"num_tokens": 878412.0,
"reward": 0.9965839385986328,
"reward_std": 0.08146949857473373,
"rewards/check_answer/mean": 0.9965839385986328,
"rewards/check_answer/std": 0.08146951347589493,
"step": 297
},
{
"completion_length": 502.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 584.0,
"completions/max_terminated_length": 584.0,
"completions/mean_length": 502.75,
"completions/mean_terminated_length": 502.75,
"completions/min_length": 399.0,
"completions/min_terminated_length": 399.0,
"epoch": 0.05295894792962502,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4674580097198486,
"kl": 0.8372730612754822,
"learning_rate": 3.1984126984126984e-06,
"loss": 0.0008,
"num_tokens": 881591.0,
"reward": 0.4154725670814514,
"reward_std": 0.09623111039400101,
"rewards/check_answer/mean": 0.4154725670814514,
"rewards/check_answer/std": 0.0962311252951622,
"step": 298
},
{
"completion_length": 787.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 877.0,
"completions/max_terminated_length": 877.0,
"completions/mean_length": 787.75,
"completions/mean_terminated_length": 787.75,
"completions/min_length": 755.0,
"completions/min_terminated_length": 755.0,
"epoch": 0.05313666251999289,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.0730735063552856,
"kl": 0.6790334284305573,
"learning_rate": 3.1904761904761905e-06,
"loss": 0.0007,
"num_tokens": 885982.0,
"reward": 0.7593076229095459,
"reward_std": 0.01759275607764721,
"rewards/check_answer/mean": 0.7593076229095459,
"rewards/check_answer/std": 0.017592772841453552,
"step": 299
},
{
"completion_length": 235.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 280.0,
"completions/max_terminated_length": 280.0,
"completions/mean_length": 235.5,
"completions/mean_terminated_length": 235.5,
"completions/min_length": 212.0,
"completions/min_terminated_length": 212.0,
"epoch": 0.05331437711036076,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.377175807952881,
"kl": 1.5857573747634888,
"learning_rate": 3.1825396825396826e-06,
"loss": 0.0016,
"num_tokens": 888016.0,
"reward": 0.1881876289844513,
"reward_std": 0.01520258467644453,
"rewards/check_answer/mean": 0.1881876289844513,
"rewards/check_answer/std": 0.015202581882476807,
"step": 300
},
{
"completion_length": 577.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 659.0,
"completions/max_terminated_length": 659.0,
"completions/mean_length": 577.5,
"completions/mean_terminated_length": 577.5,
"completions/min_length": 440.0,
"completions/min_terminated_length": 440.0,
"epoch": 0.053492091700728626,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3808391094207764,
"kl": 0.893027126789093,
"learning_rate": 3.1746031746031746e-06,
"loss": 0.0009,
"num_tokens": 891490.0,
"reward": 0.6694183945655823,
"reward_std": 0.14587241411209106,
"rewards/check_answer/mean": 0.6694183945655823,
"rewards/check_answer/std": 0.14587242901325226,
"step": 301
},
{
"completion_length": 638.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 711.0,
"completions/max_terminated_length": 711.0,
"completions/mean_length": 638.0,
"completions/mean_terminated_length": 638.0,
"completions/min_length": 533.0,
"completions/min_terminated_length": 533.0,
"epoch": 0.0536698062910965,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2870622873306274,
"kl": 0.876729428768158,
"learning_rate": 3.1666666666666667e-06,
"loss": 0.0009,
"num_tokens": 895242.0,
"reward": 0.8270393013954163,
"reward_std": 0.07924500852823257,
"rewards/check_answer/mean": 0.8270393013954163,
"rewards/check_answer/std": 0.07924499362707138,
"step": 302
},
{
"completion_length": 597.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 624.0,
"completions/max_terminated_length": 624.0,
"completions/mean_length": 597.0,
"completions/mean_terminated_length": 597.0,
"completions/min_length": 555.0,
"completions/min_terminated_length": 555.0,
"epoch": 0.05384752088146437,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.332850694656372,
"kl": 0.7819689065217972,
"learning_rate": 3.1587301587301588e-06,
"loss": 0.0008,
"num_tokens": 898822.0,
"reward": 0.515304684638977,
"reward_std": 0.06797716021537781,
"rewards/check_answer/mean": 0.515304684638977,
"rewards/check_answer/std": 0.06797716021537781,
"step": 303
},
{
"completion_length": 539.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 551.0,
"completions/max_terminated_length": 551.0,
"completions/mean_length": 539.25,
"completions/mean_terminated_length": 539.25,
"completions/min_length": 517.0,
"completions/min_terminated_length": 517.0,
"epoch": 0.05402523547183224,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3149205446243286,
"kl": 0.7560958564281464,
"learning_rate": 3.150793650793651e-06,
"loss": 0.0008,
"num_tokens": 902183.0,
"reward": 0.7622121572494507,
"reward_std": 0.14100457727909088,
"rewards/check_answer/mean": 0.7622121572494507,
"rewards/check_answer/std": 0.14100460708141327,
"step": 304
},
{
"completion_length": 611.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 677.0,
"completions/max_terminated_length": 677.0,
"completions/mean_length": 611.5,
"completions/mean_terminated_length": 611.5,
"completions/min_length": 548.0,
"completions/min_terminated_length": 548.0,
"epoch": 0.054202950062200106,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2738730907440186,
"kl": 0.7394523322582245,
"learning_rate": 3.142857142857143e-06,
"loss": 0.0007,
"num_tokens": 905865.0,
"reward": 0.9216010570526123,
"reward_std": 0.15905454754829407,
"rewards/check_answer/mean": 0.9216010570526123,
"rewards/check_answer/std": 0.15905457735061646,
"step": 305
},
{
"completion_length": 576.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 693.0,
"completions/max_terminated_length": 693.0,
"completions/mean_length": 576.25,
"completions/mean_terminated_length": 576.25,
"completions/min_length": 375.0,
"completions/min_terminated_length": 375.0,
"epoch": 0.054380664652567974,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4615154266357422,
"kl": 0.8034448176622391,
"learning_rate": 3.134920634920635e-06,
"loss": 0.0008,
"num_tokens": 909398.0,
"reward": 0.6731904149055481,
"reward_std": 0.10627910494804382,
"rewards/check_answer/mean": 0.6731904149055481,
"rewards/check_answer/std": 0.10627911239862442,
"step": 306
},
{
"completion_length": 289.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 318.0,
"completions/max_terminated_length": 318.0,
"completions/mean_length": 289.25,
"completions/mean_terminated_length": 289.25,
"completions/min_length": 273.0,
"completions/min_terminated_length": 273.0,
"epoch": 0.05455837924293584,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.1155736446380615,
"kl": 1.6435502469539642,
"learning_rate": 3.126984126984127e-06,
"loss": 0.0016,
"num_tokens": 911655.0,
"reward": 1.1246044635772705,
"reward_std": 0.39093536138534546,
"rewards/check_answer/mean": 1.1246044635772705,
"rewards/check_answer/std": 0.39093536138534546,
"step": 307
},
{
"completion_length": 434.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 524.0,
"completions/max_terminated_length": 524.0,
"completions/mean_length": 434.0,
"completions/mean_terminated_length": 434.0,
"completions/min_length": 315.0,
"completions/min_terminated_length": 315.0,
"epoch": 0.05473609383330372,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6189072132110596,
"kl": 1.117174044251442,
"learning_rate": 3.1190476190476195e-06,
"loss": 0.0011,
"num_tokens": 914515.0,
"reward": 0.6435281038284302,
"reward_std": 0.17675022780895233,
"rewards/check_answer/mean": 0.6435281038284302,
"rewards/check_answer/std": 0.17675024271011353,
"step": 308
},
{
"completion_length": 328.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 430.0,
"completions/max_terminated_length": 430.0,
"completions/mean_length": 328.0,
"completions/mean_terminated_length": 328.0,
"completions/min_length": 281.0,
"completions/min_terminated_length": 281.0,
"epoch": 0.054913808423671585,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.084550380706787,
"kl": 1.2165972590446472,
"learning_rate": 3.1111111111111116e-06,
"loss": 0.0012,
"num_tokens": 916951.0,
"reward": 0.33441269397735596,
"reward_std": 0.0777365118265152,
"rewards/check_answer/mean": 0.33441269397735596,
"rewards/check_answer/std": 0.0777365043759346,
"step": 309
},
{
"completion_length": 450.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 568.0,
"completions/max_terminated_length": 568.0,
"completions/mean_length": 450.75,
"completions/mean_terminated_length": 450.75,
"completions/min_length": 283.0,
"completions/min_terminated_length": 283.0,
"epoch": 0.055091523014039454,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5784587860107422,
"kl": 1.0402155667543411,
