| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.08, |
| "eval_steps": 500, |
| "global_step": 800, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 382.0, |
| "completions/max_terminated_length": 382.0, |
| "completions/mean_length": 287.875, |
| "completions/mean_terminated_length": 287.875, |
| "completions/min_length": 203.0, |
| "completions/min_terminated_length": 203.0, |
| "entropy": 0.1642230749130249, |
| "epoch": 0.0001, |
| "frac_reward_zero_std": 0.0, |
| "grad_norm": 0.665309488773346, |
| "learning_rate": 1e-05, |
| "loss": -0.11556684970855713, |
| "num_tokens": 3951.0, |
| "reward": 0.125, |
| "reward_std": 0.3535533845424652, |
| "rewards/boxed_accuracy_reward/mean": 0.125, |
| "rewards/boxed_accuracy_reward/std": 0.3535533845424652, |
| "step": 1, |
| "step_time": 26.083515274000092 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 438.1111111111111, |
| "completions/max_terminated_length": 438.1111111111111, |
| "completions/mean_length": 253.38888888888889, |
| "completions/mean_terminated_length": 253.38888888888889, |
| "completions/min_length": 132.33333333333334, |
| "completions/min_terminated_length": 132.33333333333334, |
| "entropy": 0.19083112478256226, |
| "epoch": 0.001, |
| "frac_reward_zero_std": 0.3333333333333333, |
| "grad_norm": 0.7672234177589417, |
| "learning_rate": 9.91e-06, |
| "loss": 0.017603478497929044, |
| "num_tokens": 34851.0, |
| "reward": 0.1527777777777778, |
| "reward_std": 0.2903762327300178, |
| "rewards/boxed_accuracy_reward/mean": 0.1527777777777778, |
| "rewards/boxed_accuracy_reward/std": 0.2903762393527561, |
| "step": 10, |
| "step_time": 29.407547047666917 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0625, |
| "completions/max_length": 598.4, |
| "completions/max_terminated_length": 454.2, |
| "completions/mean_length": 326.7125, |
| "completions/mean_terminated_length": 281.9235778808594, |
| "completions/min_length": 157.3, |
| "completions/min_terminated_length": 157.3, |
| "entropy": 0.22966246232390403, |
| "epoch": 0.002, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 9.810000000000001e-06, |
| "loss": 0.015656116604804992, |
| "num_tokens": 74804.0, |
| "reward": 0.15, |
| "reward_std": 0.12416292428970337, |
| "rewards/boxed_accuracy_reward/mean": 0.15, |
| "rewards/boxed_accuracy_reward/std": 0.12416292428970337, |
| "step": 20, |
| "step_time": 40.079171687399864 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 471.4, |
| "completions/max_terminated_length": 418.5, |
| "completions/mean_length": 261.45, |
| "completions/mean_terminated_length": 251.90892944335937, |
| "completions/min_length": 161.7, |
| "completions/min_terminated_length": 161.7, |
| "entropy": 0.17057178877294063, |
| "epoch": 0.003, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 9.71e-06, |
| "loss": -0.01473381370306015, |
| "num_tokens": 109952.0, |
| "reward": 0.25, |
| "reward_std": 0.08711025416851044, |
| "rewards/boxed_accuracy_reward/mean": 0.25, |
| "rewards/boxed_accuracy_reward/std": 0.08711026012897491, |
| "step": 30, |
| "step_time": 31.344073937199575 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0375, |
| "completions/max_length": 462.8, |
| "completions/max_terminated_length": 416.3, |
| "completions/mean_length": 292.6375, |
| "completions/mean_terminated_length": 267.9024993896484, |
| "completions/min_length": 172.9, |
| "completions/min_terminated_length": 172.9, |
| "entropy": 0.22824809104204177, |
| "epoch": 0.004, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 9.610000000000001e-06, |
| "loss": 0.014347019791603088, |
| "num_tokens": 148715.0, |
| "reward": 0.1, |
| "reward_std": 0.09258201122283935, |
| "rewards/boxed_accuracy_reward/mean": 0.1, |
| "rewards/boxed_accuracy_reward/std": 0.09258201122283935, |
| "step": 40, |
| "step_time": 30.823996917499972 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 543.5, |
| "completions/max_terminated_length": 415.6, |
| "completions/mean_length": 269.2375, |
| "completions/mean_terminated_length": 249.4482162475586, |
| "completions/min_length": 160.6, |
| "completions/min_terminated_length": 160.6, |
| "entropy": 0.266183939576149, |
| "epoch": 0.005, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 0.7730127573013306, |
| "learning_rate": 9.51e-06, |
| "loss": 0.047160887718200685, |
| "num_tokens": 184662.0, |
| "reward": 0.1625, |
| "reward_std": 0.26849985122680664, |
| "rewards/boxed_accuracy_reward/mean": 0.1625, |
| "rewards/boxed_accuracy_reward/std": 0.2684998571872711, |
| "step": 50, |
| "step_time": 36.26071728879979 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 445.4, |
| "completions/max_terminated_length": 445.4, |
| "completions/mean_length": 271.05, |
| "completions/mean_terminated_length": 271.05, |
| "completions/min_length": 168.9, |
| "completions/min_terminated_length": 168.9, |
| "entropy": 0.18989864736795425, |
| "epoch": 0.006, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 9.41e-06, |
| "loss": 0.07637788653373719, |
| "num_tokens": 221874.0, |
| "reward": 0.25, |
| "reward_std": 0.2276727557182312, |
| "rewards/boxed_accuracy_reward/mean": 0.25, |
| "rewards/boxed_accuracy_reward/std": 0.22767276763916017, |
| "step": 60, |
| "step_time": 29.8493166889999 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 559.9, |
| "completions/max_terminated_length": 456.6, |
| "completions/mean_length": 279.3125, |
| "completions/mean_terminated_length": 262.7035736083984, |
| "completions/min_length": 157.0, |
| "completions/min_terminated_length": 157.0, |
| "entropy": 0.2535190790891647, |
| "epoch": 0.007, |
| "frac_reward_zero_std": 0.9, |
| "grad_norm": 0.0, |
| "learning_rate": 9.31e-06, |
| "loss": 0.007473611086606979, |
| "num_tokens": 257171.0, |
| "reward": 0.05, |
| "reward_std": 0.05345224738121033, |
| "rewards/boxed_accuracy_reward/mean": 0.05, |
| "rewards/boxed_accuracy_reward/std": 0.05345224738121033, |
| "step": 70, |
| "step_time": 37.422609474000275 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 506.5, |
| "completions/max_terminated_length": 375.0, |
| "completions/mean_length": 246.025, |
| "completions/mean_terminated_length": 226.50714569091798, |
| "completions/min_length": 146.6, |
| "completions/min_terminated_length": 146.6, |
| "entropy": 0.21952729374170304, |
| "epoch": 0.008, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 9.210000000000002e-06, |
| "loss": -0.004495994746685028, |
| "num_tokens": 291301.0, |
| "reward": 0.175, |
| "reward_std": 0.18685350716114044, |
| "rewards/boxed_accuracy_reward/mean": 0.175, |
| "rewards/boxed_accuracy_reward/std": 0.18685351312160492, |
| "step": 80, |
| "step_time": 33.85629228980069 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 417.1, |
| "completions/max_terminated_length": 417.1, |
| "completions/mean_length": 270.3375, |
| "completions/mean_terminated_length": 270.3375, |
| "completions/min_length": 177.8, |
| "completions/min_terminated_length": 177.8, |
| "entropy": 0.2391076095402241, |
| "epoch": 0.009, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.6646445393562317, |
| "learning_rate": 9.110000000000001e-06, |
| "loss": 0.0037924151867628096, |
| "num_tokens": 326552.0, |
| "reward": 0.1125, |
| "reward_std": 0.2095758467912674, |
| "rewards/boxed_accuracy_reward/mean": 0.1125, |
| "rewards/boxed_accuracy_reward/std": 0.20957585871219636, |
| "step": 90, |
