{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.25, "eval_steps": 500, "global_step": 100, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1998.0, "completions/mean_length": 617.0625, "completions/mean_terminated_length": 570.9031982421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 6.5873212814331055, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 163788.0, "reward": 0.29218748211860657, "reward_std": 0.3750533163547516, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 1 }, { "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1681.0, "completions/mean_length": 615.75, "completions/mean_terminated_length": 569.54833984375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 16.994749069213867, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 291028.0, "reward": 0.24062499403953552, "reward_std": 0.3783254623413086, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.26726123690605164, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 2 }, { "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": 1544.0, "completions/max_terminated_length": 1544.0, "completions/mean_length": 542.484375, "completions/mean_terminated_length": 542.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0075, "frac_reward_zero_std": 0.25, "grad_norm": 5.403246879577637, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 425931.0, "reward": 0.22968749701976776, "reward_std": 0.24118764698505402, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.20351573824882507, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 3 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1915.0, "completions/mean_length": 549.96875, "completions/mean_terminated_length": 526.1904907226562, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.125, "grad_norm": 9.253963470458984, "learning_rate": 6e-07, "loss": 0.0, "num_tokens": 557993.0, "reward": 0.13515624403953552, "reward_std": 0.17939597368240356, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "step": 4 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1963.0, "completions/mean_length": 519.921875, "completions/mean_terminated_length": 495.66668701171875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0125, "frac_reward_zero_std": 0.25, "grad_norm": 1.3908439874649048, "learning_rate": 8e-07, "loss": 0.0, "num_tokens": 690364.0, "reward": 0.19218748807907104, "reward_std": 0.29305610060691833, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "step": 5 }, { "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": 1678.0, "completions/max_terminated_length": 1678.0, "completions/mean_length": 333.984375, "completions/mean_terminated_length": 333.984375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.015, "frac_reward_zero_std": 0.0, "grad_norm": 6.072598934173584, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 819347.0, "reward": 0.33828121423721313, "reward_std": 0.4058018624782562, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.2280818521976471, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 6 }, { "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": 1771.0, "completions/max_terminated_length": 1771.0, "completions/mean_length": 354.0, "completions/mean_terminated_length": 354.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0175, "frac_reward_zero_std": 0.125, "grad_norm": 16.699325561523438, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 925811.0, "reward": 0.1914062350988388, "reward_std": 0.3238601088523865, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.24688033759593964, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "step": 7 }, { "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1472.0, "completions/mean_length": 520.828125, "completions/mean_terminated_length": 471.56451416015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02, "frac_reward_zero_std": 0.0, "grad_norm": 12.57072925567627, "learning_rate": 9.989068136093872e-07, "loss": -0.0, "num_tokens": 1048440.0, "reward": 0.23906247317790985, "reward_std": 0.31088316440582275, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 8 }, { "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": 1214.0, "completions/max_terminated_length": 1214.0, "completions/mean_length": 281.671875, "completions/mean_terminated_length": 281.671875, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0225, "frac_reward_zero_std": 0.125, "grad_norm": 17.275171279907227, "learning_rate": 9.975414512725056e-07, "loss": 0.0, "num_tokens": 1141299.0, "reward": 0.25624996423721313, "reward_std": 0.3120592534542084, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.28171807527542114, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 9 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1379.0, "completions/mean_length": 302.296875, "completions/mean_terminated_length": 274.5873107910156, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.25, "grad_norm": 10.201237678527832, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1243270.0, "reward": 0.3882812261581421, "reward_std": 0.37965649366378784, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 10 }, { "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": 865.0, "completions/max_terminated_length": 865.0, "completions/mean_length": 214.390625, "completions/mean_terminated_length": 214.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.25, "grad_norm": 19.455276489257812, "learning_rate": 9.931806517013612e-07, "loss": 0.0, "num_tokens": 1346943.0, "reward": 0.5625, "reward_std": 0.31731271743774414, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 11 }, { "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": 1532.0, "completions/max_terminated_length": 1532.0, "completions/mean_length": 236.640625, "completions/mean_terminated_length": 236.640625, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.03, "frac_reward_zero_std": 0.25, "grad_norm": 56.06517028808594, "learning_rate": 9.901899829374047e-07, "loss": 0.0, "num_tokens": 1470600.0, "reward": 0.24453124403953552, "reward_std": 0.23627185821533203, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 12 }, { "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": 1141.0, "completions/max_terminated_length": 1141.0, "completions/mean_length": 160.859375, "completions/mean_terminated_length": 160.