"learning_rate": 3.1031746031746036e-06,
"loss": 0.001,
"num_tokens": 919910.0,
"reward": 0.29483291506767273,
"reward_std": 0.32212281227111816,
"rewards/check_answer/mean": 0.29483291506767273,
"rewards/check_answer/std": 0.32212281227111816,
"step": 310
},
{
"completion_length": 428.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 452.0,
"completions/max_terminated_length": 452.0,
"completions/mean_length": 428.0,
"completions/mean_terminated_length": 428.0,
"completions/min_length": 392.0,
"completions/min_terminated_length": 392.0,
"epoch": 0.05526923760440732,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.693036675453186,
"kl": 0.9303410053253174,
"learning_rate": 3.0952380952380957e-06,
"loss": 0.0009,
"num_tokens": 922766.0,
"reward": 0.5544993877410889,
"reward_std": 0.06014230474829674,
"rewards/check_answer/mean": 0.5544993877410889,
"rewards/check_answer/std": 0.060142308473587036,
"step": 311
},
{
"completion_length": 584.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 624.0,
"completions/max_terminated_length": 624.0,
"completions/mean_length": 584.25,
"completions/mean_terminated_length": 584.25,
"completions/min_length": 511.0,
"completions/min_terminated_length": 511.0,
"epoch": 0.05544695219477519,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1301647424697876,
"kl": 0.8142163455486298,
"learning_rate": 3.0873015873015878e-06,
"loss": 0.0008,
"num_tokens": 926339.0,
"reward": 0.5951245427131653,
"reward_std": 0.08697230368852615,
"rewards/check_answer/mean": 0.5951245427131653,
"rewards/check_answer/std": 0.08697231858968735,
"step": 312
},
{
"completion_length": 525.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 562.0,
"completions/max_terminated_length": 562.0,
"completions/mean_length": 525.75,
"completions/mean_terminated_length": 525.75,
"completions/min_length": 430.0,
"completions/min_terminated_length": 430.0,
"epoch": 0.05562466678514306,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6353572607040405,
"kl": 0.8869746625423431,
"learning_rate": 3.07936507936508e-06,
"loss": 0.0009,
"num_tokens": 929606.0,
"reward": 0.534943163394928,
"reward_std": 0.09478342533111572,
"rewards/check_answer/mean": 0.534943163394928,
"rewards/check_answer/std": 0.09478344023227692,
"step": 313
},
{
"completion_length": 526.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 632.0,
"completions/max_terminated_length": 632.0,
"completions/mean_length": 526.5,
"completions/mean_terminated_length": 526.5,
"completions/min_length": 387.0,
"completions/min_terminated_length": 387.0,
"epoch": 0.055802381375510926,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2205736637115479,
"kl": 0.8127477318048477,
"learning_rate": 3.071428571428572e-06,
"loss": 0.0008,
"num_tokens": 932924.0,
"reward": 0.8197857141494751,
"reward_std": 0.06929592788219452,
"rewards/check_answer/mean": 0.8197857141494751,
"rewards/check_answer/std": 0.06929592788219452,
"step": 314
},
{
"completion_length": 300.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 366.0,
"completions/max_terminated_length": 366.0,
"completions/mean_length": 300.25,
"completions/mean_terminated_length": 300.25,
"completions/min_length": 244.0,
"completions/min_terminated_length": 244.0,
"epoch": 0.0559800959658788,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8808090686798096,
"kl": 1.1378408521413803,
"learning_rate": 3.063492063492064e-06,
"loss": 0.0011,
"num_tokens": 935261.0,
"reward": 0.6404327154159546,
"reward_std": 0.09276391565799713,
"rewards/check_answer/mean": 0.6404327154159546,
"rewards/check_answer/std": 0.09276392310857773,
"step": 315
},
{
"completion_length": 489.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 611.0,
"completions/max_terminated_length": 611.0,
"completions/mean_length": 489.25,
"completions/mean_terminated_length": 489.25,
"completions/min_length": 427.0,
"completions/min_terminated_length": 427.0,
"epoch": 0.05615781055624667,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6700539588928223,
"kl": 1.1479647308588028,
"learning_rate": 3.055555555555556e-06,
"loss": 0.0011,
"num_tokens": 938350.0,
"reward": 0.3595621585845947,
"reward_std": 0.045500971376895905,
"rewards/check_answer/mean": 0.3595621585845947,
"rewards/check_answer/std": 0.045500971376895905,
"step": 316
},
{
"completion_length": 603.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 698.0,
"completions/max_terminated_length": 698.0,
"completions/mean_length": 603.5,
"completions/mean_terminated_length": 603.5,
"completions/min_length": 467.0,
"completions/min_terminated_length": 467.0,
"epoch": 0.05633552514661454,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1136270761489868,
"kl": 0.8240077495574951,
"learning_rate": 3.047619047619048e-06,
"loss": 0.0008,
"num_tokens": 941984.0,
"reward": 0.6625612378120422,
"reward_std": 0.088188037276268,
"rewards/check_answer/mean": 0.6625612378120422,
"rewards/check_answer/std": 0.08818802982568741,
"step": 317
},
{
"completion_length": 625.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 688.0,
"completions/max_terminated_length": 688.0,
"completions/mean_length": 625.25,
"completions/mean_terminated_length": 625.25,
"completions/min_length": 553.0,
"completions/min_terminated_length": 553.0,
"epoch": 0.056513239736982405,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1915112733840942,
"kl": 0.6459822058677673,
"learning_rate": 3.0396825396825397e-06,
"loss": 0.0006,
"num_tokens": 945721.0,
"reward": 0.6700223684310913,
"reward_std": 0.042193882167339325,
"rewards/check_answer/mean": 0.6700223684310913,
"rewards/check_answer/std": 0.04219387471675873,
"step": 318
},
{
"completion_length": 387.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 414.0,
"completions/max_terminated_length": 414.0,
"completions/mean_length": 387.75,
"completions/mean_terminated_length": 387.75,
"completions/min_length": 332.0,
"completions/min_terminated_length": 332.0,
"epoch": 0.056690954327350274,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0014514923095703,
"kl": 1.993764191865921,
"learning_rate": 3.031746031746032e-06,
"loss": 0.002,
"num_tokens": 948400.0,
"reward": 0.6778929233551025,
"reward_std": 0.08552777767181396,
"rewards/check_answer/mean": 0.6778929233551025,
"rewards/check_answer/std": 0.08552776277065277,
"step": 319
},
{
"completion_length": 314.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 351.0,
"completions/max_terminated_length": 351.0,
"completions/mean_length": 314.25,
"completions/mean_terminated_length": 314.25,
"completions/min_length": 301.0,
"completions/min_terminated_length": 301.0,
"epoch": 0.05686866891771814,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8478176593780518,
"kl": 1.2670190036296844,
"learning_rate": 3.023809523809524e-06,
"loss": 0.0013,
"num_tokens": 950769.0,
"reward": 0.9119844436645508,
"reward_std": 0.3170468211174011,
"rewards/check_answer/mean": 0.9119844436645508,
"rewards/check_answer/std": 0.3170468807220459,
"step": 320
},
{
"completion_length": 522.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 629.0,
"completions/max_terminated_length": 629.0,
"completions/mean_length": 522.25,
"completions/mean_terminated_length": 522.25,
"completions/min_length": 403.0,
"completions/min_terminated_length": 403.0,
"epoch": 0.05704638350808602,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6379070281982422,
"kl": 0.9933887124061584,
"learning_rate": 3.015873015873016e-06,
"loss": 0.001,
"num_tokens": 954010.0,
"reward": 0.41864892840385437,
"reward_std": 0.18404975533485413,
"rewards/check_answer/mean": 0.41864892840385437,
"rewards/check_answer/std": 0.18404975533485413,
"step": 321
},
{
"completion_length": 539.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 624.0,
"completions/max_terminated_length": 624.0,
"completions/mean_length": 539.0,
"completions/mean_terminated_length": 539.0,
"completions/min_length": 466.0,
"completions/min_terminated_length": 466.0,
"epoch": 0.057224098098453885,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3833321332931519,
"kl": 0.9609391242265701,
"learning_rate": 3.007936507936508e-06,
"loss": 0.001,
"num_tokens": 957386.0,
"reward": 0.9422582387924194,
"reward_std": 0.21418391168117523,
"rewards/check_answer/mean": 0.9422582387924194,
"rewards/check_answer/std": 0.21418389678001404,
"step": 322
},
{
"completion_length": 376.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 447.0,
"completions/max_terminated_length": 447.0,
"completions/mean_length": 376.75,
"completions/mean_terminated_length": 376.75,
"completions/min_length": 261.0,
"completions/min_terminated_length": 261.0,
"epoch": 0.05740181268882175,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7104628086090088,