| "step_time": 27.908505431900267 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0625, |
| "completions/max_length": 489.8, |
| "completions/max_terminated_length": 405.3, |
| "completions/mean_length": 305.5, |
| "completions/mean_terminated_length": 269.0066680908203, |
| "completions/min_length": 142.9, |
| "completions/min_terminated_length": 142.9, |
| "entropy": 0.2832915961742401, |
| "epoch": 0.01, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 9.01e-06, |
| "loss": 0.01493358165025711, |
| "num_tokens": 364440.0, |
| "reward": 0.2, |
| "reward_std": 0.17045392990112304, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.17045392990112304, |
| "step": 100, |
| "step_time": 32.493894719798845 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 454.8, |
| "completions/max_terminated_length": 377.0, |
| "completions/mean_length": 258.625, |
| "completions/mean_terminated_length": 248.3375015258789, |
| "completions/min_length": 170.1, |
| "completions/min_terminated_length": 170.1, |
| "entropy": 0.2617956634610891, |
| "epoch": 0.011, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.29700157046318054, |
| "learning_rate": 8.910000000000001e-06, |
| "loss": 0.03313029408454895, |
| "num_tokens": 398914.0, |
| "reward": 0.125, |
| "reward_std": 0.22598327100276946, |
| "rewards/boxed_accuracy_reward/mean": 0.125, |
| "rewards/boxed_accuracy_reward/std": 0.22598327696323395, |
| "step": 110, |
| "step_time": 30.349950915500084 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 457.5, |
| "completions/max_terminated_length": 389.1, |
| "completions/mean_length": 265.4875, |
| "completions/mean_terminated_length": 255.85714416503907, |
| "completions/min_length": 164.2, |
| "completions/min_terminated_length": 164.2, |
| "entropy": 0.2188379220664501, |
| "epoch": 0.012, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 8.81e-06, |
| "loss": 0.02148536443710327, |
| "num_tokens": 434921.0, |
| "reward": 0.0625, |
| "reward_std": 0.08880758583545685, |
| "rewards/boxed_accuracy_reward/mean": 0.0625, |
| "rewards/boxed_accuracy_reward/std": 0.08880758583545685, |
| "step": 120, |
| "step_time": 30.65767741319978 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 424.0, |
| "completions/max_terminated_length": 382.3, |
| "completions/mean_length": 237.0125, |
| "completions/mean_terminated_length": 228.27142944335938, |
| "completions/min_length": 139.3, |
| "completions/min_terminated_length": 139.3, |
| "entropy": 0.2461308181285858, |
| "epoch": 0.013, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 8.710000000000001e-06, |
| "loss": 0.013855263590812683, |
| "num_tokens": 468658.0, |
| "reward": 0.1, |
| "reward_std": 0.20411193668842315, |
| "rewards/boxed_accuracy_reward/mean": 0.1, |
| "rewards/boxed_accuracy_reward/std": 0.20411194264888763, |
| "step": 130, |
| "step_time": 28.27658463789994 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 416.1, |
| "completions/max_terminated_length": 357.2, |
| "completions/mean_length": 233.2625, |
| "completions/mean_terminated_length": 223.3732147216797, |
| "completions/min_length": 130.1, |
| "completions/min_terminated_length": 130.1, |
| "entropy": 0.29308837205171584, |
| "epoch": 0.014, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 8.61e-06, |
| "loss": 0.0346070647239685, |
| "num_tokens": 500159.0, |
| "reward": 0.1875, |
| "reward_std": 0.22768060266971588, |
| "rewards/boxed_accuracy_reward/mean": 0.1875, |
| "rewards/boxed_accuracy_reward/std": 0.22768060266971588, |
| "step": 140, |
| "step_time": 27.783400436999727 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 503.8, |
| "completions/max_terminated_length": 503.8, |
| "completions/mean_length": 295.05, |
| "completions/mean_terminated_length": 295.05, |
| "completions/min_length": 184.6, |
| "completions/min_terminated_length": 184.6, |
| "entropy": 0.20557359904050826, |
| "epoch": 0.015, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 8.51e-06, |
| "loss": 0.03165164291858673, |
| "num_tokens": 537435.0, |
| "reward": 0.1625, |
| "reward_std": 0.14433692693710326, |
| "rewards/boxed_accuracy_reward/mean": 0.1625, |
| "rewards/boxed_accuracy_reward/std": 0.14433693289756774, |
| "step": 150, |
| "step_time": 33.749151340000026 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 504.0, |
| "completions/max_terminated_length": 405.2, |
| "completions/mean_length": 280.05, |
| "completions/mean_terminated_length": 261.85714569091795, |
| "completions/min_length": 158.9, |
| "completions/min_terminated_length": 158.9, |
| "entropy": 0.2418084606528282, |
| "epoch": 0.016, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 0.9073280096054077, |
| "learning_rate": 8.41e-06, |
| "loss": 0.029856789112091064, |
| "num_tokens": 573351.0, |
| "reward": 0.2, |
| "reward_std": 0.2794355183839798, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.2794355243444443, |
| "step": 160, |
| "step_time": 36.18624399920045 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 484.3, |
| "completions/max_terminated_length": 484.3, |
| "completions/mean_length": 281.0, |
| "completions/mean_terminated_length": 281.0, |
| "completions/min_length": 167.8, |
| "completions/min_terminated_length": 167.8, |
| "entropy": 0.21754518747329712, |
| "epoch": 0.017, |
| "frac_reward_zero_std": 0.9, |
| "grad_norm": 0.0, |
| "learning_rate": 8.31e-06, |
| "loss": 0.01904902458190918, |
| "num_tokens": 610279.0, |
| "reward": 0.125, |
| "reward_std": 0.046291005611419675, |
| "rewards/boxed_accuracy_reward/mean": 0.125, |
| "rewards/boxed_accuracy_reward/std": 0.046291005611419675, |
| "step": 170, |
| "step_time": 35.24356591649958 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 472.6, |
| "completions/max_terminated_length": 421.2, |
| "completions/mean_length": 269.1625, |
| "completions/mean_terminated_length": 260.52142944335935, |
| "completions/min_length": 159.9, |
| "completions/min_terminated_length": 159.9, |
| "entropy": 0.22995990961790086, |
| "epoch": 0.018, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 8.210000000000001e-06, |
| "loss": -0.011320264637470245, |
| "num_tokens": 646660.0, |
| "reward": 0.0625, |
| "reward_std": 0.08880758583545685, |
| "rewards/boxed_accuracy_reward/mean": 0.0625, |
| "rewards/boxed_accuracy_reward/std": 0.08880758583545685, |
| "step": 180, |
| "step_time": 35.16620119949948 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0375, |
| "completions/max_length": 570.6, |
| "completions/max_terminated_length": 513.5, |
| "completions/mean_length": 297.2625, |
| "completions/mean_terminated_length": 272.977978515625, |
| "completions/min_length": 156.5, |
| "completions/min_terminated_length": 156.5, |
| "entropy": 0.24057478606700897, |
| "epoch": 0.019, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 1.0508952140808105, |
| "learning_rate": 8.110000000000002e-06, |
| "loss": 0.01611926555633545, |
| "num_tokens": 685673.0, |
| "reward": 0.1, |
| "reward_std": 0.1948736011981964, |
| "rewards/boxed_accuracy_reward/mean": 0.1, |
| "rewards/boxed_accuracy_reward/std": 0.1948736011981964, |
| "step": 190, |
| "step_time": 39.78827415060077 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 379.3, |
| "completions/max_terminated_length": 322.7, |
| "completions/mean_length": 236.0875, |
| "completions/mean_terminated_length": 227.3767883300781, |
| "completions/min_length": 150.7, |