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0325, "frac_reward_zero_std": 0.25, "grad_norm": 30.387075424194336, "learning_rate": 9.866632986240029e-07, "loss": -0.0, "num_tokens": 1572047.0, "reward": 0.3179687261581421, "reward_std": 0.29590028524398804, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.22589658200740814, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 13 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1285.0, "completions/mean_length": 340.390625, "completions/mean_terminated_length": 313.2857360839844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.25, "grad_norm": 9.51353931427002, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1691744.0, "reward": 0.32890623807907104, "reward_std": 0.1864037811756134, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.24282869696617126, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 14 }, { "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": 825.0, "completions/max_terminated_length": 825.0, "completions/mean_length": 168.75, "completions/mean_terminated_length": 168.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0375, "frac_reward_zero_std": 0.25, "grad_norm": 15.794662475585938, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1789216.0, "reward": 0.42656248807907104, "reward_std": 0.1613350808620453, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.26679685711860657, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 15 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1444.0, "completions/mean_length": 104.59375, "completions/mean_terminated_length": 73.74603271484375, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.04, "frac_reward_zero_std": 0.25, "grad_norm": 21.912553787231445, "learning_rate": 9.729086208503173e-07, "loss": -0.0, "num_tokens": 1873166.0, "reward": 0.66015625, "reward_std": 0.28690463304519653, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.20877929031848907, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 16 }, { "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": 854.0, "completions/max_terminated_length": 854.0, "completions/mean_length": 56.171875, "completions/mean_terminated_length": 56.171875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.75, "grad_norm": 6.873574256896973, "learning_rate": 9.672822322997304e-07, "loss": 0.0, "num_tokens": 1985401.0, "reward": 0.09999999403953552, "reward_std": 0.10215428471565247, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.046875, "rewards/mcq_exact_match_reward/std": 0.21304203569889069, "step": 17 }, { "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": 22.0, "completions/max_terminated_length": 22.0, "completions/mean_length": 6.25, "completions/mean_terminated_length": 6.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.5, "grad_norm": 18.71595573425293, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 2092457.0, "reward": 0.6585937142372131, "reward_std": 0.22210699319839478, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 18 }, { "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": 554.0, "completions/max_terminated_length": 554.0, "completions/mean_length": 48.65625, "completions/mean_terminated_length": 48.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.625, "grad_norm": 18.441650390625, "learning_rate": 9.545032675245813e-07, "loss": -0.0, "num_tokens": 2192075.0, "reward": 0.11328125, "reward_std": 0.14858190715312958, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.10882443934679031, "rewards/mcq_exact_match_reward/mean": 0.0625, "rewards/mcq_exact_match_reward/std": 0.24397502839565277, "step": 19 }, { "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": 142.0, "completions/max_terminated_length": 142.0, "completions/mean_length": 8.234375, "completions/mean_terminated_length": 8.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.625, "grad_norm": 25.329486846923828, "learning_rate": 9.473646649103817e-07, "loss": -0.0, "num_tokens": 2297002.0, "reward": 0.5507811903953552, "reward_std": 0.17607249319553375, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 20 }, { "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": 191.0, "completions/max_terminated_length": 191.0, "completions/mean_length": 10.46875, "completions/mean_terminated_length": 10.46875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.75, "grad_norm": 13.107504844665527, "learning_rate": 9.397368756032444e-07, "loss": 0.0, "num_tokens": 2392752.0, "reward": 0.33046871423721313, "reward_std": 0.0887577086687088, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 21 }, { "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": 349.0, "completions/max_terminated_length": 349.0, "completions/mean_length": 17.8125, "completions/mean_terminated_length": 17.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.625, "grad_norm": 26.923320770263672, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2481588.0, "reward": 0.5453124642372131, "reward_std": 0.10573774576187134, "rewards/format_reward/mean": 0.609375, "rewards/format_reward/std": 0.2083333432674408, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 22 }, { "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": 990.0, "completions/max_terminated_length": 990.0, "completions/mean_length": 56.515625, "completions/mean_terminated_length": 56.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.25, "grad_norm": 21.975736618041992, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 2581757.0, "reward": 0.47343748807907104, "reward_std": 0.2496831715106964, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 23 }, { "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": 345.0, "completions/max_terminated_length": 345.0, "completions/mean_length": 24.53125, "completions/mean_terminated_length": 24.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.5, "grad_norm": 19.507545471191406, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2674911.0, "reward": 0.40234375, "reward_std": 0.23218470811843872, "rewards/format_reward/mean": 0.5859375, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 24 }, { "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": 747.0, "completions/max_terminated_length": 747.0, "completions/mean_length": 49.328125, "completions/mean_terminated_length": 49.