"kl": 1.065696358680725,
"learning_rate": 3e-06,
"loss": 0.0011,
"num_tokens": 960033.0,
"reward": 0.8748215436935425,
"reward_std": 0.06767245382070541,
"rewards/check_answer/mean": 0.8748215436935425,
"rewards/check_answer/std": 0.06767246872186661,
"step": 323
},
{
"completion_length": 251.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 319.0,
"completions/max_terminated_length": 319.0,
"completions/mean_length": 251.0,
"completions/mean_terminated_length": 251.0,
"completions/min_length": 177.0,
"completions/min_terminated_length": 177.0,
"epoch": 0.05757952727918962,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.016751289367676,
"kl": 1.2200357019901276,
"learning_rate": 2.992063492063492e-06,
"loss": 0.0012,
"num_tokens": 962141.0,
"reward": 0.9851218461990356,
"reward_std": 0.5946353673934937,
"rewards/check_answer/mean": 0.9851218461990356,
"rewards/check_answer/std": 0.5946353673934937,
"step": 324
},
{
"completion_length": 499.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 596.0,
"completions/max_terminated_length": 596.0,
"completions/mean_length": 499.25,
"completions/mean_terminated_length": 499.25,
"completions/min_length": 433.0,
"completions/min_terminated_length": 433.0,
"epoch": 0.05775724186955749,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.48799467086792,
"kl": 0.8197063505649567,
"learning_rate": 2.984126984126984e-06,
"loss": 0.0008,
"num_tokens": 965306.0,
"reward": 0.5132365226745605,
"reward_std": 0.08968237042427063,
"rewards/check_answer/mean": 0.5132365226745605,
"rewards/check_answer/std": 0.08968238532543182,
"step": 325
},
{
"completion_length": 347.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 420.0,
"completions/max_terminated_length": 420.0,
"completions/mean_length": 347.25,
"completions/mean_terminated_length": 347.25,
"completions/min_length": 250.0,
"completions/min_terminated_length": 250.0,
"epoch": 0.05793495645992536,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0635764598846436,
"kl": 1.091228261590004,
"learning_rate": 2.9761904761904763e-06,
"loss": 0.0011,
"num_tokens": 967815.0,
"reward": 0.2954791486263275,
"reward_std": 0.09739544987678528,
"rewards/check_answer/mean": 0.2954791486263275,
"rewards/check_answer/std": 0.09739544987678528,
"step": 326
},
{
"completion_length": 586.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 683.0,
"completions/max_terminated_length": 683.0,
"completions/mean_length": 586.75,
"completions/mean_terminated_length": 586.75,
"completions/min_length": 445.0,
"completions/min_terminated_length": 445.0,
"epoch": 0.058112671050293226,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3021204471588135,
"kl": 0.7141899466514587,
"learning_rate": 2.9682539682539683e-06,
"loss": 0.0007,
"num_tokens": 971330.0,
"reward": 0.5484046936035156,
"reward_std": 0.08494994789361954,
"rewards/check_answer/mean": 0.5484046936035156,
"rewards/check_answer/std": 0.08494995534420013,
"step": 327
},
{
"completion_length": 618.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 637.0,
"completions/max_terminated_length": 637.0,
"completions/mean_length": 618.5,
"completions/mean_terminated_length": 618.5,
"completions/min_length": 591.0,
"completions/min_terminated_length": 591.0,
"epoch": 0.0582903856406611,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.261189341545105,
"kl": 0.7775250673294067,
"learning_rate": 2.9603174603174604e-06,
"loss": 0.0008,
"num_tokens": 975020.0,
"reward": 0.6435350775718689,
"reward_std": 0.07737845182418823,
"rewards/check_answer/mean": 0.6435350775718689,
"rewards/check_answer/std": 0.07737845182418823,
"step": 328
},
{
"completion_length": 318.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 399.0,
"completions/max_terminated_length": 399.0,
"completions/mean_length": 318.0,
"completions/mean_terminated_length": 318.0,
"completions/min_length": 216.0,
"completions/min_terminated_length": 216.0,
"epoch": 0.05846810023102897,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.718347191810608,
"kl": 0.993873119354248,
"learning_rate": 2.9523809523809525e-06,
"loss": 0.001,
"num_tokens": 977440.0,
"reward": 1.4608731269836426,
"reward_std": 0.07825374603271484,
"rewards/check_answer/mean": 1.4608731269836426,
"rewards/check_answer/std": 0.07825374603271484,
"step": 329
},
{
"completion_length": 290.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 319.0,
"completions/max_terminated_length": 319.0,
"completions/mean_length": 290.75,
"completions/mean_terminated_length": 290.75,
"completions/min_length": 240.0,
"completions/min_terminated_length": 240.0,
"epoch": 0.05864581482139684,
"frac_reward_zero_std": 1.0,
"grad_norm": 0.06520308554172516,
"kl": 1.8041629493236542,
"learning_rate": 2.944444444444445e-06,
"loss": 0.0018,
"num_tokens": 979699.0,
"reward": 0.5776078701019287,
"reward_std": 0.0,
"rewards/check_answer/mean": 0.5776078701019287,
"rewards/check_answer/std": 0.0,
"step": 330
},
{
"completion_length": 491.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 573.0,
"completions/max_terminated_length": 573.0,
"completions/mean_length": 491.5,
"completions/mean_terminated_length": 491.5,
"completions/min_length": 366.0,
"completions/min_terminated_length": 366.0,
"epoch": 0.058823529411764705,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6860709190368652,
"kl": 3.7682925313711166,
"learning_rate": 2.936507936507937e-06,
"loss": 0.0038,
"num_tokens": 982845.0,
"reward": 0.6522830724716187,
"reward_std": 0.05248499661684036,
"rewards/check_answer/mean": 0.6522830724716187,
"rewards/check_answer/std": 0.052484992891550064,
"step": 331
},
{
"completion_length": 365.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 423.0,
"completions/max_terminated_length": 423.0,
"completions/mean_length": 365.75,
"completions/mean_terminated_length": 365.75,
"completions/min_length": 298.0,
"completions/min_terminated_length": 298.0,
"epoch": 0.05900124400213257,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.921233892440796,
"kl": 1.0943335890769958,
"learning_rate": 2.928571428571429e-06,
"loss": 0.0011,
"num_tokens": 985436.0,
"reward": 0.9379060864448547,
"reward_std": 0.23881644010543823,
"rewards/check_answer/mean": 0.9379060864448547,
"rewards/check_answer/std": 0.23881642520427704,
"step": 332
},
{
"completion_length": 380.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 465.0,
"completions/max_terminated_length": 465.0,
"completions/mean_length": 380.5,
"completions/mean_terminated_length": 380.5,
"completions/min_length": 285.0,
"completions/min_terminated_length": 285.0,
"epoch": 0.05917895859250044,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3847863674163818,
"kl": 0.9714604318141937,
"learning_rate": 2.920634920634921e-06,
"loss": 0.001,
"num_tokens": 988110.0,
"reward": 0.865478515625,
"reward_std": 0.14856794476509094,
"rewards/check_answer/mean": 0.865478515625,
"rewards/check_answer/std": 0.14856792986392975,
"step": 333
},
{
"completion_length": 450.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 527.0,
"completions/max_terminated_length": 527.0,
"completions/mean_length": 450.0,
"completions/mean_terminated_length": 450.0,
"completions/min_length": 325.0,
"completions/min_terminated_length": 325.0,
"epoch": 0.059356673182868316,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4313362836837769,
"kl": 1.003603994846344,
"learning_rate": 2.9126984126984132e-06,
"loss": 0.001,
"num_tokens": 991110.0,
"reward": 1.0104808807373047,
"reward_std": 0.12709029018878937,
"rewards/check_answer/mean": 1.0104808807373047,
"rewards/check_answer/std": 0.12709030508995056,
"step": 334
},
{
"completion_length": 479.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 657.0,
"completions/max_terminated_length": 657.0,
"completions/mean_length": 479.25,
"completions/mean_terminated_length": 479.25,
"completions/min_length": 387.0,
"completions/min_terminated_length": 387.0,
"epoch": 0.059534387773236185,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5928345918655396,
"kl": 0.8456375598907471,
"learning_rate": 2.9047619047619053e-06,
"loss": 0.0008,
"num_tokens": 994223.0,
"reward": 0.6031970977783203,
"reward_std": 0.11164192855358124,
"rewards/check_answer/mean": 0.6031970977783203,
"rewards/check_answer/std": 0.11164193600416183,
"step": 335
},
{
"completion_length": 638.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 685.0,
"completions/max_terminated_length": 685.0,
"completions/mean_length": 638.0,
"completions/mean_terminated_length": 638.0,
"completions/min_length": 569.0,
"completions/min_terminated_length": 569.0,
"epoch": 0.05971210236360405,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3519030809402466,
"kl": 0.8861069083213806,