| "completions/min_terminated_length": 150.7, |
| "entropy": 0.21252653896808624, |
| "epoch": 0.02, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 0.7992851138114929, |
| "learning_rate": 8.010000000000001e-06, |
| "loss": 0.02353036105632782, |
| "num_tokens": 719336.0, |
| "reward": 0.35, |
| "reward_std": 0.30299633741378784, |
| "rewards/boxed_accuracy_reward/mean": 0.35, |
| "rewards/boxed_accuracy_reward/std": 0.3029963493347168, |
| "step": 200, |
| "step_time": 27.117062961000194 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0875, |
| "completions/max_length": 609.6, |
| "completions/max_terminated_length": 431.3, |
| "completions/mean_length": 345.1875, |
| "completions/mean_terminated_length": 294.5657165527344, |
| "completions/min_length": 165.6, |
| "completions/min_terminated_length": 165.6, |
| "entropy": 0.17211427241563798, |
| "epoch": 0.021, |
| "frac_reward_zero_std": 0.9, |
| "grad_norm": 0.0, |
| "learning_rate": 7.91e-06, |
| "loss": -0.0028946617618203162, |
| "num_tokens": 759991.0, |
| "reward": 0.0125, |
| "reward_std": 0.03535533845424652, |
| "rewards/boxed_accuracy_reward/mean": 0.0125, |
| "rewards/boxed_accuracy_reward/std": 0.03535533845424652, |
| "step": 210, |
| "step_time": 44.109834954999315 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 562.9, |
| "completions/max_terminated_length": 454.9, |
| "completions/mean_length": 275.2125, |
| "completions/mean_terminated_length": 257.24286193847655, |
| "completions/min_length": 149.9, |
| "completions/min_terminated_length": 149.9, |
| "entropy": 0.2036581441760063, |
| "epoch": 0.022, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.9508808255195618, |
| "learning_rate": 7.810000000000001e-06, |
| "loss": 0.05190147757530213, |
| "num_tokens": 795232.0, |
| "reward": 0.2375, |
| "reward_std": 0.17422835528850555, |
| "rewards/boxed_accuracy_reward/mean": 0.2375, |
| "rewards/boxed_accuracy_reward/std": 0.17422835528850555, |
| "step": 220, |
| "step_time": 38.12656000370043 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 372.2, |
| "completions/max_terminated_length": 322.1, |
| "completions/mean_length": 216.725, |
| "completions/mean_terminated_length": 206.85714416503907, |
| "completions/min_length": 123.2, |
| "completions/min_terminated_length": 123.2, |
| "entropy": 0.24748676642775536, |
| "epoch": 0.023, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.825386643409729, |
| "learning_rate": 7.71e-06, |
| "loss": -0.019961801171302796, |
| "num_tokens": 828634.0, |
| "reward": 0.225, |
| "reward_std": 0.1632926881313324, |
| "rewards/boxed_accuracy_reward/mean": 0.225, |
| "rewards/boxed_accuracy_reward/std": 0.1632926881313324, |
| "step": 230, |
| "step_time": 24.96606878720049 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 373.3, |
| "completions/max_terminated_length": 373.3, |
| "completions/mean_length": 232.2, |
| "completions/mean_terminated_length": 232.2, |
| "completions/min_length": 150.1, |
| "completions/min_terminated_length": 150.1, |
| "entropy": 0.2603755287826061, |
| "epoch": 0.024, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 0.5379490256309509, |
| "learning_rate": 7.610000000000001e-06, |
| "loss": 0.03033584952354431, |
| "num_tokens": 861162.0, |
| "reward": 0.1875, |
| "reward_std": 0.2722664296627045, |
| "rewards/boxed_accuracy_reward/mean": 0.1875, |
| "rewards/boxed_accuracy_reward/std": 0.2722664475440979, |
| "step": 240, |
| "step_time": 24.956236593399808 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 532.6, |
| "completions/max_terminated_length": 465.1, |
| "completions/mean_length": 262.9375, |
| "completions/mean_terminated_length": 252.74464416503906, |
| "completions/min_length": 135.1, |
| "completions/min_terminated_length": 135.1, |
| "entropy": 0.2639094963669777, |
| "epoch": 0.025, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 7.510000000000001e-06, |
| "loss": 0.014231480658054352, |
| "num_tokens": 896165.0, |
| "reward": 0.075, |
| "reward_std": 0.08711025416851044, |
| "rewards/boxed_accuracy_reward/mean": 0.075, |
| "rewards/boxed_accuracy_reward/std": 0.08711026012897491, |
| "step": 250, |
| "step_time": 35.73487980219979 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 416.4, |
| "completions/max_terminated_length": 416.4, |
| "completions/mean_length": 263.975, |
| "completions/mean_terminated_length": 263.975, |
| "completions/min_length": 163.7, |
| "completions/min_terminated_length": 163.7, |
| "entropy": 0.249638444930315, |
| "epoch": 0.026, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 7.41e-06, |
| "loss": 0.01857263296842575, |
| "num_tokens": 930771.0, |
| "reward": 0.1625, |
| "reward_std": 0.15235702097415924, |
| "rewards/boxed_accuracy_reward/mean": 0.1625, |
| "rewards/boxed_accuracy_reward/std": 0.15235702097415924, |
| "step": 260, |
| "step_time": 28.139153176400214 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 422.7, |
| "completions/max_terminated_length": 422.7, |
| "completions/mean_length": 251.9875, |
| "completions/mean_terminated_length": 251.9875, |
| "completions/min_length": 153.3, |
| "completions/min_terminated_length": 153.3, |
| "entropy": 0.24129935801029206, |
| "epoch": 0.027, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 7.31e-06, |
| "loss": -0.032139381766319274, |
| "num_tokens": 963458.0, |
| "reward": 0.2125, |
| "reward_std": 0.14433692693710326, |
| "rewards/boxed_accuracy_reward/mean": 0.2125, |
| "rewards/boxed_accuracy_reward/std": 0.14433693289756774, |
| "step": 270, |
| "step_time": 28.544226764599806 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 332.8, |
| "completions/max_terminated_length": 332.8, |
| "completions/mean_length": 231.9125, |
| "completions/mean_terminated_length": 231.9125, |
| "completions/min_length": 148.2, |
| "completions/min_terminated_length": 148.2, |
| "entropy": 0.26727280877530574, |
| "epoch": 0.028, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.9024311900138855, |
| "learning_rate": 7.2100000000000004e-06, |
| "loss": 0.021772387623786926, |
| "num_tokens": 996579.0, |
| "reward": 0.15, |
| "reward_std": 0.17422050833702088, |
| "rewards/boxed_accuracy_reward/mean": 0.15, |
| "rewards/boxed_accuracy_reward/std": 0.17422052025794982, |
| "step": 280, |
| "step_time": 22.47074644069944 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0625, |
| "completions/max_length": 443.4, |
| "completions/max_terminated_length": 364.0, |
| "completions/mean_length": 265.1375, |
| "completions/mean_terminated_length": 211.40833435058593, |
| "completions/min_length": 125.6, |
| "completions/min_terminated_length": 125.6, |
| "entropy": 0.2202783677726984, |
| "epoch": 0.029, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.23354178667068481, |
| "learning_rate": 7.1100000000000005e-06, |
| "loss": 0.039704719185829164, |
| "num_tokens": 1033446.0, |
| "reward": 0.2375, |
| "reward_std": 0.2478467583656311, |
| "rewards/boxed_accuracy_reward/mean": 0.2375, |
| "rewards/boxed_accuracy_reward/std": 0.24784677624702453, |
| "step": 290, |
| "step_time": 29.699856723000266 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 403.5, |
| "completions/max_terminated_length": 403.5, |
| "completions/mean_length": 281.5125, |
| "completions/mean_terminated_length": 281.5125, |
| "completions/min_length": 162.4, |
| "completions/min_terminated_length": 162.4, |