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.5, "grad_norm": 14.060385704040527, "learning_rate": 9.045084971874737e-07, "loss": -0.0, "num_tokens": 2776588.0, "reward": 0.24453124403953552, "reward_std": 0.11687872558832169, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.19654127955436707, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 25 }, { "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 7.640625, "completions/mean_terminated_length": 7.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.065, "frac_reward_zero_std": 0.625, "grad_norm": 24.995521545410156, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 2853725.0, "reward": 0.125, "reward_std": 0.1462520956993103, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.0625, "rewards/mcq_exact_match_reward/std": 0.24397502839565277, "step": 26 }, { "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.859375, "completions/mean_terminated_length": 6.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.875, "grad_norm": 12.034884452819824, "learning_rate": 8.842005554284295e-07, "loss": -0.0, "num_tokens": 2958020.0, "reward": 0.5874999761581421, "reward_std": 0.0578637532889843, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 27 }, { "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": 432.0, "completions/max_terminated_length": 432.0, "completions/mean_length": 23.96875, "completions/mean_terminated_length": 23.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.625, "grad_norm": 16.252504348754883, "learning_rate": 8.734107384920769e-07, "loss": 0.0, "num_tokens": 3058178.0, "reward": 0.653124988079071, "reward_std": 0.17567718029022217, "rewards/format_reward/mean": 0.59375, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 28 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.5, "grad_norm": 19.131885528564453, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3165074.0, "reward": 0.3937499523162842, "reward_std": 0.24251841008663177, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 29 }, { "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": 327.0, "completions/max_terminated_length": 327.0, "completions/mean_length": 29.59375, "completions/mean_terminated_length": 29.59375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.375, "grad_norm": 12.563758850097656, "learning_rate": 8.506183921362442e-07, "loss": -0.0, "num_tokens": 3254080.0, "reward": 0.2601562440395355, "reward_std": 0.2749083638191223, "rewards/format_reward/mean": 0.7265625, "rewards/format_reward/std": 0.250866562128067, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 30 }, { "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.859375, "completions/mean_terminated_length": 6.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0775, "frac_reward_zero_std": 0.875, "grad_norm": 7.785823822021484, "learning_rate": 8.386407858128706e-07, "loss": -0.0, "num_tokens": 3320999.0, "reward": 0.4937499761581421, "reward_std": 0.06681530177593231, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 31 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08, "frac_reward_zero_std": 0.625, "grad_norm": 15.087549209594727, "learning_rate": 8.262928807620843e-07, "loss": -0.0, "num_tokens": 3432231.0, "reward": 0.3624999523162842, "reward_std": 0.1872510462999344, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 32 }, { "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": 342.0, "completions/max_terminated_length": 342.0, "completions/mean_length": 13.96875, "completions/mean_terminated_length": 13.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.5, "grad_norm": 16.296710968017578, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 3512573.0, "reward": 0.28125, "reward_std": 0.24511480331420898, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 33 }, { "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": 100.0, "completions/max_terminated_length": 100.0, "completions/mean_length": 8.21875, "completions/mean_terminated_length": 8.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.085, "frac_reward_zero_std": 0.875, "grad_norm": 15.109249114990234, "learning_rate": 8.005405736415125e-07, "loss": -0.0, "num_tokens": 3601043.0, "reward": 0.6187499761581421, "reward_std": 0.06681530177593231, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "step": 34 }, { "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": 268.0, "completions/max_terminated_length": 268.0, "completions/mean_length": 17.671875, "completions/mean_terminated_length": 17.671875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0875, "frac_reward_zero_std": 0.625, "grad_norm": 18.687572479248047, "learning_rate": 7.871643313414718e-07, "loss": 0.0, "num_tokens": 3685510.0, "reward": 0.5249999761581421, "reward_std": 0.1462520956993103, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 35 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 3793934.0, "reward": 0.4249999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 36 }, { "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": 21.0, "completions/max_terminated_length": 21.0, "completions/mean_length": 6.234375, "completions/mean_terminated_length": 6.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0925, "frac_reward_zero_std": 0.125, "grad_norm": 36.460479736328125, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 3908741.0, "reward": 0.5648437142372131, "reward_std": 0.38751497864723206, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 37 }, { "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": 881.0, "completions/max_terminated_length": 881.0, "completions/mean_length": 23.515625, "completions/mean_terminated_length": 23.