"learning_rate": 2.8968253968253974e-06,
"loss": 0.0009,
"num_tokens": 997995.0,
"reward": 0.587228000164032,
"reward_std": 0.04346587508916855,
"rewards/check_answer/mean": 0.587228000164032,
"rewards/check_answer/std": 0.043465860188007355,
"step": 336
},
{
"completion_length": 543.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 603.0,
"completions/max_terminated_length": 603.0,
"completions/mean_length": 543.75,
"completions/mean_terminated_length": 543.75,
"completions/min_length": 416.0,
"completions/min_terminated_length": 416.0,
"epoch": 0.05988981695397192,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4519314765930176,
"kl": 0.9474461674690247,
"learning_rate": 2.888888888888889e-06,
"loss": 0.0009,
"num_tokens": 1001334.0,
"reward": 0.4379974603652954,
"reward_std": 0.038793571293354034,
"rewards/check_answer/mean": 0.4379974603652954,
"rewards/check_answer/std": 0.038793571293354034,
"step": 337
},
{
"completion_length": 462.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 510.0,
"completions/max_terminated_length": 510.0,
"completions/mean_length": 462.0,
"completions/mean_terminated_length": 462.0,
"completions/min_length": 430.0,
"completions/min_terminated_length": 430.0,
"epoch": 0.06006753154433979,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6436101198196411,
"kl": 0.9353700429201126,
"learning_rate": 2.880952380952381e-06,
"loss": 0.0009,
"num_tokens": 1004330.0,
"reward": 0.7715719938278198,
"reward_std": 0.09180789440870285,
"rewards/check_answer/mean": 0.7715719938278198,
"rewards/check_answer/std": 0.09180789440870285,
"step": 338
},
{
"completion_length": 566.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 584.0,
"completions/max_terminated_length": 584.0,
"completions/mean_length": 566.25,
"completions/mean_terminated_length": 566.25,
"completions/min_length": 525.0,
"completions/min_terminated_length": 525.0,
"epoch": 0.06024524613470766,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5423588752746582,
"kl": 0.9218340069055557,
"learning_rate": 2.873015873015873e-06,
"loss": 0.0009,
"num_tokens": 1007771.0,
"reward": 0.6818769574165344,
"reward_std": 0.07576140016317368,
"rewards/check_answer/mean": 0.6818769574165344,
"rewards/check_answer/std": 0.07576142251491547,
"step": 339
},
{
"completion_length": 282.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 313.0,
"completions/max_terminated_length": 313.0,
"completions/mean_length": 282.25,
"completions/mean_terminated_length": 282.25,
"completions/min_length": 256.0,
"completions/min_terminated_length": 256.0,
"epoch": 0.06042296072507553,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.2459816932678223,
"kl": 1.3345037996768951,
"learning_rate": 2.865079365079365e-06,
"loss": 0.0013,
"num_tokens": 1009996.0,
"reward": 0.8702346086502075,
"reward_std": 0.32417604327201843,
"rewards/check_answer/mean": 0.8702346086502075,
"rewards/check_answer/std": 0.32417601346969604,
"step": 340
},
{
"completion_length": 390.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 503.0,
"completions/max_terminated_length": 503.0,
"completions/mean_length": 390.5,
"completions/mean_terminated_length": 390.5,
"completions/min_length": 276.0,
"completions/min_terminated_length": 276.0,
"epoch": 0.0606006753154434,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4166250228881836,
"kl": 1.2890968769788742,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.0013,
"num_tokens": 1012746.0,
"reward": 0.7851279973983765,
"reward_std": 0.10244892537593842,
"rewards/check_answer/mean": 0.7851279973983765,
"rewards/check_answer/std": 0.10244891792535782,
"step": 341
},
{
"completion_length": 459.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 510.0,
"completions/max_terminated_length": 510.0,
"completions/mean_length": 459.75,
"completions/mean_terminated_length": 459.75,
"completions/min_length": 402.0,
"completions/min_terminated_length": 402.0,
"epoch": 0.06077838990581127,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3638399839401245,
"kl": 1.0932938754558563,
"learning_rate": 2.8492063492063493e-06,
"loss": 0.0011,
"num_tokens": 1015749.0,
"reward": 0.7626172304153442,
"reward_std": 0.11766716837882996,
"rewards/check_answer/mean": 0.7626172304153442,
"rewards/check_answer/std": 0.11766716837882996,
"step": 342
},
{
"completion_length": 524.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 598.0,
"completions/max_terminated_length": 598.0,
"completions/mean_length": 524.5,
"completions/mean_terminated_length": 524.5,
"completions/min_length": 341.0,
"completions/min_terminated_length": 341.0,
"epoch": 0.060956104496179137,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5445250272750854,
"kl": 0.9230055660009384,
"learning_rate": 2.8412698412698414e-06,
"loss": 0.0009,
"num_tokens": 1019039.0,
"reward": 0.6102451086044312,
"reward_std": 0.03914730250835419,
"rewards/check_answer/mean": 0.6102451086044312,
"rewards/check_answer/std": 0.03914730250835419,
"step": 343
},
{
"completion_length": 360.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 459.0,
"completions/max_terminated_length": 459.0,
"completions/mean_length": 360.5,
"completions/mean_terminated_length": 360.5,
"completions/min_length": 273.0,
"completions/min_terminated_length": 273.0,
"epoch": 0.061133819086547005,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7623164653778076,
"kl": 2.2060094326734543,
"learning_rate": 2.8333333333333335e-06,
"loss": 0.0022,
"num_tokens": 1021613.0,
"reward": 0.5367588996887207,
"reward_std": 0.09137478470802307,
"rewards/check_answer/mean": 0.5367588996887207,
"rewards/check_answer/std": 0.09137478470802307,
"step": 344
},
{
"completion_length": 386.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 481.0,
"completions/max_terminated_length": 481.0,
"completions/mean_length": 386.75,
"completions/mean_terminated_length": 386.75,
"completions/min_length": 254.0,
"completions/min_terminated_length": 254.0,
"epoch": 0.06131153367691487,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6934525966644287,
"kl": 0.9983743131160736,
"learning_rate": 2.8253968253968255e-06,
"loss": 0.001,
"num_tokens": 1024292.0,
"reward": 0.8309897184371948,
"reward_std": 0.049490053206682205,
"rewards/check_answer/mean": 0.8309897184371948,
"rewards/check_answer/std": 0.049490053206682205,
"step": 345
},
{
"completion_length": 679.0,
"completions/clipped_ratio": 0.25,
"completions/max_length": 1733.0,
"completions/max_terminated_length": 495.0,
"completions/mean_length": 679.0,
"completions/mean_terminated_length": 327.66668701171875,
"completions/min_length": 240.0,
"completions/min_terminated_length": 240.0,
"epoch": 0.06148924826728274,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2086621522903442,
"kl": 0.9801141023635864,
"learning_rate": 2.8174603174603176e-06,
"loss": 0.001,
"num_tokens": 1028160.0,
"reward": 0.40882742404937744,
"reward_std": 0.5656044483184814,
"rewards/check_answer/mean": 0.40882742404937744,
"rewards/check_answer/std": 0.5656045079231262,
"step": 346
},
{
"completion_length": 499.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 670.0,
"completions/max_terminated_length": 670.0,
"completions/mean_length": 499.5,
"completions/mean_terminated_length": 499.5,
"completions/min_length": 398.0,
"completions/min_terminated_length": 398.0,
"epoch": 0.061666962857650616,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5193108320236206,
"kl": 0.8783832341432571,
"learning_rate": 2.8095238095238096e-06,
"loss": 0.0009,
"num_tokens": 1031374.0,
"reward": 0.5393548011779785,
"reward_std": 0.18399730324745178,
"rewards/check_answer/mean": 0.5393548011779785,
"rewards/check_answer/std": 0.18399730324745178,
"step": 347
},
{
"completion_length": 426.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 537.0,
"completions/max_terminated_length": 537.0,
"completions/mean_length": 426.0,
"completions/mean_terminated_length": 426.0,
"completions/min_length": 353.0,
"completions/min_terminated_length": 353.0,
"epoch": 0.061844677448018484,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5595829486846924,
"kl": 2.0859084129333496,
"learning_rate": 2.8015873015873017e-06,
"loss": 0.0021,
"num_tokens": 1034230.0,
"reward": 0.35466447472572327,
"reward_std": 0.007600174751132727,
"rewards/check_answer/mean": 0.35466447472572327,
"rewards/check_answer/std": 0.007600174751132727,
"step": 348
},
{
"completion_length": 242.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 296.0,
"completions/max_terminated_length": 296.0,
"completions/mean_length": 242.25,
"completions/mean_terminated_length": 242.25,
"completions/min_length": 168.0,
"completions/min_terminated_length": 168.0,