| "entropy": 0.2921067249029875, |
| "epoch": 0.03, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 7.01e-06, |
| "loss": 0.006478501856327057, |
| "num_tokens": 1071503.0, |
| "reward": 0.2625, |
| "reward_std": 0.19864802658557892, |
| "rewards/boxed_accuracy_reward/mean": 0.2625, |
| "rewards/boxed_accuracy_reward/std": 0.19864802658557892, |
| "step": 300, |
| "step_time": 27.02418423309937 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 438.9, |
| "completions/max_terminated_length": 438.9, |
| "completions/mean_length": 247.8, |
| "completions/mean_terminated_length": 247.8, |
| "completions/min_length": 139.4, |
| "completions/min_terminated_length": 139.4, |
| "entropy": 0.22176421582698821, |
| "epoch": 0.031, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 1.1117596626281738, |
| "learning_rate": 6.91e-06, |
| "loss": -0.027645233273506164, |
| "num_tokens": 1109583.0, |
| "reward": 0.1625, |
| "reward_std": 0.26302809417247774, |
| "rewards/boxed_accuracy_reward/mean": 0.1625, |
| "rewards/boxed_accuracy_reward/std": 0.26302810609340666, |
| "step": 310, |
| "step_time": 29.716965569899912 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 388.2, |
| "completions/max_terminated_length": 343.9, |
| "completions/mean_length": 237.475, |
| "completions/mean_terminated_length": 228.22321472167968, |
| "completions/min_length": 132.2, |
| "completions/min_terminated_length": 132.2, |
| "entropy": 0.249158176779747, |
| "epoch": 0.032, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 6.810000000000001e-06, |
| "loss": -0.009709689766168594, |
| "num_tokens": 1142949.0, |
| "reward": 0.2, |
| "reward_std": 0.11700168251991272, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.11700168251991272, |
| "step": 320, |
| "step_time": 26.217313883200042 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 355.9, |
| "completions/max_terminated_length": 355.9, |
| "completions/mean_length": 207.9625, |
| "completions/mean_terminated_length": 207.9625, |
| "completions/min_length": 134.0, |
| "completions/min_terminated_length": 134.0, |
| "entropy": 0.257133349776268, |
| "epoch": 0.033, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 6.710000000000001e-06, |
| "loss": 0.0044826693832874295, |
| "num_tokens": 1177898.0, |
| "reward": 0.325, |
| "reward_std": 0.17422050833702088, |
| "rewards/boxed_accuracy_reward/mean": 0.325, |
| "rewards/boxed_accuracy_reward/std": 0.17422052025794982, |
| "step": 330, |
| "step_time": 24.09444000950098 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 455.1, |
| "completions/max_terminated_length": 405.7, |
| "completions/mean_length": 259.3375, |
| "completions/mean_terminated_length": 250.85535888671876, |
| "completions/min_length": 135.2, |
| "completions/min_terminated_length": 135.2, |
| "entropy": 0.2380140107125044, |
| "epoch": 0.034, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 0.0, |
| "learning_rate": 6.610000000000001e-06, |
| "loss": 0.07607665061950683, |
| "num_tokens": 1214925.0, |
| "reward": 0.375, |
| "reward_std": 0.25040294229984283, |
| "rewards/boxed_accuracy_reward/mean": 0.375, |
| "rewards/boxed_accuracy_reward/std": 0.2504029482603073, |
| "step": 340, |
| "step_time": 30.860040666599524 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0375, |
| "completions/max_length": 509.9, |
| "completions/max_terminated_length": 390.4, |
| "completions/mean_length": 283.525, |
| "completions/mean_terminated_length": 255.67381134033204, |
| "completions/min_length": 167.1, |
| "completions/min_terminated_length": 167.1, |
| "entropy": 0.2493958815932274, |
| "epoch": 0.035, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 6.51e-06, |
| "loss": 0.040489372611045835, |
| "num_tokens": 1249775.0, |
| "reward": 0.1125, |
| "reward_std": 0.15526474714279176, |
| "rewards/boxed_accuracy_reward/mean": 0.1125, |
| "rewards/boxed_accuracy_reward/std": 0.15526476502418518, |
| "step": 350, |
| "step_time": 34.709320317299714 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 379.4, |
| "completions/max_terminated_length": 379.4, |
| "completions/mean_length": 231.2125, |
| "completions/mean_terminated_length": 231.2125, |
| "completions/min_length": 124.0, |
| "completions/min_terminated_length": 124.0, |
| "entropy": 0.268961301445961, |
| "epoch": 0.036, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 6.4100000000000005e-06, |
| "loss": -0.025786453485488893, |
| "num_tokens": 1284312.0, |
| "reward": 0.1, |
| "reward_std": 0.07071067690849304, |
| "rewards/boxed_accuracy_reward/mean": 0.1, |
| "rewards/boxed_accuracy_reward/std": 0.07071067690849304, |
| "step": 360, |
| "step_time": 25.807292600200164 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 331.3, |
| "completions/max_terminated_length": 331.3, |
| "completions/mean_length": 220.925, |
| "completions/mean_terminated_length": 220.925, |
| "completions/min_length": 154.9, |
| "completions/min_terminated_length": 154.9, |
| "entropy": 0.20115781389176846, |
| "epoch": 0.037, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.40251433849334717, |
| "learning_rate": 6.3100000000000006e-06, |
| "loss": -0.008264368772506714, |
| "num_tokens": 1314754.0, |
| "reward": 0.1125, |
| "reward_std": 0.10606601536273956, |
| "rewards/boxed_accuracy_reward/mean": 0.1125, |
| "rewards/boxed_accuracy_reward/std": 0.10606601536273956, |
| "step": 370, |
| "step_time": 22.319113775398726 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 397.9, |
| "completions/max_terminated_length": 378.2, |
| "completions/mean_length": 229.525, |
| "completions/mean_terminated_length": 221.51964416503907, |
| "completions/min_length": 131.6, |
| "completions/min_terminated_length": 131.6, |
| "entropy": 0.23055790960788727, |
| "epoch": 0.038, |
| "frac_reward_zero_std": 0.9, |
| "grad_norm": 0.0, |
| "learning_rate": 6.210000000000001e-06, |
| "loss": 0.020930925011634828, |
| "num_tokens": 1347092.0, |
| "reward": 0.025, |
| "reward_std": 0.046291005611419675, |
| "rewards/boxed_accuracy_reward/mean": 0.025, |
| "rewards/boxed_accuracy_reward/std": 0.046291005611419675, |
| "step": 380, |
| "step_time": 26.844786134899188 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 486.9, |
| "completions/max_terminated_length": 446.0, |
| "completions/mean_length": 300.1, |
| "completions/mean_terminated_length": 291.9678588867188, |
| "completions/min_length": 183.8, |
| "completions/min_terminated_length": 183.8, |
| "entropy": 0.1606756892055273, |
| "epoch": 0.039, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 6.110000000000001e-06, |
| "loss": -0.025502890348434448, |
| "num_tokens": 1385580.0, |
| "reward": 0.1875, |
| "reward_std": 0.245769664645195, |
| "rewards/boxed_accuracy_reward/mean": 0.1875, |
| "rewards/boxed_accuracy_reward/std": 0.24576967656612397, |
| "step": 390, |
| "step_time": 32.540994760800096 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 311.1, |
| "completions/max_terminated_length": 311.1, |
| "completions/mean_length": 216.7125, |
| "completions/mean_terminated_length": 216.7125, |
| "completions/min_length": 156.9, |
| "completions/min_terminated_length": 156.9, |
| "entropy": 0.19945215657353402, |
| "epoch": 0.04, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.6455156803131104, |
| "learning_rate": 6.01e-06, |
| "loss": 0.026773136854171754, |
| "num_tokens": 1417461.0, |
| "reward": 0.225, |