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.095, "frac_reward_zero_std": 0.75, "grad_norm": 8.36151123046875, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 3993790.0, "reward": 0.3499999940395355, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 38 }, { "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 6.828125, "completions/mean_terminated_length": 6.828125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0975, "frac_reward_zero_std": 0.5, "grad_norm": 15.606691360473633, "learning_rate": 7.306705814893439e-07, "loss": -0.0, "num_tokens": 4080659.0, "reward": 0.5874999761581421, "reward_std": 0.22201895713806152, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 39 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1, "frac_reward_zero_std": 0.875, "grad_norm": 12.567124366760254, "learning_rate": 7.158771761692464e-07, "loss": -0.0, "num_tokens": 4160227.0, "reward": 0.5812499523162842, "reward_std": 0.0578637532889843, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 40 }, { "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": 451.0, "completions/max_terminated_length": 451.0, "completions/mean_length": 12.953125, "completions/mean_terminated_length": 12.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1025, "frac_reward_zero_std": 0.5, "grad_norm": 57.7166633605957, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 4269648.0, "reward": 0.6593749523162842, "reward_std": 0.2109457552433014, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 41 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.105, "frac_reward_zero_std": 0.875, "grad_norm": 17.067180633544922, "learning_rate": 6.855986244591103e-07, "loss": 0.0, "num_tokens": 4395936.0, "reward": 0.31562498211860657, "reward_std": 0.04419416934251785, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 42 }, { "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 6.40625, "completions/mean_terminated_length": 6.40625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1075, "frac_reward_zero_std": 0.5, "grad_norm": 30.13214874267578, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 4467898.0, "reward": 0.45390623807907104, "reward_std": 0.09393882006406784, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 43 }, { "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": 120.0, "completions/max_terminated_length": 120.0, "completions/mean_length": 8.53125, "completions/mean_terminated_length": 8.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11, "frac_reward_zero_std": 0.625, "grad_norm": 15.19147777557373, "learning_rate": 6.545084971874736e-07, "loss": -0.0, "num_tokens": 4568684.0, "reward": 0.6499999761581421, "reward_std": 0.16675157845020294, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 44 }, { "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": 135.0, "completions/max_terminated_length": 135.0, "completions/mean_length": 8.796875, "completions/mean_terminated_length": 8.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1125, "frac_reward_zero_std": 0.625, "grad_norm": 14.918896675109863, "learning_rate": 6.387014543809223e-07, "loss": 0.0, "num_tokens": 4672255.0, "reward": 0.5093749761581421, "reward_std": 0.06911306828260422, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.18898223340511322, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 45 }, { "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 7.515625, "completions/mean_terminated_length": 7.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.115, "frac_reward_zero_std": 0.5, "grad_norm": 20.45856475830078, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 4758496.0, "reward": 0.484375, "reward_std": 0.21778543293476105, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 46 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1175, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 4822192.0, "reward": 0.42499998211860657, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 47 }, { "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": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.12, "frac_reward_zero_std": 0.75, "grad_norm": 12.950883865356445, "learning_rate": 5.90440267166055e-07, "loss": -0.0, "num_tokens": 4919512.0, "reward": 0.7218749523162842, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 48 }, { "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.953125, "completions/mean_terminated_length": 6.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1225, "frac_reward_zero_std": 0.5, "grad_norm": 28.649272918701172, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 4991973.0, "reward": 0.6359374523162842, "reward_std": 0.10647735744714737, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 49 }, { "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": 83.0, "completions/max_terminated_length": 83.0, "completions/mean_length": 8.984375, "completions/mean_terminated_length": 8.984375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.125, "frac_reward_zero_std": 0.375, "grad_norm": 26.786386489868164, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 5084700.0, "reward": 0.5796874761581421, "reward_std": 0.19239477813243866, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.22658175230026245, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 50 }, { "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": 87.0, "completions/max_terminated_length": 87.0, "completions/mean_length": 7.546875, "completions/mean_terminated_length": 7.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1275, "frac_reward_zero_std": 0.875, "grad_norm": 21.658172607421875, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 5167639.0, "reward": 0.6781249642372131, "reward_std": 0.0033407723531126976, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 51 }, { "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": 17.0, "completions/max_terminated_length": 17.0, "completions/mean_length": 7.65625, "completions/mean_terminated_length": 7.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.13, "frac_reward_zero_std": 0.375, "grad_norm": 33.7187385559082, "learning_rate": 5.247918773366111e-07, "loss": 0.0, "num_tokens": 5256449.0, "reward": 0.671875, "reward_std": 0.1372174769639969, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 52 }, { "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": 142.0, "completions/max_terminated_length": 142.0, "completions/mean_length": 10.1875, "completions/mean_terminated_length": 10.