{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 16.571428571428573, "eval_steps": 500, "global_step": 50, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio/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": 60.0, "completions/max_terminated_length": 60.0, "completions/mean_length": 59.25, "completions/mean_terminated_length": 59.25, "completions/min_length": 58.5, "completions/min_terminated_length": 58.5, "epoch": 0.5714285714285714, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 2.666666666666666e-07, "loss": 0.0, "num_tokens": 2522.0, "reward": 8.320502758026123, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.14650316163897514, "rewards/question_recreation_reward_func/std": 0.13350021466612816, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 61.5, "completions/max_terminated_length": 61.5, "completions/mean_length": 60.33333396911621, "completions/mean_terminated_length": 60.33333396911621, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 1.2857142857142856, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 8e-07, "loss": 0.0, "num_tokens": 5054.0, "reward": 7.801945447921753, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 0.0, "rewards/question_recreation_reward_func/mean": 0.12794561684131622, "rewards/question_recreation_reward_func/std": 0.08619211986660957, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 0.0, "completions/max_length": 64.0, "completions/max_terminated_length": 64.0, "completions/mean_length": 61.83333396911621, "completions/mean_terminated_length": 61.83333396911621, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 1.8571428571428572, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 7.96430995261912e-07, "loss": 0.0, "num_tokens": 7594.0, "reward": 7.469648122787476, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 0.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.29664809443056583, "rewards/question_recreation_reward_func/std": 0.12906248797662556, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 62.5, "completions/max_terminated_length": 62.5, "completions/mean_length": 60.875, "completions/mean_terminated_length": 60.875, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 2.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 7.857876700217506e-07, "loss": 0.0, "num_tokens": 10129.0, "reward": 9.379204273223877, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.20520437508821487, "rewards/question_recreation_reward_func/std": 0.1838914006948471, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 64.0, "completions/max_terminated_length": 64.0, "completions/mean_length": 61.91666603088379, "completions/mean_terminated_length": 61.91666603088379, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 3.2857142857142856, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 7.682599546705714e-07, "loss": 0.0, "num_tokens": 12671.0, "reward": 7.95428204536438, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 0.0, "rewards/question_recreation_reward_func/mean": 0.2812819126993418, "rewards/question_recreation_reward_func/std": 0.1468058144673705, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 61.5, "completions/max_terminated_length": 61.5, "completions/mean_length": 59.54166603088379, "completions/mean_terminated_length": 59.54166603088379, "completions/min_length": 58.5, "completions/min_terminated_length": 58.5, "epoch": 3.857142857142857, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 7.441606317040557e-07, "loss": 0.0, "num_tokens": 15198.0, "reward": 7.847028970718384, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 0.0, "rewards/question_recreation_reward_func/mean": 0.1730293445289135, "rewards/question_recreation_reward_func/std": 0.1298600658774376, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 64.0, "completions/max_terminated_length": 64.0, "completions/mean_length": 62.25, "completions/mean_terminated_length": 62.25, "completions/min_length": 60.5, "completions/min_terminated_length": 60.5, "epoch": 4.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 7.139197541114644e-07, "loss": 0.0, "num_tokens": 17744.0, "reward": 7.7899744510650635, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 0.0, "rewards/question_recreation_reward_func/mean": 0.11697426624596119, "rewards/question_recreation_reward_func/std": 0.09874683525413275, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 62.5, "completions/max_terminated_length": 62.5, "completions/mean_length": 60.5, "completions/mean_terminated_length": 60.5, "completions/min_length": 58.5, "completions/min_terminated_length": 58.5, "epoch": 5.285714285714286, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 6.780769710698569e-07, "loss": 0.0, "num_tokens": 20276.0, "reward": 9.529415130615234, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.35541531443595886, "rewards/question_recreation_reward_func/std": 0.14058449491858482, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 60.5, "completions/max_terminated_length": 60.5, "completions/mean_length": 60.0, "completions/mean_terminated_length": 60.0, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 5.857142857142857, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 6.37271897891742e-07, "loss": 0.0, "num_tokens": 22804.0, "reward": 7.80859112739563, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 0.0, "rewards/question_recreation_reward_func/mean": 0.13559100963175297, "rewards/question_recreation_reward_func/std": 0.10423970874398947, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 61.5, "completions/max_terminated_length": 61.5, "completions/mean_length": 60.5, "completions/mean_terminated_length": 60.5, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 6.