| "reward_std": 0.1632926881313324, |
| "rewards/boxed_accuracy_reward/mean": 0.225, |
| "rewards/boxed_accuracy_reward/std": 0.1632926881313324, |
| "step": 400, |
| "step_time": 21.00104573589997 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.05, |
| "completions/max_length": 493.7, |
| "completions/max_terminated_length": 403.2, |
| "completions/mean_length": 278.675, |
| "completions/mean_terminated_length": 238.26178894042968, |
| "completions/min_length": 146.1, |
| "completions/min_terminated_length": 146.1, |
| "entropy": 0.20032316148281099, |
| "epoch": 0.041, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 5.91e-06, |
| "loss": 0.045339572429656985, |
| "num_tokens": 1456083.0, |
| "reward": 0.1375, |
| "reward_std": 0.15235702097415924, |
| "rewards/boxed_accuracy_reward/mean": 0.1375, |
| "rewards/boxed_accuracy_reward/std": 0.15235702097415924, |
| "step": 410, |
| "step_time": 33.424314678498924 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 380.4, |
| "completions/max_terminated_length": 327.8, |
| "completions/mean_length": 234.2375, |
| "completions/mean_terminated_length": 225.44107360839843, |
| "completions/min_length": 155.1, |
| "completions/min_terminated_length": 155.1, |
| "entropy": 0.20021430365741252, |
| "epoch": 0.042, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 5.81e-06, |
| "loss": 0.017741450667381288, |
| "num_tokens": 1487414.0, |
| "reward": 0.3375, |
| "reward_std": 0.1977891743183136, |
| "rewards/boxed_accuracy_reward/mean": 0.3375, |
| "rewards/boxed_accuracy_reward/std": 0.19778918027877807, |
| "step": 420, |
| "step_time": 25.730258163498366 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 358.2, |
| "completions/max_terminated_length": 358.2, |
| "completions/mean_length": 223.7375, |
| "completions/mean_terminated_length": 223.7375, |
| "completions/min_length": 126.5, |
| "completions/min_terminated_length": 126.5, |
| "entropy": 0.27519515678286555, |
| "epoch": 0.043, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 5.71e-06, |
| "loss": 0.019644300639629363, |
| "num_tokens": 1518057.0, |
| "reward": 0.1375, |
| "reward_std": 0.19864802658557892, |
| "rewards/boxed_accuracy_reward/mean": 0.1375, |
| "rewards/boxed_accuracy_reward/std": 0.19864802658557892, |
| "step": 430, |
| "step_time": 24.234244189500167 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 320.9, |
| "completions/max_terminated_length": 320.9, |
| "completions/mean_length": 207.3625, |
| "completions/mean_terminated_length": 207.3625, |
| "completions/min_length": 140.0, |
| "completions/min_terminated_length": 140.0, |
| "entropy": 0.23536108247935772, |
| "epoch": 0.044, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 5.610000000000001e-06, |
| "loss": 0.01632075011730194, |
| "num_tokens": 1549726.0, |
| "reward": 0.2875, |
| "reward_std": 0.13509859144687653, |
| "rewards/boxed_accuracy_reward/mean": 0.2875, |
| "rewards/boxed_accuracy_reward/std": 0.13509859144687653, |
| "step": 440, |
| "step_time": 21.775102456000603 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 491.2, |
| "completions/max_terminated_length": 397.3, |
| "completions/mean_length": 289.6375, |
| "completions/mean_terminated_length": 272.23036041259763, |
| "completions/min_length": 176.8, |
| "completions/min_terminated_length": 176.8, |
| "entropy": 0.20252985879778862, |
| "epoch": 0.045, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 5.510000000000001e-06, |
| "loss": 0.07112041711807252, |
| "num_tokens": 1586873.0, |
| "reward": 0.2, |
| "reward_std": 0.23691109120845794, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.23691110908985138, |
| "step": 450, |
| "step_time": 33.158058354499374 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 447.2, |
| "completions/max_terminated_length": 447.2, |
| "completions/mean_length": 284.3875, |
| "completions/mean_terminated_length": 284.3875, |
| "completions/min_length": 166.0, |
| "completions/min_terminated_length": 166.0, |
| "entropy": 0.1773521050810814, |
| "epoch": 0.046, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 5.410000000000001e-06, |
| "loss": -0.04271512031555176, |
| "num_tokens": 1622272.0, |
| "reward": 0.25, |
| "reward_std": 0.13340125977993011, |
| "rewards/boxed_accuracy_reward/mean": 0.25, |
| "rewards/boxed_accuracy_reward/std": 0.1334012657403946, |
| "step": 460, |
| "step_time": 30.262695012098995 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 510.3, |
| "completions/max_terminated_length": 417.6, |
| "completions/mean_length": 271.6625, |
| "completions/mean_terminated_length": 254.18214416503906, |
| "completions/min_length": 141.5, |
| "completions/min_terminated_length": 141.5, |
| "entropy": 0.23274324238300323, |
| "epoch": 0.047, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 5.310000000000001e-06, |
| "loss": 0.08956294059753418, |
| "num_tokens": 1657333.0, |
| "reward": 0.3, |
| "reward_std": 0.20411193668842315, |
| "rewards/boxed_accuracy_reward/mean": 0.3, |
| "rewards/boxed_accuracy_reward/std": 0.20411194264888763, |
| "step": 470, |
| "step_time": 34.52710425489859 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 446.0, |
| "completions/max_terminated_length": 396.1, |
| "completions/mean_length": 276.6125, |
| "completions/mean_terminated_length": 268.86607360839844, |
| "completions/min_length": 169.2, |
| "completions/min_terminated_length": 169.2, |
| "entropy": 0.25105012357234957, |
| "epoch": 0.048, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.0, |
| "learning_rate": 5.210000000000001e-06, |
| "loss": -0.00809616819024086, |
| "num_tokens": 1694526.0, |
| "reward": 0.0375, |
| "reward_std": 0.0816463440656662, |
| "rewards/boxed_accuracy_reward/mean": 0.0375, |
| "rewards/boxed_accuracy_reward/std": 0.0816463440656662, |
| "step": 480, |
| "step_time": 30.273345093600803 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 442.4, |
| "completions/max_terminated_length": 381.1, |
| "completions/mean_length": 224.5125, |
| "completions/mean_terminated_length": 214.30535736083985, |
| "completions/min_length": 104.2, |
| "completions/min_terminated_length": 104.2, |
| "entropy": 0.24412711150944233, |
| "epoch": 0.049, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.4479704797267914, |
| "learning_rate": 5.11e-06, |
| "loss": 0.03438442945480347, |
| "num_tokens": 1726799.0, |
| "reward": 0.2125, |
| "reward_std": 0.1759178400039673, |
| "rewards/boxed_accuracy_reward/mean": 0.2125, |
| "rewards/boxed_accuracy_reward/std": 0.17591784596443177, |
| "step": 490, |
| "step_time": 29.914187529801712 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 327.1, |
| "completions/max_terminated_length": 327.1, |
| "completions/mean_length": 217.2875, |
| "completions/mean_terminated_length": 217.2875, |
| "completions/min_length": 128.0, |
| "completions/min_terminated_length": 128.0, |
| "entropy": 0.23657882679253817, |
| "epoch": 0.05, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.6822343468666077, |
| "learning_rate": 5.01e-06, |
| "loss": 0.04140647947788238, |
| "num_tokens": 1758502.0, |
| "reward": 0.275, |
| "reward_std": 0.1632926881313324, |
| "rewards/boxed_accuracy_reward/mean": 0.275, |
| "rewards/boxed_accuracy_reward/std": 0.1632926881313324, |
| "step": 500, |
| "step_time": 22.14599506190134 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 296.2, |
| "completions/max_terminated_length": 296.2, |