"epoch": 0.06202239203838635,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0255789756774902,
"kl": 1.20065575838089,
"learning_rate": 2.7936507936507938e-06,
"loss": 0.0012,
"num_tokens": 1036303.0,
"reward": 1.0878612995147705,
"reward_std": 0.5185619592666626,
"rewards/check_answer/mean": 1.0878612995147705,
"rewards/check_answer/std": 0.5185619592666626,
"step": 349
},
{
"completion_length": 493.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 607.0,
"completions/max_terminated_length": 607.0,
"completions/mean_length": 493.25,
"completions/mean_terminated_length": 493.25,
"completions/min_length": 254.0,
"completions/min_terminated_length": 254.0,
"epoch": 0.06220010662875422,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5794947147369385,
"kl": 0.7937883138656616,
"learning_rate": 2.785714285714286e-06,
"loss": 0.0008,
"num_tokens": 1039448.0,
"reward": 0.552399218082428,
"reward_std": 0.027781648561358452,
"rewards/check_answer/mean": 0.552399218082428,
"rewards/check_answer/std": 0.027781641110777855,
"step": 350
},
{
"completion_length": 412.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 506.0,
"completions/max_terminated_length": 506.0,
"completions/mean_length": 412.5,
"completions/mean_terminated_length": 412.5,
"completions/min_length": 317.0,
"completions/min_terminated_length": 317.0,
"epoch": 0.06237782121912209,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.685268759727478,
"kl": 1.08314049243927,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.0011,
"num_tokens": 1042238.0,
"reward": 0.6266283392906189,
"reward_std": 0.12399213761091232,
"rewards/check_answer/mean": 0.6266283392906189,
"rewards/check_answer/std": 0.12399215251207352,
"step": 351
},
{
"completion_length": 261.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 311.0,
"completions/max_terminated_length": 311.0,
"completions/mean_length": 261.75,
"completions/mean_terminated_length": 261.75,
"completions/min_length": 182.0,
"completions/min_terminated_length": 182.0,
"epoch": 0.06255553580948996,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.057827949523926,
"kl": 1.1714966595172882,
"learning_rate": 2.7698412698412704e-06,
"loss": 0.0012,
"num_tokens": 1044401.0,
"reward": 1.1655079126358032,
"reward_std": 0.1512351781129837,
"rewards/check_answer/mean": 1.1655079126358032,
"rewards/check_answer/std": 0.1512351632118225,
"step": 352
},
{
"completion_length": 454.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 590.0,
"completions/max_terminated_length": 590.0,
"completions/mean_length": 454.75,
"completions/mean_terminated_length": 454.75,
"completions/min_length": 349.0,
"completions/min_terminated_length": 349.0,
"epoch": 0.06273325039985783,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6537492275238037,
"kl": 0.836554616689682,
"learning_rate": 2.7619047619047625e-06,
"loss": 0.0008,
"num_tokens": 1047384.0,
"reward": 0.5927922129631042,
"reward_std": 0.03192928433418274,
"rewards/check_answer/mean": 0.5927922129631042,
"rewards/check_answer/std": 0.031929273158311844,
"step": 353
},
{
"completion_length": 449.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 516.0,
"completions/max_terminated_length": 516.0,
"completions/mean_length": 449.5,
"completions/mean_terminated_length": 449.5,
"completions/min_length": 358.0,
"completions/min_terminated_length": 358.0,
"epoch": 0.06291096499022569,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5732802152633667,
"kl": 0.9828883409500122,
"learning_rate": 2.7539682539682545e-06,
"loss": 0.001,
"num_tokens": 1050310.0,
"reward": 0.7062097787857056,
"reward_std": 0.07088971138000488,
"rewards/check_answer/mean": 0.7062097787857056,
"rewards/check_answer/std": 0.07088971138000488,
"step": 354
},
{
"completion_length": 416.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 466.0,
"completions/max_terminated_length": 466.0,
"completions/mean_length": 416.5,
"completions/mean_terminated_length": 416.5,
"completions/min_length": 377.0,
"completions/min_terminated_length": 377.0,
"epoch": 0.06308867958059357,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5273946523666382,
"kl": 0.7866203188896179,
"learning_rate": 2.7460317460317466e-06,
"loss": 0.0008,
"num_tokens": 1053128.0,
"reward": 0.5653597116470337,
"reward_std": 0.10099492967128754,
"rewards/check_answer/mean": 0.5653597116470337,
"rewards/check_answer/std": 0.10099492967128754,
"step": 355
},
{
"completion_length": 303.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 431.0,
"completions/max_terminated_length": 431.0,
"completions/mean_length": 303.75,
"completions/mean_terminated_length": 303.75,
"completions/min_length": 238.0,
"completions/min_terminated_length": 238.0,
"epoch": 0.06326639417096143,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6983771324157715,
"kl": 1.0617644041776657,
"learning_rate": 2.7380952380952387e-06,
"loss": 0.0011,
"num_tokens": 1055463.0,
"reward": 0.7611073851585388,
"reward_std": 0.21248966455459595,
"rewards/check_answer/mean": 0.7611073851585388,
"rewards/check_answer/std": 0.21248964965343475,
"step": 356
},
{
"completion_length": 429.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 554.0,
"completions/max_terminated_length": 554.0,
"completions/mean_length": 429.75,
"completions/mean_terminated_length": 429.75,
"completions/min_length": 357.0,
"completions/min_terminated_length": 357.0,
"epoch": 0.0634441087613293,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7513980865478516,
"kl": 0.9004791676998138,
"learning_rate": 2.7301587301587303e-06,
"loss": 0.0009,
"num_tokens": 1058338.0,
"reward": 0.7837266325950623,
"reward_std": 0.41704121232032776,
"rewards/check_answer/mean": 0.7837266325950623,
"rewards/check_answer/std": 0.41704121232032776,
"step": 357
},
{
"completion_length": 329.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 368.0,
"completions/max_terminated_length": 368.0,
"completions/mean_length": 329.75,
"completions/mean_terminated_length": 329.75,
"completions/min_length": 278.0,
"completions/min_terminated_length": 278.0,
"epoch": 0.06362182335169718,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.628482699394226,
"kl": 1.0090267211198807,
"learning_rate": 2.7222222222222224e-06,
"loss": 0.001,
"num_tokens": 1060765.0,
"reward": 0.7090367078781128,
"reward_std": 0.26245641708374023,
"rewards/check_answer/mean": 0.7090367078781128,
"rewards/check_answer/std": 0.26245638728141785,
"step": 358
},
{
"completion_length": 407.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 488.0,
"completions/max_terminated_length": 488.0,
"completions/mean_length": 407.5,
"completions/mean_terminated_length": 407.5,
"completions/min_length": 284.0,
"completions/min_terminated_length": 284.0,
"epoch": 0.06379953794206504,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.618558406829834,
"kl": 2.1619486063718796,
"learning_rate": 2.7142857142857144e-06,
"loss": 0.0022,
"num_tokens": 1063575.0,
"reward": 0.6832454800605774,
"reward_std": 0.13225074112415314,
"rewards/check_answer/mean": 0.6832454800605774,
"rewards/check_answer/std": 0.13225072622299194,
"step": 359
},
{
"completion_length": 356.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 409.0,
"completions/max_terminated_length": 409.0,
"completions/mean_length": 356.5,
"completions/mean_terminated_length": 356.5,
"completions/min_length": 289.0,
"completions/min_terminated_length": 289.0,
"epoch": 0.06397725253243292,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7556419372558594,
"kl": 0.9982209950685501,
"learning_rate": 2.7063492063492065e-06,
"loss": 0.001,
"num_tokens": 1066133.0,
"reward": 0.9417442083358765,
"reward_std": 0.0924437865614891,
"rewards/check_answer/mean": 0.9417442083358765,
"rewards/check_answer/std": 0.0924437940120697,
"step": 360
},
{
"completion_length": 438.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 506.0,
"completions/max_terminated_length": 506.0,
"completions/mean_length": 438.75,
"completions/mean_terminated_length": 438.75,
"completions/min_length": 383.0,
"completions/min_terminated_length": 383.0,
"epoch": 0.06415496712280078,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6464420557022095,
"kl": 1.1058107018470764,
"learning_rate": 2.6984126984126986e-06,
"loss": 0.0011,
"num_tokens": 1069028.0,
"reward": 0.5221630930900574,
"reward_std": 0.08845002204179764,
"rewards/check_answer/mean": 0.5221630930900574,
"rewards/check_answer/std": 0.08845002204179764,
"step": 361
},
{
"completion_length": 395.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 545.0,
"completions/max_terminated_length": 545.0,
"completions/mean_length": 395.25,
"completions/mean_terminated_length": 395.25,
"completions/min_length": 289.0,
"completions/min_terminated_length": 289.0,