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 5.922327020746735e-07, "loss": 0.0, "num_tokens": 25336.0, "reward": 6.758650302886963, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 1.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.08565033413469791, "rewards/question_recreation_reward_func/std": 0.05564996972680092, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 59.5, "completions/max_terminated_length": 59.5, "completions/mean_length": 59.08333396911621, "completions/mean_terminated_length": 59.08333396911621, "completions/min_length": 58.5, "completions/min_terminated_length": 58.5, "epoch": 7.285714285714286, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 5.437631091350051e-07, "loss": 0.0, "num_tokens": 27856.0, "reward": 8.844876766204834, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.17087683081626892, "rewards/question_recreation_reward_func/std": 0.14287568256258965, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 60.5, "completions/max_terminated_length": 60.5, "completions/mean_length": 59.91666603088379, "completions/mean_terminated_length": 59.91666603088379, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 7.857142857142857, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 4.927280601070113e-07, "loss": 0.0, "num_tokens": 30384.0, "reward": 7.822388648986816, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.14838874712586403, "rewards/question_recreation_reward_func/std": 0.13132291659712791, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 62.5, "completions/max_terminated_length": 62.5, "completions/mean_length": 61.0, "completions/mean_terminated_length": 61.0, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 8.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 4.400382766496394e-07, "loss": 0.0, "num_tokens": 32920.0, "reward": 9.513794898986816, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.3397950530052185, "rewards/question_recreation_reward_func/std": 0.15862121805548668, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 64.0, "completions/max_terminated_length": 64.0, "completions/mean_length": 61.33333396911621, "completions/mean_terminated_length": 61.33333396911621, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 9.285714285714286, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 3.866340091969303e-07, "loss": 0.0, "num_tokens": 35456.0, "reward": 6.254339218139648, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 1.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 0.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.08133926056325436, "rewards/question_recreation_reward_func/std": 0.04871071805246174, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 60.5, "completions/max_terminated_length": 60.5, "completions/mean_length": 60.125, "completions/mean_terminated_length": 60.125, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 9.857142857142858, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 3.33468258166748e-07, "loss": 0.0, "num_tokens": 37985.0, "reward": 8.307038307189941, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.13403821364045143, "rewards/question_recreation_reward_func/std": 0.10603272262960672, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 62.5, "completions/max_terminated_length": 62.5, "completions/mean_length": 61.0, "completions/mean_terminated_length": 61.0, "completions/min_length": 59.5, "completions/min_terminated_length": 59.5, "epoch": 10.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 2.8148976764574644e-07, "loss": 0.0, "num_tokens": 40521.0, "reward": 9.50374174118042, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.3297421932220459, "rewards/question_recreation_reward_func/std": 0.1702292338013649, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 60.0, "completions/max_terminated_length": 60.0, "completions/mean_length": 59.58333396911621, "completions/mean_terminated_length": 59.58333396911621, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 11.285714285714286, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 2.3162609502866607e-07, "loss": 0.0, "num_tokens": 43045.0, "reward": 6.792631387710571, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 1.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.11863159388303757, "rewards/question_recreation_reward_func/std": 0.0983341597020626, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 64.0, "completions/max_terminated_length": 64.0, "completions/mean_length": 61.41666603088379, "completions/mean_terminated_length": 61.41666603088379, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 11.857142857142858, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 1.8476705873465095e-07, "loss": 0.0, "num_tokens": 45589.0, "reward": 9.386459350585938, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.2124595046043396, "rewards/question_recreation_reward_func/std": 0.17551389336585999, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 63.0, "completions/max_terminated_length": 63.0, "completions/mean_length": 61.5, "completions/mean_terminated_length": 61.5, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 12.