| "completions/mean_length": 206.925, |
| "completions/mean_terminated_length": 206.925, |
| "completions/min_length": 134.8, |
| "completions/min_terminated_length": 134.8, |
| "entropy": 0.21120237074792386, |
| "epoch": 0.051, |
| "frac_reward_zero_std": 0.4, |
| "grad_norm": 0.0, |
| "learning_rate": 4.9100000000000004e-06, |
| "loss": -0.02729707658290863, |
| "num_tokens": 1787704.0, |
| "reward": 0.4125, |
| "reward_std": 0.2592615157365799, |
| "rewards/boxed_accuracy_reward/mean": 0.4125, |
| "rewards/boxed_accuracy_reward/std": 0.2592615157365799, |
| "step": 510, |
| "step_time": 20.121878111200203 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 397.7, |
| "completions/max_terminated_length": 397.7, |
| "completions/mean_length": 263.6125, |
| "completions/mean_terminated_length": 263.6125, |
| "completions/min_length": 164.6, |
| "completions/min_terminated_length": 164.6, |
| "entropy": 0.214491768181324, |
| "epoch": 0.052, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 4.8100000000000005e-06, |
| "loss": 0.003942453116178512, |
| "num_tokens": 1822977.0, |
| "reward": 0.0875, |
| "reward_std": 0.16875659823417663, |
| "rewards/boxed_accuracy_reward/mean": 0.0875, |
| "rewards/boxed_accuracy_reward/std": 0.1687566041946411, |
| "step": 520, |
| "step_time": 26.994221128601204 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0375, |
| "completions/max_length": 485.0, |
| "completions/max_terminated_length": 358.8, |
| "completions/mean_length": 280.6625, |
| "completions/mean_terminated_length": 254.4261932373047, |
| "completions/min_length": 165.4, |
| "completions/min_terminated_length": 165.4, |
| "entropy": 0.3250377386808395, |
| "epoch": 0.053, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.6389704346656799, |
| "learning_rate": 4.71e-06, |
| "loss": 0.08574633002281189, |
| "num_tokens": 1859638.0, |
| "reward": 0.15, |
| "reward_std": 0.15782093107700348, |
| "rewards/boxed_accuracy_reward/mean": 0.15, |
| "rewards/boxed_accuracy_reward/std": 0.15782093703746797, |
| "step": 530, |
| "step_time": 32.832222134500626 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 361.9, |
| "completions/max_terminated_length": 361.9, |
| "completions/mean_length": 252.375, |
| "completions/mean_terminated_length": 252.375, |
| "completions/min_length": 164.6, |
| "completions/min_terminated_length": 164.6, |
| "entropy": 0.2565031200647354, |
| "epoch": 0.054, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 4.610000000000001e-06, |
| "loss": 0.021490342915058136, |
| "num_tokens": 1893300.0, |
| "reward": 0.1625, |
| "reward_std": 0.20871699452400208, |
| "rewards/boxed_accuracy_reward/mean": 0.1625, |
| "rewards/boxed_accuracy_reward/std": 0.2087170124053955, |
| "step": 540, |
| "step_time": 24.475655020699925 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 373.8, |
| "completions/max_terminated_length": 299.6, |
| "completions/mean_length": 206.3, |
| "completions/mean_terminated_length": 185.17916717529297, |
| "completions/min_length": 111.3, |
| "completions/min_terminated_length": 111.3, |
| "entropy": 0.22958396673202514, |
| "epoch": 0.055, |
| "frac_reward_zero_std": 0.3, |
| "grad_norm": 0.6256431937217712, |
| "learning_rate": 4.510000000000001e-06, |
| "loss": 0.03615856468677521, |
| "num_tokens": 1923932.0, |
| "reward": 0.375, |
| "reward_std": 0.29838343262672423, |
| "rewards/boxed_accuracy_reward/mean": 0.375, |
| "rewards/boxed_accuracy_reward/std": 0.2983834445476532, |
| "step": 550, |
| "step_time": 25.349692796901216 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.05, |
| "completions/max_length": 475.5, |
| "completions/max_terminated_length": 364.6, |
| "completions/mean_length": 278.4375, |
| "completions/mean_terminated_length": 239.52916870117187, |
| "completions/min_length": 158.6, |
| "completions/min_terminated_length": 158.6, |
| "entropy": 0.2459364876151085, |
| "epoch": 0.056, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.8983960151672363, |
| "learning_rate": 4.41e-06, |
| "loss": -0.031027239561080933, |
| "num_tokens": 1963111.0, |
| "reward": 0.075, |
| "reward_std": 0.18771235942840575, |
| "rewards/boxed_accuracy_reward/mean": 0.075, |
| "rewards/boxed_accuracy_reward/std": 0.18771235942840575, |
| "step": 560, |
| "step_time": 32.03998489129954 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 419.6, |
| "completions/max_terminated_length": 378.6, |
| "completions/mean_length": 257.3625, |
| "completions/mean_terminated_length": 249.33035888671876, |
| "completions/min_length": 171.2, |
| "completions/min_terminated_length": 171.2, |
| "entropy": 0.2526143785566092, |
| "epoch": 0.057, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 4.31e-06, |
| "loss": 0.01212334781885147, |
| "num_tokens": 1998908.0, |
| "reward": 0.2, |
| "reward_std": 0.14056250154972078, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.14056250751018523, |
| "step": 570, |
| "step_time": 28.352883074600687 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 394.2, |
| "completions/max_terminated_length": 394.2, |
| "completions/mean_length": 257.6875, |
| "completions/mean_terminated_length": 257.6875, |
| "completions/min_length": 160.4, |
| "completions/min_terminated_length": 160.4, |
| "entropy": 0.2496233806014061, |
| "epoch": 0.058, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.6807348132133484, |
| "learning_rate": 4.21e-06, |
| "loss": 0.0322126567363739, |
| "num_tokens": 2034387.0, |
| "reward": 0.1375, |
| "reward_std": 0.1767766922712326, |
| "rewards/boxed_accuracy_reward/mean": 0.1375, |
| "rewards/boxed_accuracy_reward/std": 0.1767766922712326, |
| "step": 580, |
| "step_time": 26.78779618450062 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 371.7, |
| "completions/max_terminated_length": 302.6, |
| "completions/mean_length": 208.2625, |
| "completions/mean_terminated_length": 198.36964416503906, |
| "completions/min_length": 121.9, |
| "completions/min_terminated_length": 121.9, |
| "entropy": 0.22029370740056037, |
| "epoch": 0.059, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 4.1100000000000005e-06, |
| "loss": 0.0259624183177948, |
| "num_tokens": 2067680.0, |
| "reward": 0.275, |
| "reward_std": 0.18685350716114044, |
| "rewards/boxed_accuracy_reward/mean": 0.275, |
| "rewards/boxed_accuracy_reward/std": 0.18685351312160492, |
| "step": 590, |
| "step_time": 25.24782887190013 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 447.7, |
| "completions/max_terminated_length": 380.5, |
| "completions/mean_length": 247.1375, |
| "completions/mean_terminated_length": 229.69286193847657, |
| "completions/min_length": 137.8, |
| "completions/min_terminated_length": 137.8, |
| "entropy": 0.24857322350144387, |
| "epoch": 0.06, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 4.0100000000000006e-06, |
| "loss": 0.04133725166320801, |
| "num_tokens": 2101531.0, |
| "reward": 0.1, |
| "reward_std": 0.13340125977993011, |
| "rewards/boxed_accuracy_reward/mean": 0.1, |
| "rewards/boxed_accuracy_reward/std": 0.1334012657403946, |
| "step": 600, |
| "step_time": 30.182934339900385 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 423.4, |
| "completions/max_terminated_length": 362.5, |
| "completions/mean_length": 264.475, |
| "completions/mean_terminated_length": 255.80714416503906, |
| "completions/min_length": 171.7, |