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1325, "frac_reward_zero_std": 0.5, "grad_norm": 30.225751876831055, "learning_rate": 5.082669723831793e-07, "loss": 0.0, "num_tokens": 5339973.0, "reward": 0.7390625476837158, "reward_std": 0.07245384156703949, "rewards/format_reward/mean": 0.671875, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 53 }, { "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": 169.0, "completions/max_terminated_length": 169.0, "completions/mean_length": 14.609375, "completions/mean_terminated_length": 14.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.135, "frac_reward_zero_std": 0.25, "grad_norm": 26.415239334106445, "learning_rate": 4.917330276168208e-07, "loss": 0.0, "num_tokens": 5418324.0, "reward": 0.6453125476837158, "reward_std": 0.23691105842590332, "rewards/format_reward/mean": 0.828125, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "step": 54 }, { "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": 71.0, "completions/max_terminated_length": 71.0, "completions/mean_length": 11.15625, "completions/mean_terminated_length": 11.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1375, "frac_reward_zero_std": 0.375, "grad_norm": 26.392406463623047, "learning_rate": 4.752081226633888e-07, "loss": -0.0, "num_tokens": 5525550.0, "reward": 0.6539062261581421, "reward_std": 0.21252106130123138, "rewards/format_reward/mean": 0.7578125, "rewards/format_reward/std": 0.25185325741767883, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 55 }, { "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": 185.0, "completions/max_terminated_length": 185.0, "completions/mean_length": 14.375, "completions/mean_terminated_length": 14.375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.14, "frac_reward_zero_std": 0.375, "grad_norm": 17.015522003173828, "learning_rate": 4.5871032726383385e-07, "loss": -0.0, "num_tokens": 5628374.0, "reward": 0.633593738079071, "reward_std": 0.2549750804901123, "rewards/format_reward/mean": 0.7109375, "rewards/format_reward/std": 0.24888142943382263, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "step": 56 }, { "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 490.0, "completions/mean_length": 85.4375, "completions/mean_terminated_length": 22.129032135009766, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1425, "frac_reward_zero_std": 0.375, "grad_norm": 19.1580810546875, "learning_rate": 4.4225768151520694e-07, "loss": 0.0, "num_tokens": 5713074.0, "reward": 0.8203125, "reward_std": 0.19372814893722534, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 57 }, { "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": 277.0, "completions/max_terminated_length": 277.0, "completions/mean_length": 20.953125, "completions/mean_terminated_length": 20.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.145, "frac_reward_zero_std": 0.75, "grad_norm": 9.586272239685059, "learning_rate": 4.258681761440789e-07, "loss": -0.0, "num_tokens": 5802279.0, "reward": 0.3226562738418579, "reward_std": 0.0481996089220047, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.2313210517168045, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "step": 58 }, { "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": 243.0, "completions/max_terminated_length": 243.0, "completions/mean_length": 16.6875, "completions/mean_terminated_length": 16.6875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1475, "frac_reward_zero_std": 0.75, "grad_norm": 18.477115631103516, "learning_rate": 4.095597328339452e-07, "loss": 0.0, "num_tokens": 5882435.0, "reward": 0.6031249761581421, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 59 }, { "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": 747.0, "completions/max_terminated_length": 747.0, "completions/mean_length": 28.328125, "completions/mean_terminated_length": 28.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.15, "frac_reward_zero_std": 0.375, "grad_norm": 21.94671630859375, "learning_rate": 3.9335018462812664e-07, "loss": 0.0, "num_tokens": 5955832.0, "reward": 0.699999988079071, "reward_std": 0.19821153581142426, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 60 }, { "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": 18.0, "completions/max_terminated_length": 18.0, "completions/mean_length": 11.453125, "completions/mean_terminated_length": 11.453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1525, "frac_reward_zero_std": 0.5, "grad_norm": 30.7242431640625, "learning_rate": 3.772572564296004e-07, "loss": -0.0, "num_tokens": 6046037.0, "reward": 0.6031249761581421, "reward_std": 0.0530330091714859, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 61 }, { "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": 405.0, "completions/max_terminated_length": 405.0, "completions/mean_length": 38.359375, "completions/mean_terminated_length": 38.359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.155, "frac_reward_zero_std": 0.25, "grad_norm": 39.60837936401367, "learning_rate": 3.612985456190778e-07, "loss": 0.0, "num_tokens": 6141316.0, "reward": 0.5601562261581421, "reward_std": 0.27629151940345764, "rewards/format_reward/mean": 0.9140625, "rewards/format_reward/std": 0.209963858127594, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 62 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 329.0, "completions/mean_length": 67.171875, "completions/mean_terminated_length": 35.730159759521484, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1575, "frac_reward_zero_std": 0.5, "grad_norm": 32.035133361816406, "learning_rate": 3.454915028125263e-07, "loss": 0.0, "num_tokens": 6256351.0, "reward": 0.40625, "reward_std": 0.17226949334144592, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 63 }, { "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 321.0, "completions/mean_length": 87.671875, "completions/mean_terminated_length": 24.435483932495117, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.16, "frac_reward_zero_std": 0.625, "grad_norm": 17.660398483276367, "learning_rate": 3.2985341277917846e-07, "loss": -0.0, "num_tokens": 6380450.0, "reward": 0.82421875, "reward_std": 0.11005400866270065, "rewards/format_reward/mean": 0.8984375, "rewards/format_reward/std": 0.2387082874774933, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 64 }, { "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 575.0, "completions/mean_length": 104.171875, "completions/mean_terminated_length": 