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 1.41748859376479e-07, "loss": 0.0, "num_tokens": 48129.0, "reward": 9.332355499267578, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.15835533291101456, "rewards/question_recreation_reward_func/std": 0.13034464046359062, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 62.0, "completions/max_terminated_length": 62.0, "completions/mean_length": 60.75, "completions/mean_terminated_length": 60.75, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 13.285714285714286, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 1.0333915774081696e-07, "loss": 0.0, "num_tokens": 50661.0, "reward": 7.8131725788116455, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.14017230831086636, "rewards/question_recreation_reward_func/std": 0.09894967451691628, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 63.5, "completions/max_terminated_length": 63.5, "completions/mean_length": 61.33333396911621, "completions/mean_terminated_length": 61.33333396911621, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 13.857142857142858, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 7.022337586329597e-08, "loss": 0.0, "num_tokens": 53199.0, "reward": 7.823443651199341, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.14944354072213173, "rewards/question_recreation_reward_func/std": 0.13010496646165848, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 60.0, "completions/max_terminated_length": 60.0, "completions/mean_length": 59.5, "completions/mean_terminated_length": 59.5, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 14.571428571428571, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 4.299246565604755e-08, "loss": 0.0, "num_tokens": 55723.0, "reward": 7.8444740772247314, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.1704743131995201, "rewards/question_recreation_reward_func/std": 0.14472759515047073, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 62.5, "completions/max_terminated_length": 62.5, "completions/mean_length": 61.83333206176758, "completions/mean_terminated_length": 61.83333206176758, "completions/min_length": 60.5, "completions/min_terminated_length": 60.5, "epoch": 15.285714285714286, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 2.213236335683253e-08, "loss": 0.0, "num_tokens": 58263.0, "reward": 9.4916090965271, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 3.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.31760965287685394, "rewards/question_recreation_reward_func/std": 0.184238713234663, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 64.0, "completions/max_terminated_length": 64.0, "completions/mean_length": 61.83333396911621, "completions/mean_terminated_length": 61.83333396911621, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 15.857142857142858, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 8.0153179853653e-09, "loss": 0.0, "num_tokens": 60803.0, "reward": 7.469648122787476, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9230000376701355, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 0.5, "rewards/final_correctness_reward_func/std": 0.5773502588272095, "rewards/question_recreation_reward_func/mean": 0.29664809443056583, "rewards/question_recreation_reward_func/std": 0.12906248797662556, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "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": 63.5, "completions/max_terminated_length": 63.5, "completions/mean_length": 61.25, "completions/mean_terminated_length": 61.25, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 16.571428571428573, "grad_norm": 0.0, "kl": 0.0, "learning_rate": 8.932485507387344e-10, "loss": 0.0, "num_tokens": 63341.0, "reward": 7.827478647232056, "reward_std": 0.0, "rewards/concensus_correctness_reward_func/mean": 2.9240000247955322, "rewards/concensus_correctness_reward_func/std": 0.0, "rewards/consensus_reward_func/mean": 2.0, "rewards/consensus_reward_func/std": 0.0, "rewards/cumulative_reward_2/mean": 0.0, "rewards/cumulative_reward_2/std": 0.0, "rewards/final_correctness_reward_func/mean": 1.0, "rewards/final_correctness_reward_func/std": 1.154700517654419, "rewards/question_recreation_reward_func/mean": 0.15347910672426224, "rewards/question_recreation_reward_func/std": 0.13736233860254288, "rewards/soft_format_reward_func/mean": 0.0, "rewards/soft_format_reward_func/std": 0.0, "rewards/strict_format_reward_func/mean": 0.5, "rewards/strict_format_reward_func/std": 0.0, "rewards/xmlcount_reward_func/mean": 1.25, "rewards/xmlcount_reward_func/std": 0.0, "step": 50 }, { "epoch": 16.571428571428573, "step": 50, "total_flos": 0.0, "train_loss": 0.0, "train_runtime": 1419.4479, "train_samples_per_second": 0.141, "train_steps_per_second": 0.035 } ], "logging_steps": 2, "max_steps": 50, "num_input_tokens_seen": 63341, "num_train_epochs": 17, "save_steps": 25, "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 }