"epoch": 0.06433268171316865,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5931658744812012,
"kl": 3.2120852321386337,
"learning_rate": 2.6904761904761906e-06,
"loss": 0.0032,
"num_tokens": 1071749.0,
"reward": 1.010395884513855,
"reward_std": 0.39355483651161194,
"rewards/check_answer/mean": 1.010395884513855,
"rewards/check_answer/std": 0.3935548663139343,
"step": 362
},
{
"completion_length": 366.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 516.0,
"completions/max_terminated_length": 516.0,
"completions/mean_length": 366.75,
"completions/mean_terminated_length": 366.75,
"completions/min_length": 257.0,
"completions/min_terminated_length": 257.0,
"epoch": 0.06451039630353653,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.682044506072998,
"kl": 1.1539453864097595,
"learning_rate": 2.6825396825396827e-06,
"loss": 0.0012,
"num_tokens": 1074356.0,
"reward": 0.6987335681915283,
"reward_std": 0.0868241935968399,
"rewards/check_answer/mean": 0.6987335681915283,
"rewards/check_answer/std": 0.0868242084980011,
"step": 363
},
{
"completion_length": 361.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 570.0,
"completions/max_terminated_length": 570.0,
"completions/mean_length": 361.75,
"completions/mean_terminated_length": 361.75,
"completions/min_length": 237.0,
"completions/min_terminated_length": 237.0,
"epoch": 0.06468811089390439,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6390739679336548,
"kl": 1.1847519278526306,
"learning_rate": 2.6746031746031748e-06,
"loss": 0.0012,
"num_tokens": 1076923.0,
"reward": 0.8292936682701111,
"reward_std": 0.20514844357967377,
"rewards/check_answer/mean": 0.8292936682701111,
"rewards/check_answer/std": 0.20514844357967377,
"step": 364
},
{
"completion_length": 292.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 390.0,
"completions/max_terminated_length": 390.0,
"completions/mean_length": 292.5,
"completions/mean_terminated_length": 292.5,
"completions/min_length": 226.0,
"completions/min_terminated_length": 226.0,
"epoch": 0.06486582548427226,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.981711745262146,
"kl": 1.2689296901226044,
"learning_rate": 2.666666666666667e-06,
"loss": 0.0013,
"num_tokens": 1079209.0,
"reward": 0.9906752705574036,
"reward_std": 0.468296080827713,
"rewards/check_answer/mean": 0.9906752705574036,
"rewards/check_answer/std": 0.468296080827713,
"step": 365
},
{
"completion_length": 304.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 353.0,
"completions/max_terminated_length": 353.0,
"completions/mean_length": 304.25,
"completions/mean_terminated_length": 304.25,
"completions/min_length": 236.0,
"completions/min_terminated_length": 236.0,
"epoch": 0.06504354007464012,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.9962329864501953,
"kl": 1.0153531730175018,
"learning_rate": 2.658730158730159e-06,
"loss": 0.001,
"num_tokens": 1081590.0,
"reward": 0.4718705713748932,
"reward_std": 0.1988232433795929,
"rewards/check_answer/mean": 0.4718705713748932,
"rewards/check_answer/std": 0.1988232433795929,
"step": 366
},
{
"completion_length": 586.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 696.0,
"completions/max_terminated_length": 696.0,
"completions/mean_length": 586.5,
"completions/mean_terminated_length": 586.5,
"completions/min_length": 508.0,
"completions/min_terminated_length": 508.0,
"epoch": 0.065221254665008,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2512035369873047,
"kl": 1.0086099058389664,
"learning_rate": 2.650793650793651e-06,
"loss": 0.001,
"num_tokens": 1085176.0,
"reward": 0.49589335918426514,
"reward_std": 0.02800540067255497,
"rewards/check_answer/mean": 0.49589335918426514,
"rewards/check_answer/std": 0.02800540253520012,
"step": 367
},
{
"completion_length": 653.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 811.0,
"completions/max_terminated_length": 811.0,
"completions/mean_length": 653.0,
"completions/mean_terminated_length": 653.0,
"completions/min_length": 457.0,
"completions/min_terminated_length": 457.0,
"epoch": 0.06539896925537586,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2003167867660522,
"kl": 0.7881456017494202,
"learning_rate": 2.642857142857143e-06,
"loss": 0.0008,
"num_tokens": 1089024.0,
"reward": 0.7394192218780518,
"reward_std": 0.09663909673690796,
"rewards/check_answer/mean": 0.7394192218780518,
"rewards/check_answer/std": 0.09663910418748856,
"step": 368
},
{
"completion_length": 456.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 541.0,
"completions/max_terminated_length": 541.0,
"completions/mean_length": 456.25,
"completions/mean_terminated_length": 456.25,
"completions/min_length": 340.0,
"completions/min_terminated_length": 340.0,
"epoch": 0.06557668384574374,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5003619194030762,
"kl": 0.8792458027601242,
"learning_rate": 2.634920634920635e-06,
"loss": 0.0009,
"num_tokens": 1092085.0,
"reward": 0.271815687417984,
"reward_std": 0.048961468040943146,
"rewards/check_answer/mean": 0.271815687417984,
"rewards/check_answer/std": 0.04896145686507225,
"step": 369
},
{
"completion_length": 241.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 281.0,
"completions/max_terminated_length": 281.0,
"completions/mean_length": 241.0,
"completions/mean_terminated_length": 241.0,
"completions/min_length": 185.0,
"completions/min_terminated_length": 185.0,
"epoch": 0.06575439843611161,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.5538017749786377,
"kl": 1.2450654804706573,
"learning_rate": 2.626984126984127e-06,
"loss": 0.0012,
"num_tokens": 1094145.0,
"reward": 1.040844202041626,
"reward_std": 0.6358775496482849,
"rewards/check_answer/mean": 1.040844202041626,
"rewards/check_answer/std": 0.6358775496482849,
"step": 370
},
{
"completion_length": 413.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 517.0,
"completions/max_terminated_length": 517.0,
"completions/mean_length": 413.75,
"completions/mean_terminated_length": 413.75,
"completions/min_length": 310.0,
"completions/min_terminated_length": 310.0,
"epoch": 0.06593211302647947,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.8476204872131348,
"kl": 1.6918334662914276,
"learning_rate": 2.6190476190476192e-06,
"loss": 0.0017,
"num_tokens": 1096956.0,
"reward": 1.0639305114746094,
"reward_std": 0.0493079274892807,
"rewards/check_answer/mean": 1.0639305114746094,
"rewards/check_answer/std": 0.049307931214571,
"step": 371
},
{
"completion_length": 253.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 295.0,
"completions/max_terminated_length": 295.0,
"completions/mean_length": 253.25,
"completions/mean_terminated_length": 253.25,
"completions/min_length": 205.0,
"completions/min_terminated_length": 205.0,
"epoch": 0.06610982761684735,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.1900248527526855,
"kl": 1.0722286999225616,
"learning_rate": 2.6111111111111113e-06,
"loss": 0.0011,
"num_tokens": 1099077.0,
"reward": 0.3582559823989868,
"reward_std": 0.037198010832071304,
"rewards/check_answer/mean": 0.3582559823989868,
"rewards/check_answer/std": 0.037198010832071304,
"step": 372
},
{
"completion_length": 339.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 406.0,
"completions/max_terminated_length": 406.0,
"completions/mean_length": 339.5,
"completions/mean_terminated_length": 339.5,
"completions/min_length": 286.0,
"completions/min_terminated_length": 286.0,
"epoch": 0.06628754220721521,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.416177749633789,
"kl": 0.759627029299736,
"learning_rate": 2.6031746031746038e-06,
"loss": 0.0008,
"num_tokens": 1101567.0,
"reward": 0.5204699039459229,
"reward_std": 0.20935100317001343,
"rewards/check_answer/mean": 0.5204699039459229,
"rewards/check_answer/std": 0.20935101807117462,
"step": 373
},
{
"completion_length": 581.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 717.0,
"completions/max_terminated_length": 717.0,
"completions/mean_length": 581.75,
"completions/mean_terminated_length": 581.75,
"completions/min_length": 361.0,
"completions/min_terminated_length": 361.0,
"epoch": 0.06646525679758308,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3829495906829834,
"kl": 0.7596141695976257,
"learning_rate": 2.595238095238096e-06,
"loss": 0.0008,
"num_tokens": 1105114.0,
"reward": 0.6244233846664429,
"reward_std": 0.1169346272945404,
"rewards/check_answer/mean": 0.6244233846664429,
"rewards/check_answer/std": 0.1169346421957016,
"step": 374
},
{
"completion_length": 553.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 615.0,
"completions/max_terminated_length": 615.0,
"completions/mean_length": 553.25,
"completions/mean_terminated_length": 553.25,
"completions/min_length": 532.0,