| "completions/min_terminated_length": 171.7, |
| "entropy": 0.23762509562075138, |
| "epoch": 0.061, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.8978026509284973, |
| "learning_rate": 3.910000000000001e-06, |
| "loss": 0.04321916103363037, |
| "num_tokens": 2137113.0, |
| "reward": 0.4, |
| "reward_std": 0.24030575454235076, |
| "rewards/boxed_accuracy_reward/mean": 0.4, |
| "rewards/boxed_accuracy_reward/std": 0.24030576050281524, |
| "step": 610, |
| "step_time": 28.61728952409976 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 368.3, |
| "completions/max_terminated_length": 368.3, |
| "completions/mean_length": 232.825, |
| "completions/mean_terminated_length": 232.825, |
| "completions/min_length": 147.5, |
| "completions/min_terminated_length": 147.5, |
| "entropy": 0.2583306774497032, |
| "epoch": 0.062, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.4808867573738098, |
| "learning_rate": 3.8100000000000004e-06, |
| "loss": 0.0032923467457294463, |
| "num_tokens": 2169507.0, |
| "reward": 0.0625, |
| "reward_std": 0.12793734967708587, |
| "rewards/boxed_accuracy_reward/mean": 0.0625, |
| "rewards/boxed_accuracy_reward/std": 0.12793734967708587, |
| "step": 620, |
| "step_time": 24.657616854002118 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 393.6, |
| "completions/max_terminated_length": 393.6, |
| "completions/mean_length": 239.7875, |
| "completions/mean_terminated_length": 239.7875, |
| "completions/min_length": 150.1, |
| "completions/min_terminated_length": 150.1, |
| "entropy": 0.24895308986306192, |
| "epoch": 0.063, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 3.7100000000000005e-06, |
| "loss": -0.008826246112585067, |
| "num_tokens": 2202426.0, |
| "reward": 0.1625, |
| "reward_std": 0.19062793254852295, |
| "rewards/boxed_accuracy_reward/mean": 0.1625, |
| "rewards/boxed_accuracy_reward/std": 0.19062793850898743, |
| "step": 630, |
| "step_time": 26.481925834499272 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 490.3, |
| "completions/max_terminated_length": 412.0, |
| "completions/mean_length": 271.6625, |
| "completions/mean_terminated_length": 261.0321441650391, |
| "completions/min_length": 156.2, |
| "completions/min_terminated_length": 156.2, |
| "entropy": 0.2555754192173481, |
| "epoch": 0.064, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.0, |
| "learning_rate": 3.61e-06, |
| "loss": -0.015103229880332946, |
| "num_tokens": 2237311.0, |
| "reward": 0.225, |
| "reward_std": 0.13340125977993011, |
| "rewards/boxed_accuracy_reward/mean": 0.225, |
| "rewards/boxed_accuracy_reward/std": 0.1334012657403946, |
| "step": 640, |
| "step_time": 32.94274350120031 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 445.9, |
| "completions/max_terminated_length": 445.9, |
| "completions/mean_length": 247.9125, |
| "completions/mean_terminated_length": 247.9125, |
| "completions/min_length": 134.0, |
| "completions/min_terminated_length": 134.0, |
| "entropy": 0.28245309442281724, |
| "epoch": 0.065, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 3.5100000000000003e-06, |
| "loss": 0.05236095190048218, |
| "num_tokens": 2271440.0, |
| "reward": 0.15, |
| "reward_std": 0.18685350716114044, |
| "rewards/boxed_accuracy_reward/mean": 0.15, |
| "rewards/boxed_accuracy_reward/std": 0.18685351312160492, |
| "step": 650, |
| "step_time": 30.18701468149884 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 364.8, |
| "completions/max_terminated_length": 317.3, |
| "completions/mean_length": 223.35, |
| "completions/mean_terminated_length": 214.17857360839844, |
| "completions/min_length": 135.9, |
| "completions/min_terminated_length": 135.9, |
| "entropy": 0.265711085498333, |
| "epoch": 0.066, |
| "frac_reward_zero_std": 0.3, |
| "grad_norm": 0.8590646982192993, |
| "learning_rate": 3.4100000000000004e-06, |
| "loss": 0.03245477676391602, |
| "num_tokens": 2304332.0, |
| "reward": 0.3375, |
| "reward_std": 0.3311904340982437, |
| "rewards/boxed_accuracy_reward/mean": 0.3375, |
| "rewards/boxed_accuracy_reward/std": 0.3311904460191727, |
| "step": 660, |
| "step_time": 24.758795369901055 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 369.7, |
| "completions/max_terminated_length": 369.7, |
| "completions/mean_length": 255.3375, |
| "completions/mean_terminated_length": 255.3375, |
| "completions/min_length": 146.1, |
| "completions/min_terminated_length": 146.1, |
| "entropy": 0.2967137709259987, |
| "epoch": 0.067, |
| "frac_reward_zero_std": 0.8, |
| "grad_norm": 0.8069291710853577, |
| "learning_rate": 3.3100000000000005e-06, |
| "loss": 0.022205421328544618, |
| "num_tokens": 2338295.0, |
| "reward": 0.2, |
| "reward_std": 0.10350983142852783, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.10350984334945679, |
| "step": 670, |
| "step_time": 24.84447250970188 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 378.8, |
| "completions/max_terminated_length": 378.8, |
| "completions/mean_length": 223.475, |
| "completions/mean_terminated_length": 223.475, |
| "completions/min_length": 140.8, |
| "completions/min_terminated_length": 140.8, |
| "entropy": 0.3041005529463291, |
| "epoch": 0.068, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 3.21e-06, |
| "loss": -0.020687612891197204, |
| "num_tokens": 2369605.0, |
| "reward": 0.2, |
| "reward_std": 0.24954409003257752, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.24954410195350646, |
| "step": 680, |
| "step_time": 25.69439109330124 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 346.6, |
| "completions/max_terminated_length": 346.6, |
| "completions/mean_length": 237.9625, |
| "completions/mean_terminated_length": 237.9625, |
| "completions/min_length": 147.4, |
| "completions/min_terminated_length": 147.4, |
| "entropy": 0.24550328701734542, |
| "epoch": 0.069, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.9554150700569153, |
| "learning_rate": 3.1100000000000003e-06, |
| "loss": 0.013637937605381012, |
| "num_tokens": 2401898.0, |
| "reward": 0.1875, |
| "reward_std": 0.13509859144687653, |
| "rewards/boxed_accuracy_reward/mean": 0.1875, |
| "rewards/boxed_accuracy_reward/std": 0.13509859144687653, |
| "step": 690, |
| "step_time": 23.408739431599678 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 390.2, |
| "completions/max_terminated_length": 390.2, |
| "completions/mean_length": 242.8125, |
| "completions/mean_terminated_length": 242.8125, |
| "completions/min_length": 145.6, |
| "completions/min_terminated_length": 145.6, |
| "entropy": 0.2589107438921928, |
| "epoch": 0.07, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 1.1996217966079712, |
| "learning_rate": 3.01e-06, |
| "loss": 0.05301226377487182, |
| "num_tokens": 2436907.0, |
| "reward": 0.1125, |
| "reward_std": 0.18138959705829621, |
| "rewards/boxed_accuracy_reward/mean": 0.1125, |
| "rewards/boxed_accuracy_reward/std": 0.18138959705829621, |
| "step": 700, |
| "step_time": 26.322794716500358 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 369.9, |
| "completions/max_terminated_length": 291.4, |
| "completions/mean_length": 211.5125, |
| "completions/mean_terminated_length": 200.55178680419922, |
| "completions/min_length": 133.0, |
| "completions/min_terminated_length": 133.0, |
| "entropy": 0.19192061349749565, |
| "epoch": 0.071, |
| "frac_reward_zero_std": 0.9, |
| "grad_norm": 0.0, |
| "learning_rate": 2.91e-06, |
| "loss": -0.003464340791106224, |