41.46773910522461, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1625, "frac_reward_zero_std": 0.5, "grad_norm": 14.603215217590332, "learning_rate": 3.1440137554088953e-07, "loss": 0.0, "num_tokens": 6463037.0, "reward": 0.8929687738418579, "reward_std": 0.2293713092803955, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.796875, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 65 }, { "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": 823.0, "completions/max_terminated_length": 823.0, "completions/mean_length": 60.671875, "completions/mean_terminated_length": 60.671875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.165, "frac_reward_zero_std": 0.5, "grad_norm": 10.126643180847168, "learning_rate": 2.9915228767351535e-07, "loss": -0.0, "num_tokens": 6560536.0, "reward": 0.45781251788139343, "reward_std": 0.19486570358276367, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 66 }, { "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": 587.0, "completions/max_terminated_length": 587.0, "completions/mean_length": 36.65625, "completions/mean_terminated_length": 36.65625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1675, "frac_reward_zero_std": 0.625, "grad_norm": 7.164755344390869, "learning_rate": 2.841228238307536e-07, "loss": 0.0, "num_tokens": 6654954.0, "reward": 0.6312500238418579, "reward_std": 0.1462520956993103, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 67 }, { "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": 714.0, "completions/max_terminated_length": 714.0, "completions/mean_length": 44.484375, "completions/mean_terminated_length": 44.484375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.17, "frac_reward_zero_std": 0.5, "grad_norm": 13.311678886413574, "learning_rate": 2.6932941851065615e-07, "loss": 0.0, "num_tokens": 6750993.0, "reward": 0.42656251788139343, "reward_std": 0.19129054248332977, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 68 }, { "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": 334.0, "completions/max_terminated_length": 334.0, "completions/mean_length": 30.359375, "completions/mean_terminated_length": 30.359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1725, "frac_reward_zero_std": 0.375, "grad_norm": 14.94647216796875, "learning_rate": 2.547882480847461e-07, "loss": 0.0, "num_tokens": 6869880.0, "reward": 0.48124998807907104, "reward_std": 0.1414213478565216, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 69 }, { "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": 323.0, "completions/max_terminated_length": 323.0, "completions/mean_length": 36.75, "completions/mean_terminated_length": 36.75, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.175, "frac_reward_zero_std": 0.375, "grad_norm": 25.44900894165039, "learning_rate": 2.4051521310939254e-07, "loss": -0.0, "num_tokens": 6946712.0, "reward": 0.6460937857627869, "reward_std": 0.21315544843673706, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 70 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 497.0, "completions/mean_length": 57.78125, "completions/mean_terminated_length": 26.19047737121582, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1775, "frac_reward_zero_std": 0.375, "grad_norm": 14.250919342041016, "learning_rate": 2.2652592093878665e-07, "loss": 0.0, "num_tokens": 7050066.0, "reward": 0.42500001192092896, "reward_std": 0.16192083060741425, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 71 }, { "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": 979.0, "completions/max_terminated_length": 979.0, "completions/mean_length": 52.5, "completions/mean_terminated_length": 52.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.18, "frac_reward_zero_std": 0.625, "grad_norm": 13.77537727355957, "learning_rate": 2.128356686585282e-07, "loss": -0.0, "num_tokens": 7142074.0, "reward": 0.7390625476837158, "reward_std": 0.17234039306640625, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 72 }, { "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": 623.0, "completions/max_terminated_length": 623.0, "completions/mean_length": 37.328125, "completions/mean_terminated_length": 37.328125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1825, "frac_reward_zero_std": 0.75, "grad_norm": 10.029747009277344, "learning_rate": 1.9945942635848745e-07, "loss": 0.0, "num_tokens": 7228127.0, "reward": 0.20625001192092896, "reward_std": 0.0530330091714859, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "step": 73 }, { "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": 502.0, "completions/max_terminated_length": 502.0, "completions/mean_length": 35.921875, "completions/mean_terminated_length": 35.921875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.185, "frac_reward_zero_std": 0.75, "grad_norm": 24.13599967956543, "learning_rate": 1.8641182076323148e-07, "loss": 0.0, "num_tokens": 7318698.0, "reward": 0.6156250238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 74 }, { "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": 650.0, "completions/max_terminated_length": 650.0, "completions/mean_length": 47.109375, "completions/mean_terminated_length": 47.109375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1875, "frac_reward_zero_std": 0.625, "grad_norm": 13.97047233581543, "learning_rate": 1.7370711923791564e-07, "loss": 0.0, "num_tokens": 7417273.0, "reward": 0.6140625476837158, "reward_std": 0.13592836260795593, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 75 }, { "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": 1312.0, "completions/max_terminated_length": 1312.0, "completions/mean_length": 47.6875, "completions/mean_terminated_length": 47.6875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.19, "frac_reward_zero_std": 0.75, "grad_norm": 14.043416023254395, "learning_rate": 1.6135921418712955e-07, "loss": -0.0, "num_tokens": 7494261.0, "reward": 0.6312500238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 76 }, { "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": 528.0, "completions/max_terminated_length": 528.0, "completions/mean_length": 54.71875, "completions/mean_terminated_length": 54.71875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1925, "frac_reward_zero_std": 