"completions/min_terminated_length": 532.0,
"epoch": 0.06664297138795094,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.049517273902893,
"kl": 0.731134757399559,
"learning_rate": 2.587301587301588e-06,
"loss": 0.0007,
"num_tokens": 1108555.0,
"reward": 0.9090848565101624,
"reward_std": 0.053088050335645676,
"rewards/check_answer/mean": 0.9090848565101624,
"rewards/check_answer/std": 0.05308803915977478,
"step": 375
},
{
"completion_length": 581.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 793.0,
"completions/max_terminated_length": 793.0,
"completions/mean_length": 581.5,
"completions/mean_terminated_length": 581.5,
"completions/min_length": 442.0,
"completions/min_terminated_length": 442.0,
"epoch": 0.06682068597831882,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.1811550855636597,
"kl": 0.7464872002601624,
"learning_rate": 2.57936507936508e-06,
"loss": 0.0007,
"num_tokens": 1112113.0,
"reward": 0.5068844556808472,
"reward_std": 0.08424738049507141,
"rewards/check_answer/mean": 0.5068844556808472,
"rewards/check_answer/std": 0.08424738049507141,
"step": 376
},
{
"completion_length": 456.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 531.0,
"completions/max_terminated_length": 531.0,
"completions/mean_length": 456.25,
"completions/mean_terminated_length": 456.25,
"completions/min_length": 331.0,
"completions/min_terminated_length": 331.0,
"epoch": 0.0669984005686867,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6065852642059326,
"kl": 1.0195069313049316,
"learning_rate": 2.571428571428571e-06,
"loss": 0.001,
"num_tokens": 1115090.0,
"reward": 0.9967861771583557,
"reward_std": 0.1993342638015747,
"rewards/check_answer/mean": 0.9967861771583557,
"rewards/check_answer/std": 0.1993342638015747,
"step": 377
},
{
"completion_length": 570.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 772.0,
"completions/max_terminated_length": 772.0,
"completions/mean_length": 570.25,
"completions/mean_terminated_length": 570.25,
"completions/min_length": 417.0,
"completions/min_terminated_length": 417.0,
"epoch": 0.06717611515905456,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.288891315460205,
"kl": 0.9126172959804535,
"learning_rate": 2.5634920634920637e-06,
"loss": 0.0009,
"num_tokens": 1118543.0,
"reward": 0.9744613170623779,
"reward_std": 0.10629359632730484,
"rewards/check_answer/mean": 0.9744613170623779,
"rewards/check_answer/std": 0.10629359632730484,
"step": 378
},
{
"completion_length": 584.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 758.0,
"completions/max_terminated_length": 758.0,
"completions/mean_length": 584.25,
"completions/mean_terminated_length": 584.25,
"completions/min_length": 336.0,
"completions/min_terminated_length": 336.0,
"epoch": 0.06735382974942243,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.518787145614624,
"kl": 1.7292539477348328,
"learning_rate": 2.5555555555555557e-06,
"loss": 0.0017,
"num_tokens": 1122052.0,
"reward": 0.9780303835868835,
"reward_std": 0.11002466827630997,
"rewards/check_answer/mean": 0.9780303835868835,
"rewards/check_answer/std": 0.11002466827630997,
"step": 379
},
{
"completion_length": 464.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 674.0,
"completions/max_terminated_length": 674.0,
"completions/mean_length": 464.75,
"completions/mean_terminated_length": 464.75,
"completions/min_length": 313.0,
"completions/min_terminated_length": 313.0,
"epoch": 0.06753154433979029,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.394750952720642,
"kl": 0.9494964927434921,
"learning_rate": 2.547619047619048e-06,
"loss": 0.0009,
"num_tokens": 1125083.0,
"reward": 0.49630308151245117,
"reward_std": 0.14733171463012695,
"rewards/check_answer/mean": 0.49630308151245117,
"rewards/check_answer/std": 0.14733172953128815,
"step": 380
},
{
"completion_length": 366.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 422.0,
"completions/max_terminated_length": 422.0,
"completions/mean_length": 366.75,
"completions/mean_terminated_length": 366.75,
"completions/min_length": 228.0,
"completions/min_terminated_length": 228.0,
"epoch": 0.06770925893015817,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5692752599716187,
"kl": 1.070998653769493,
"learning_rate": 2.53968253968254e-06,
"loss": 0.0011,
"num_tokens": 1127670.0,
"reward": 0.8239761590957642,
"reward_std": 0.22099705040454865,
"rewards/check_answer/mean": 0.8239761590957642,
"rewards/check_answer/std": 0.22099705040454865,
"step": 381
},
{
"completion_length": 559.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 689.0,
"completions/max_terminated_length": 689.0,
"completions/mean_length": 559.75,
"completions/mean_terminated_length": 559.75,
"completions/min_length": 372.0,
"completions/min_terminated_length": 372.0,
"epoch": 0.06788697352052603,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4717862606048584,
"kl": 0.7960640043020248,
"learning_rate": 2.531746031746032e-06,
"loss": 0.0008,
"num_tokens": 1131065.0,
"reward": 0.7636763453483582,
"reward_std": 0.03722745552659035,
"rewards/check_answer/mean": 0.7636763453483582,
"rewards/check_answer/std": 0.03722746670246124,
"step": 382
},
{
"completion_length": 446.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 593.0,
"completions/max_terminated_length": 593.0,
"completions/mean_length": 446.0,
"completions/mean_terminated_length": 446.0,
"completions/min_length": 374.0,
"completions/min_terminated_length": 374.0,
"epoch": 0.0680646881108939,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.5540060997009277,
"kl": 0.8313327878713608,
"learning_rate": 2.523809523809524e-06,
"loss": 0.0008,
"num_tokens": 1134045.0,
"reward": 0.6946367621421814,
"reward_std": 0.10725139826536179,
"rewards/check_answer/mean": 0.6946367621421814,
"rewards/check_answer/std": 0.10725140571594238,
"step": 383
},
{
"completion_length": 614.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 701.0,
"completions/max_terminated_length": 701.0,
"completions/mean_length": 614.0,
"completions/mean_terminated_length": 614.0,
"completions/min_length": 552.0,
"completions/min_terminated_length": 552.0,
"epoch": 0.06824240270126178,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.311665415763855,
"kl": 0.8095138221979141,
"learning_rate": 2.515873015873016e-06,
"loss": 0.0008,
"num_tokens": 1137665.0,
"reward": 0.4693688750267029,
"reward_std": 0.06689858436584473,
"rewards/check_answer/mean": 0.4693688750267029,
"rewards/check_answer/std": 0.06689856946468353,
"step": 384
},
{
"completion_length": 582.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 657.0,
"completions/max_terminated_length": 657.0,
"completions/mean_length": 582.25,
"completions/mean_terminated_length": 582.25,
"completions/min_length": 468.0,
"completions/min_terminated_length": 468.0,
"epoch": 0.06842011729162964,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.0779439210891724,
"kl": 0.6766816079616547,
"learning_rate": 2.507936507936508e-06,
"loss": 0.0007,
"num_tokens": 1141214.0,
"reward": 0.6747899055480957,
"reward_std": 0.07833781838417053,
"rewards/check_answer/mean": 0.6747899055480957,
"rewards/check_answer/std": 0.07833781838417053,
"step": 385
},
{
"completion_length": 403.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 469.0,
"completions/max_terminated_length": 469.0,
"completions/mean_length": 403.0,
"completions/mean_terminated_length": 403.0,
"completions/min_length": 289.0,
"completions/min_terminated_length": 289.0,
"epoch": 0.06859783188199751,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.6302034854888916,
"kl": 0.8939348012208939,
"learning_rate": 2.5e-06,
"loss": 0.0009,
"num_tokens": 1143954.0,
"reward": 0.3099154233932495,
"reward_std": 0.15409527719020844,
"rewards/check_answer/mean": 0.3099154233932495,
"rewards/check_answer/std": 0.15409527719020844,
"step": 386
},
{
"completion_length": 601.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 738.0,
"completions/max_terminated_length": 738.0,
"completions/mean_length": 601.75,
"completions/mean_terminated_length": 601.75,
"completions/min_length": 368.0,
"completions/min_terminated_length": 368.0,
"epoch": 0.06877554647236538,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.424817442893982,
"kl": 0.8362443447113037,
"learning_rate": 2.4920634920634923e-06,
"loss": 0.0008,
"num_tokens": 1147545.0,
"reward": 0.7676776647567749,
"reward_std": 0.20846693217754364,
"rewards/check_answer/mean": 0.7676776647567749,
"rewards/check_answer/std": 0.20846691727638245,
"step": 387
},
{
"completion_length": 360.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 424.0,
"completions/max_terminated_length": 424.0,
"completions/mean_length": 360.5,
"completions/mean_terminated_length": 360.5,