| "num_tokens": 2467292.0, |
| "reward": 0.2875, |
| "reward_std": 0.03535533845424652, |
| "rewards/boxed_accuracy_reward/mean": 0.2875, |
| "rewards/boxed_accuracy_reward/std": 0.03535533845424652, |
| "step": 710, |
| "step_time": 25.09874547720101 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.05, |
| "completions/max_length": 574.9, |
| "completions/max_terminated_length": 388.2, |
| "completions/mean_length": 293.5125, |
| "completions/mean_terminated_length": 260.60000762939455, |
| "completions/min_length": 169.5, |
| "completions/min_terminated_length": 169.5, |
| "entropy": 0.22459689602255822, |
| "epoch": 0.072, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.2952307164669037, |
| "learning_rate": 2.8100000000000006e-06, |
| "loss": 0.04532091617584229, |
| "num_tokens": 2502805.0, |
| "reward": 0.25, |
| "reward_std": 0.15782093107700348, |
| "rewards/boxed_accuracy_reward/mean": 0.25, |
| "rewards/boxed_accuracy_reward/std": 0.15782093703746797, |
| "step": 720, |
| "step_time": 38.782466992000266 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 315.9, |
| "completions/max_terminated_length": 315.9, |
| "completions/mean_length": 216.875, |
| "completions/mean_terminated_length": 216.875, |
| "completions/min_length": 130.2, |
| "completions/min_terminated_length": 130.2, |
| "entropy": 0.26642010658979415, |
| "epoch": 0.073, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.7213253974914551, |
| "learning_rate": 2.7100000000000003e-06, |
| "loss": -0.004600425064563751, |
| "num_tokens": 2535131.0, |
| "reward": 0.1375, |
| "reward_std": 0.17422835528850555, |
| "rewards/boxed_accuracy_reward/mean": 0.1375, |
| "rewards/boxed_accuracy_reward/std": 0.17422835528850555, |
| "step": 730, |
| "step_time": 21.296572963699145 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 427.0, |
| "completions/max_terminated_length": 427.0, |
| "completions/mean_length": 246.8375, |
| "completions/mean_terminated_length": 246.8375, |
| "completions/min_length": 159.4, |
| "completions/min_terminated_length": 159.4, |
| "entropy": 0.2320882935076952, |
| "epoch": 0.074, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 0.0, |
| "learning_rate": 2.6100000000000004e-06, |
| "loss": -0.0016866859048604965, |
| "num_tokens": 2567462.0, |
| "reward": 0.225, |
| "reward_std": 0.1796922653913498, |
| "rewards/boxed_accuracy_reward/mean": 0.225, |
| "rewards/boxed_accuracy_reward/std": 0.17969227135181426, |
| "step": 740, |
| "step_time": 28.78491548419988 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0125, |
| "completions/max_length": 407.3, |
| "completions/max_terminated_length": 341.9, |
| "completions/mean_length": 229.175, |
| "completions/mean_terminated_length": 218.69285736083984, |
| "completions/min_length": 129.1, |
| "completions/min_terminated_length": 129.1, |
| "entropy": 0.2695804834365845, |
| "epoch": 0.075, |
| "frac_reward_zero_std": 0.6, |
| "grad_norm": 1.118378758430481, |
| "learning_rate": 2.51e-06, |
| "loss": 0.02082826942205429, |
| "num_tokens": 2601372.0, |
| "reward": 0.175, |
| "reward_std": 0.17045392990112304, |
| "rewards/boxed_accuracy_reward/mean": 0.175, |
| "rewards/boxed_accuracy_reward/std": 0.17045392990112304, |
| "step": 750, |
| "step_time": 27.563064585998653 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 250.8, |
| "completions/max_terminated_length": 250.8, |
| "completions/mean_length": 172.0375, |
| "completions/mean_terminated_length": 172.0375, |
| "completions/min_length": 121.6, |
| "completions/min_terminated_length": 121.6, |
| "entropy": 0.2495213456451893, |
| "epoch": 0.076, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 1.1198997497558594, |
| "learning_rate": 2.4100000000000002e-06, |
| "loss": -0.0004969652742147446, |
| "num_tokens": 2631303.0, |
| "reward": 0.3125, |
| "reward_std": 0.14433692693710326, |
| "rewards/boxed_accuracy_reward/mean": 0.3125, |
| "rewards/boxed_accuracy_reward/std": 0.14433693289756774, |
| "step": 760, |
| "step_time": 17.046090553899557 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 341.0, |
| "completions/max_terminated_length": 341.0, |
| "completions/mean_length": 209.725, |
| "completions/mean_terminated_length": 209.725, |
| "completions/min_length": 120.9, |
| "completions/min_terminated_length": 120.9, |
| "entropy": 0.2746701970696449, |
| "epoch": 0.077, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 2.3100000000000003e-06, |
| "loss": 0.04669084250926971, |
| "num_tokens": 2661489.0, |
| "reward": 0.25, |
| "reward_std": 0.23691109120845794, |
| "rewards/boxed_accuracy_reward/mean": 0.25, |
| "rewards/boxed_accuracy_reward/std": 0.23691110908985138, |
| "step": 770, |
| "step_time": 23.079953304801894 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0625, |
| "completions/max_length": 599.1, |
| "completions/max_terminated_length": 445.5, |
| "completions/mean_length": 323.7625, |
| "completions/mean_terminated_length": 287.5857177734375, |
| "completions/min_length": 165.4, |
| "completions/min_terminated_length": 165.4, |
| "entropy": 0.20161464884877206, |
| "epoch": 0.078, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.6654395461082458, |
| "learning_rate": 2.21e-06, |
| "loss": -0.01412966102361679, |
| "num_tokens": 2703534.0, |
| "reward": 0.2, |
| "reward_std": 0.14056250154972078, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.14056250751018523, |
| "step": 780, |
| "step_time": 40.28982557420022 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.025, |
| "completions/max_length": 488.7, |
| "completions/max_terminated_length": 422.1, |
| "completions/mean_length": 281.3625, |
| "completions/mean_terminated_length": 261.6958343505859, |
| "completions/min_length": 144.9, |
| "completions/min_terminated_length": 144.9, |
| "entropy": 0.23346679359674455, |
| "epoch": 0.079, |
| "frac_reward_zero_std": 0.5, |
| "grad_norm": 0.0, |
| "learning_rate": 2.11e-06, |
| "loss": 0.03522632122039795, |
| "num_tokens": 2739035.0, |
| "reward": 0.25, |
| "reward_std": 0.2529387533664703, |
| "rewards/boxed_accuracy_reward/mean": 0.25, |
| "rewards/boxed_accuracy_reward/std": 0.2529387533664703, |
| "step": 790, |
| "step_time": 33.08498501170034 |
| }, |
| { |
| "clip_ratio/high_max": 0.0, |
| "clip_ratio/high_mean": 0.0, |
| "clip_ratio/low_mean": 0.0, |
| "clip_ratio/low_min": 0.0, |
| "clip_ratio/region_mean": 0.0, |
| "completions/clipped_ratio": 0.0, |
| "completions/max_length": 344.0, |
| "completions/max_terminated_length": 344.0, |
| "completions/mean_length": 210.85, |
| "completions/mean_terminated_length": 210.85, |
| "completions/min_length": 133.1, |
| "completions/min_terminated_length": 133.1, |
| "entropy": 0.25274474173784256, |
| "epoch": 0.08, |
| "frac_reward_zero_std": 0.7, |
| "grad_norm": 0.33814314007759094, |
| "learning_rate": 2.0100000000000002e-06, |
| "loss": 0.009773983806371688, |
| "num_tokens": 2768951.0, |
| "reward": 0.2, |
| "reward_std": 0.13340125977993011, |
| "rewards/boxed_accuracy_reward/mean": 0.2, |
| "rewards/boxed_accuracy_reward/std": 0.1334012657403946, |
| "step": 800, |
| "step_time": 23.216696536900418 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1000, |
| "num_input_tokens_seen": 2768951, |
| "num_train_epochs": 1, |
| "save_steps": 200, |
| "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": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