0.375, "grad_norm": 16.878311157226562, "learning_rate": 1.493816078637557e-07, "loss": 0.0, "num_tokens": 7575331.0, "reward": 0.20625001192092896, "reward_std": 0.2167942076921463, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "step": 77 }, { "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": 158.0, "completions/max_terminated_length": 158.0, "completions/mean_length": 18.125, "completions/mean_terminated_length": 18.125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.195, "frac_reward_zero_std": 0.625, "grad_norm": 9.90275764465332, "learning_rate": 1.3778739760445552e-07, "loss": 0.0, "num_tokens": 7653915.0, "reward": 0.801562488079071, "reward_std": 0.13700194656848907, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 78 }, { "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": 104.0, "completions/max_terminated_length": 104.0, "completions/mean_length": 15.09375, "completions/mean_terminated_length": 15.09375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1975, "frac_reward_zero_std": 0.5, "grad_norm": 20.73286247253418, "learning_rate": 1.2658926150792322e-07, "loss": 0.0, "num_tokens": 7722105.0, "reward": 0.503125011920929, "reward_std": 0.17253795266151428, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 79 }, { "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": 286.0, "completions/max_terminated_length": 286.0, "completions/mean_length": 25.953125, "completions/mean_terminated_length": 25.953125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2, "frac_reward_zero_std": 0.875, "grad_norm": 6.114646911621094, "learning_rate": 1.1579944457157059e-07, "loss": -0.0, "num_tokens": 7799846.0, "reward": 0.6156250238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 80 }, { "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": 33.0, "completions/max_terminated_length": 33.0, "completions/mean_length": 13.515625, "completions/mean_terminated_length": 13.515625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2025, "frac_reward_zero_std": 0.875, "grad_norm": 16.864362716674805, "learning_rate": 1.0542974530180327e-07, "loss": 0.0, "num_tokens": 7871903.0, "reward": 0.7875000238418579, "reward_std": 0.06681530922651291, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 81 }, { "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": 505.0, "completions/max_terminated_length": 505.0, "completions/mean_length": 32.328125, "completions/mean_terminated_length": 32.328125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.205, "frac_reward_zero_std": 0.5, "grad_norm": 29.771133422851562, "learning_rate": 9.549150281252632e-08, "loss": 0.0, "num_tokens": 7969340.0, "reward": 0.8640625476837158, "reward_std": 0.21393072605133057, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.765625, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "step": 82 }, { "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": 476.0, "completions/max_terminated_length": 476.0, "completions/mean_length": 31.28125, "completions/mean_terminated_length": 31.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2075, "frac_reward_zero_std": 0.375, "grad_norm": 21.623123168945312, "learning_rate": 8.599558442598998e-08, "loss": 0.0, "num_tokens": 8097382.0, "reward": 0.5757812857627869, "reward_std": 0.18537145853042603, "rewards/format_reward/mean": 0.9140625, "rewards/format_reward/std": 0.2280818521976471, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 83 }, { "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": 480.0, "completions/max_terminated_length": 480.0, "completions/mean_length": 30.90625, "completions/mean_terminated_length": 30.90625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.21, "frac_reward_zero_std": 0.5, "grad_norm": 16.379615783691406, "learning_rate": 7.695237378953224e-08, "loss": 0.0, "num_tokens": 8191288.0, "reward": 0.6156250238418579, "reward_std": 0.2109457552433014, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 84 }, { "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": 24.0, "completions/max_terminated_length": 24.0, "completions/mean_length": 13.265625, "completions/mean_terminated_length": 13.265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2125, "frac_reward_zero_std": 0.875, "grad_norm": 6.981666088104248, "learning_rate": 6.837175952121304e-08, "loss": 0.0, "num_tokens": 8278001.0, "reward": 0.5789062976837158, "reward_std": 0.044563546776771545, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 85 }, { "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": 228.0, "completions/max_terminated_length": 228.0, "completions/mean_length": 25.0, "completions/mean_terminated_length": 25.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.215, "frac_reward_zero_std": 0.75, "grad_norm": 11.610461235046387, "learning_rate": 6.026312439675551e-08, "loss": 0.0, "num_tokens": 8386233.0, "reward": 0.6468750238418579, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 86 }, { "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": 106.0, "completions/max_terminated_length": 106.0, "completions/mean_length": 20.1875, "completions/mean_terminated_length": 20.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2175, "frac_reward_zero_std": 0.75, "grad_norm": 25.617238998413086, "learning_rate": 5.263533508961826e-08, "loss": -0.0, "num_tokens": 8490789.0, "reward": 0.6929687857627869, "reward_std": 0.09059806168079376, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 87 }, { "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": 164.0, "completions/max_terminated_length": 164.0, "completions/mean_length": 18.859375, "completions/mean_terminated_length": 18.