"completions/min_length": 276.0,
"completions/min_terminated_length": 276.0,
"epoch": 0.06895326106273325,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7432420253753662,
"kl": 1.0870371013879776,
"learning_rate": 2.4841269841269843e-06,
"loss": 0.0011,
"num_tokens": 1150095.0,
"reward": 0.5128135085105896,
"reward_std": 0.12844218313694,
"rewards/check_answer/mean": 0.5128135085105896,
"rewards/check_answer/std": 0.12844218313694,
"step": 388
},
{
"completion_length": 562.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 687.0,
"completions/max_terminated_length": 687.0,
"completions/mean_length": 562.25,
"completions/mean_terminated_length": 562.25,
"completions/min_length": 448.0,
"completions/min_terminated_length": 448.0,
"epoch": 0.06913097565310113,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3039882183074951,
"kl": 0.8053465038537979,
"learning_rate": 2.4761904761904764e-06,
"loss": 0.0008,
"num_tokens": 1153600.0,
"reward": 0.6226294040679932,
"reward_std": 0.10893689841032028,
"rewards/check_answer/mean": 0.6226294040679932,
"rewards/check_answer/std": 0.10893691331148148,
"step": 389
},
{
"completion_length": 392.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 545.0,
"completions/max_terminated_length": 545.0,
"completions/mean_length": 392.0,
"completions/mean_terminated_length": 392.0,
"completions/min_length": 334.0,
"completions/min_terminated_length": 334.0,
"epoch": 0.06930869024346899,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7045742273330688,
"kl": 0.9221175163984299,
"learning_rate": 2.4682539682539685e-06,
"loss": 0.0009,
"num_tokens": 1156316.0,
"reward": 0.5222982168197632,
"reward_std": 0.054217997938394547,
"rewards/check_answer/mean": 0.5222982168197632,
"rewards/check_answer/std": 0.054218001663684845,
"step": 390
},
{
"completion_length": 370.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 535.0,
"completions/max_terminated_length": 535.0,
"completions/mean_length": 370.75,
"completions/mean_terminated_length": 370.75,
"completions/min_length": 246.0,
"completions/min_terminated_length": 246.0,
"epoch": 0.06948640483383686,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7914966344833374,
"kl": 1.043500304222107,
"learning_rate": 2.4603174603174605e-06,
"loss": 0.001,
"num_tokens": 1158939.0,
"reward": 0.39007383584976196,
"reward_std": 0.06512384861707687,
"rewards/check_answer/mean": 0.39007383584976196,
"rewards/check_answer/std": 0.06512384861707687,
"step": 391
},
{
"completion_length": 403.25,
"completions/clipped_ratio": 0.0,
"completions/max_length": 500.0,
"completions/max_terminated_length": 500.0,
"completions/mean_length": 403.25,
"completions/mean_terminated_length": 403.25,
"completions/min_length": 302.0,
"completions/min_terminated_length": 302.0,
"epoch": 0.06966411942420472,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.3422685861587524,
"kl": 1.0109106451272964,
"learning_rate": 2.4523809523809526e-06,
"loss": 0.001,
"num_tokens": 1161676.0,
"reward": 0.7319890856742859,
"reward_std": 0.07778804004192352,
"rewards/check_answer/mean": 0.7319890856742859,
"rewards/check_answer/std": 0.07778804004192352,
"step": 392
},
{
"completion_length": 450.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 605.0,
"completions/max_terminated_length": 605.0,
"completions/mean_length": 450.0,
"completions/mean_terminated_length": 450.0,
"completions/min_length": 350.0,
"completions/min_terminated_length": 350.0,
"epoch": 0.0698418340145726,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.531283974647522,
"kl": 0.8451306223869324,
"learning_rate": 2.4444444444444447e-06,
"loss": 0.0008,
"num_tokens": 1164652.0,
"reward": 0.8435445427894592,
"reward_std": 0.14115361869335175,
"rewards/check_answer/mean": 0.8435445427894592,
"rewards/check_answer/std": 0.14115363359451294,
"step": 393
},
{
"completion_length": 396.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 510.0,
"completions/max_terminated_length": 510.0,
"completions/mean_length": 396.5,
"completions/mean_terminated_length": 396.5,
"completions/min_length": 213.0,
"completions/min_terminated_length": 213.0,
"epoch": 0.07001954860494046,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4916499853134155,
"kl": 0.9246963560581207,
"learning_rate": 2.4365079365079367e-06,
"loss": 0.0009,
"num_tokens": 1167410.0,
"reward": 0.6280086636543274,
"reward_std": 0.1793169528245926,
"rewards/check_answer/mean": 0.6280086636543274,
"rewards/check_answer/std": 0.1793169379234314,
"step": 394
},
{
"completion_length": 379.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 521.0,
"completions/max_terminated_length": 521.0,
"completions/mean_length": 379.0,
"completions/mean_terminated_length": 379.0,
"completions/min_length": 266.0,
"completions/min_terminated_length": 266.0,
"epoch": 0.07019726319530833,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.7005797624588013,
"kl": 1.0196750611066818,
"learning_rate": 2.428571428571429e-06,
"loss": 0.001,
"num_tokens": 1170094.0,
"reward": 0.8843287229537964,
"reward_std": 0.1384207308292389,
"rewards/check_answer/mean": 0.8843287229537964,
"rewards/check_answer/std": 0.13842076063156128,
"step": 395
},
{
"completion_length": 586.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 690.0,
"completions/max_terminated_length": 690.0,
"completions/mean_length": 586.75,
"completions/mean_terminated_length": 586.75,
"completions/min_length": 441.0,
"completions/min_terminated_length": 441.0,
"epoch": 0.07037497778567621,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.447705626487732,
"kl": 0.9538527727127075,
"learning_rate": 2.420634920634921e-06,
"loss": 0.001,
"num_tokens": 1173633.0,
"reward": 0.8358705043792725,
"reward_std": 0.10273723304271698,
"rewards/check_answer/mean": 0.8358705043792725,
"rewards/check_answer/std": 0.10273724794387817,
"step": 396
},
{
"completion_length": 242.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 313.0,
"completions/max_terminated_length": 313.0,
"completions/mean_length": 242.75,
"completions/mean_terminated_length": 242.75,
"completions/min_length": 185.0,
"completions/min_terminated_length": 185.0,
"epoch": 0.07055269237604407,
"frac_reward_zero_std": 0.0,
"grad_norm": 2.0574350357055664,
"kl": 1.2170641124248505,
"learning_rate": 2.412698412698413e-06,
"loss": 0.0012,
"num_tokens": 1175724.0,
"reward": 0.4298151135444641,
"reward_std": 0.20456048846244812,
"rewards/check_answer/mean": 0.4298151135444641,
"rewards/check_answer/std": 0.20456050336360931,
"step": 397
},
{
"completion_length": 594.0,
"completions/clipped_ratio": 0.0,
"completions/max_length": 826.0,
"completions/max_terminated_length": 826.0,
"completions/mean_length": 594.0,
"completions/mean_terminated_length": 594.0,
"completions/min_length": 389.0,
"completions/min_terminated_length": 389.0,
"epoch": 0.07073040696641195,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2806570529937744,
"kl": 0.8345987200737,
"learning_rate": 2.404761904761905e-06,
"loss": 0.0008,
"num_tokens": 1179312.0,
"reward": 0.6557444930076599,
"reward_std": 0.09894954413175583,
"rewards/check_answer/mean": 0.6557444930076599,
"rewards/check_answer/std": 0.09894953668117523,
"step": 398
},
{
"completion_length": 528.5,
"completions/clipped_ratio": 0.0,
"completions/max_length": 689.0,
"completions/max_terminated_length": 689.0,
"completions/mean_length": 528.5,
"completions/mean_terminated_length": 528.5,
"completions/min_length": 418.0,
"completions/min_terminated_length": 418.0,
"epoch": 0.07090812155677981,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.4865775108337402,
"kl": 0.9594889581203461,
"learning_rate": 2.396825396825397e-06,
"loss": 0.001,
"num_tokens": 1182602.0,
"reward": 0.5999321937561035,
"reward_std": 0.1533043533563614,
"rewards/check_answer/mean": 0.5999321937561035,
"rewards/check_answer/std": 0.1533043384552002,
"step": 399
},
{
"completion_length": 696.75,
"completions/clipped_ratio": 0.0,
"completions/max_length": 733.0,
"completions/max_terminated_length": 733.0,
"completions/mean_length": 696.75,
"completions/mean_terminated_length": 696.75,
"completions/min_length": 639.0,
"completions/min_terminated_length": 639.0,
"epoch": 0.07108583614714768,
"frac_reward_zero_std": 0.0,
"grad_norm": 1.2060387134552002,
"kl": 0.9064538776874542,
"learning_rate": 2.388888888888889e-06,
"loss": 0.0009,
"num_tokens": 1186613.0,
"reward": 0.5622410774230957,
"reward_std": 0.04867928475141525,
"rewards/check_answer/mean": 0.5622410774230957,
"rewards/check_answer/std": 0.04867929220199585,
"step": 400
}
],
"logging_steps": 1,
"max_steps": 700,
"num_input_tokens_seen": 1186613,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}