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.22, "frac_reward_zero_std": 0.75, "grad_norm": 10.143875122070312, "learning_rate": 4.549673247541874e-08, "loss": 0.0, "num_tokens": 8590268.0, "reward": 0.58203125, "reward_std": 0.10459993779659271, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 88 }, { "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": 68.0, "completions/max_terminated_length": 68.0, "completions/mean_length": 14.5, "completions/mean_terminated_length": 14.5, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2225, "frac_reward_zero_std": 0.75, "grad_norm": 17.269296646118164, "learning_rate": 3.8855122511307626e-08, "loss": 0.0, "num_tokens": 8682396.0, "reward": 0.6000000238418579, "reward_std": 0.1293872892856598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 89 }, { "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": 137.0, "completions/max_terminated_length": 137.0, "completions/mean_length": 19.921875, "completions/mean_terminated_length": 19.921875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.225, "frac_reward_zero_std": 0.5, "grad_norm": 26.118375778198242, "learning_rate": 3.271776770026963e-08, "loss": 0.0, "num_tokens": 8767935.0, "reward": 0.48906251788139343, "reward_std": 0.16434110701084137, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 90 }, { "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": 1574.0, "completions/max_terminated_length": 1574.0, "completions/mean_length": 47.9375, "completions/mean_terminated_length": 47.9375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2275, "frac_reward_zero_std": 0.625, "grad_norm": 11.224387168884277, "learning_rate": 2.7091379149682682e-08, "loss": 0.0, "num_tokens": 8854627.0, "reward": 0.48906251788139343, "reward_std": 0.18268883228302002, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 91 }, { "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": 123.0, "completions/max_terminated_length": 123.0, "completions/mean_length": 15.734375, "completions/mean_terminated_length": 15.734375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.23, "frac_reward_zero_std": 0.375, "grad_norm": 24.336576461791992, "learning_rate": 2.1982109232821176e-08, "loss": -0.0, "num_tokens": 8961306.0, "reward": 0.4281250238418579, "reward_std": 0.3230033218860626, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 92 }, { "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 743.0, "completions/mean_length": 105.71875, "completions/mean_terminated_length": 43.06451416015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2325, "frac_reward_zero_std": 0.375, "grad_norm": 11.845682144165039, "learning_rate": 1.7395544861325718e-08, "loss": -0.0, "num_tokens": 9064032.0, "reward": 0.8742187023162842, "reward_std": 0.21999241411685944, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.2498759627342224, "rewards/mcq_exact_match_reward/mean": 0.78125, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 93 }, { "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 516.0, "completions/mean_length": 71.984375, "completions/mean_terminated_length": 40.619049072265625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.235, "frac_reward_zero_std": 0.625, "grad_norm": 10.289705276489258, "learning_rate": 1.3336701375997127e-08, "loss": 0.0, "num_tokens": 9153103.0, "reward": 0.5046875476837158, "reward_std": 0.15067152678966522, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 94 }, { "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": 309.0, "completions/max_terminated_length": 309.0, "completions/mean_length": 20.46875, "completions/mean_terminated_length": 20.46875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2375, "frac_reward_zero_std": 0.625, "grad_norm": 27.53934669494629, "learning_rate": 9.810017062595321e-09, "loss": 0.0, "num_tokens": 9242301.0, "reward": 0.8500000238418579, "reward_std": 0.1462520956993103, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 95 }, { "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": 253.0, "completions/max_terminated_length": 253.0, "completions/mean_length": 18.46875, "completions/mean_terminated_length": 18.46875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.24, "frac_reward_zero_std": 0.25, "grad_norm": 21.703001022338867, "learning_rate": 6.819348298638839e-09, "loss": 0.0, "num_tokens": 9330875.0, "reward": 0.59375, "reward_std": 0.2891606092453003, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 96 }, { "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": 232.0, "completions/max_terminated_length": 232.0, "completions/mean_length": 21.515625, "completions/mean_terminated_length": 21.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2425, "frac_reward_zero_std": 0.375, "grad_norm": 28.218229293823242, "learning_rate": 4.367965336512403e-09, "loss": 0.0, "num_tokens": 9437012.0, "reward": 0.5640624761581421, "reward_std": 0.20705212652683258, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.19352105259895325, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 97 }, { "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": 528.0, "completions/max_terminated_length": 528.0, "completions/mean_length": 23.984375, "completions/mean_terminated_length": 23.984375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.245, "frac_reward_zero_std": 0.75, "grad_norm": 16.938936233520508, "learning_rate": 2.458548727494292e-09, "loss": 0.0, "num_tokens": 9542347.0, "reward": 0.5234375, "reward_std": 0.07123472541570663, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.3503824472427368, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 98 }, { "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": 375.0, "completions/max_terminated_length": 375.0, "completions/mean_length": 27.046875, "completions/mean_terminated_length": 27.046875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2475, "frac_reward_zero_std": 0.625, "grad_norm": 22.74932098388672, "learning_rate": 1.0931863906127325e-09, "loss": 0.0, "num_tokens": 9620238.0, "reward": 0.8812500238418579, "reward_std": 0.16675157845020294, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.78125, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 99 }, { "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": 86.0, "completions/max_terminated_length": 86.0, "completions/mean_length": 13.40625, "completions/mean_terminated_length": 13.40625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.25, "frac_reward_zero_std": 0.875, "grad_norm": 6.1370320320129395, "learning_rate": 2.733713295369755e-10, "loss": -0.0, "num_tokens": 9717184.0, "reward": 0.5007812976837158, "reward_std": 0.05742206051945686, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 9717184, "num_train_epochs": 1, "save_steps": 50, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }