{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.125, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1470.0, "completions/mean_length": 444.328125, "completions/mean_terminated_length": 418.873046875, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.00125, "frac_reward_zero_std": 0.0, "grad_norm": 12.492976188659668, "learning_rate": 0.0, "loss": -0.0, "num_tokens": 111893.0, "reward": 0.28359371423721313, "reward_std": 0.38594749569892883, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.321882963180542, "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": 0.0, "completions/max_length": 1021.0, "completions/max_terminated_length": 1021.0, "completions/mean_length": 391.3125, "completions/mean_terminated_length": 391.3125, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 6.380683898925781, "learning_rate": 1e-07, "loss": 0.0, "num_tokens": 200761.0, "reward": 0.1484375, "reward_std": 0.16173049807548523, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.37234511971473694, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1195.0, "completions/mean_length": 412.703125, "completions/mean_terminated_length": 386.7460632324219, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.00375, "frac_reward_zero_std": 0.0, "grad_norm": 7.972509860992432, "learning_rate": 2e-07, "loss": 0.0, "num_tokens": 308846.0, "reward": 0.4562499523162842, "reward_std": 0.4088757038116455, "rewards/format_reward/mean": 0.34375, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.0, "completions/max_length": 1460.0, "completions/max_terminated_length": 1460.0, "completions/mean_length": 404.453125, "completions/mean_terminated_length": 404.453125, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 1.6674329042434692, "learning_rate": 3e-07, "loss": -0.0, "num_tokens": 403371.0, "reward": 0.3695312440395355, "reward_std": 0.43215417861938477, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.3398726284503937, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1727.0, "completions/max_terminated_length": 1727.0, "completions/mean_length": 491.453125, "completions/mean_terminated_length": 491.453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.00625, "frac_reward_zero_std": 0.0, "grad_norm": 4.789487838745117, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 512176.0, "reward": 0.27265623211860657, "reward_std": 0.37521791458129883, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.2313210517168045, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 511.53125, "completions/mean_terminated_length": 511.53125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 6.271266460418701, "learning_rate": 5e-07, "loss": -0.0, "num_tokens": 618002.0, "reward": 0.3203124701976776, "reward_std": 0.4231240153312683, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.28824523091316223, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1506.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 363.5, "completions/mean_terminated_length": 363.5, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.00875, "frac_reward_zero_std": 0.0, "grad_norm": 33.56215286254883, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 720306.0, "reward": 0.39687496423721313, "reward_std": 0.4708775281906128, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.28171807527542114, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.0, "completions/max_length": 1197.0, "completions/max_terminated_length": 1197.0, "completions/mean_length": 460.0, "completions/mean_terminated_length": 460.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.0, "grad_norm": 8.380439758300781, "learning_rate": 7e-07, "loss": -0.0, "num_tokens": 830122.0, "reward": 0.3749999701976776, "reward_std": 0.48552989959716797, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1763.0, "completions/max_terminated_length": 1763.0, "completions/mean_length": 383.03125, "completions/mean_terminated_length": 383.03125, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.01125, "frac_reward_zero_std": 0.0, "grad_norm": 5.271634578704834, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 924628.0, "reward": 0.19374999403953552, "reward_std": 0.2776850759983063, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.3333333432674408, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 0.0, "completions/max_length": 1895.0, "completions/max_terminated_length": 1895.0, "completions/mean_length": 612.125, "completions/mean_terminated_length": 612.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0125, "frac_reward_zero_std": 0.25, "grad_norm": 10.379047393798828, "learning_rate": 9e-07, "loss": -0.0, "num_tokens": 1064932.0, "reward": 0.20703125, "reward_std": 0.2409304976463318, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.2302463799715042, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 1811.0, "completions/max_terminated_length": 1811.0, "completions/mean_length": 483.625, "completions/mean_terminated_length": 483.625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.01375, "frac_reward_zero_std": 0.0, "grad_norm": 11.598185539245605, "learning_rate": 1e-06, "loss": 0.0, "num_tokens": 1171476.0, "reward": 0.12734374403953552, "reward_std": 0.177528977394104, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.2824873626232147, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 958.0, "completions/max_terminated_length": 958.0, "completions/mean_length": 329.09375, "completions/mean_terminated_length": 329.09375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.015, "frac_reward_zero_std": 0.25, "grad_norm": 5.369181156158447, "learning_rate": 9.999316524962345e-07, "loss": 0.0, "num_tokens": 1260026.0, "reward": 0.21171873807907104, "reward_std": 0.2331954836845398, "rewards/format_reward/mean": 0.2421875, "rewards/format_reward/std": 0.3211115300655365, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 1264.0, "completions/max_terminated_length": 1264.0, "completions/mean_length": 389.453125, "completions/mean_terminated_length": 389.453125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.01625, "frac_reward_zero_std": 0.0, "grad_norm": 8.79492473602295, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 1357183.0, "reward": 0.3945312201976776, "reward_std": 0.4425230622291565, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.3172261714935303, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.0, "completions/max_length": 1744.0, "completions/max_terminated_length": 1744.0, "completions/mean_length": 576.140625, "completions/mean_terminated_length": 576.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0175, "frac_reward_zero_std": 0.0, "grad_norm": 1.769424557685852, "learning_rate": 9.993849845741523e-07, "loss": -0.0, "num_tokens": 1488248.0, "reward": 0.3843749761581421, "reward_std": 0.39364832639694214, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.265398770570755, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1769.0, "completions/mean_length": 675.703125, "completions/mean_terminated_length": 653.920654296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01875, "frac_reward_zero_std": 0.0, "grad_norm": 3.2238948345184326, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 1621333.0, "reward": 0.23124998807907104, "reward_std": 0.39479735493659973, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 0.0, "completions/max_length": 1123.0, "completions/max_terminated_length": 1123.0, "completions/mean_length": 436.640625, "completions/mean_terminated_length": 436.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02, "frac_reward_zero_std": 0.0, "grad_norm": 7.444122314453125, "learning_rate": 9.982922465033348e-07, "loss": -0.0, "num_tokens": 1706278.0, "reward": 0.2320312261581421, "reward_std": 0.30927249789237976, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.3347255289554596, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 1502.0, "completions/max_terminated_length": 1502.0, "completions/mean_length": 445.421875, "completions/mean_terminated_length": 445.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02125, "frac_reward_zero_std": 0.0, "grad_norm": 11.103367805480957, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1816705.0, "reward": 0.2890625, "reward_std": 0.25972992181777954, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2741328477859497, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 978.0, "completions/mean_length": 469.125, "completions/mean_terminated_length": 444.0635070800781, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.0, "grad_norm": 5.469852447509766, "learning_rate": 9.966546331768192e-07, "loss": 0.0, "num_tokens": 1919785.0, "reward": 0.375, "reward_std": 0.20704269409179688, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.33184191584587097, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1197.0, "completions/max_terminated_length": 1197.0, "completions/mean_length": 389.359375, "completions/mean_terminated_length": 389.359375, "completions/min_length": 18.0, "completions/min_terminated_length": 18.0, "epoch": 0.02375, "frac_reward_zero_std": 0.0, "grad_norm": 1.9016674757003784, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 2012040.0, "reward": 0.1914062350988388, "reward_std": 0.26537030935287476, "rewards/format_reward/mean": 0.6640625, "rewards/format_reward/std": 0.29620200395584106, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 1039.0, "completions/max_terminated_length": 1039.0, "completions/mean_length": 317.625, "completions/mean_terminated_length": 317.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.125, "grad_norm": 6.792388439178467, "learning_rate": 9.944739353007341e-07, "loss": 0.0, "num_tokens": 2113872.0, "reward": 0.42265623807907104, "reward_std": 0.23530299961566925, "rewards/format_reward/mean": 0.6328125, "rewards/format_reward/std": 0.34780704975128174, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1429.0, "completions/mean_length": 353.8125, "completions/mean_terminated_length": 299.1612854003906, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02625, "frac_reward_zero_std": 0.125, "grad_norm": 8.842331886291504, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 2226644.0, "reward": 0.17812499403953552, "reward_std": 0.24255229532718658, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.3549536168575287, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1069.0, "completions/mean_length": 283.078125, "completions/mean_terminated_length": 255.06350708007812, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.0, "grad_norm": 11.046497344970703, "learning_rate": 9.917525374361911e-07, "loss": -0.0, "num_tokens": 2323505.0, "reward": 0.36640623211860657, "reward_std": 0.3313905894756317, "rewards/format_reward/mean": 0.6953125, "rewards/format_reward/std": 0.3406015634536743, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1919.0, "completions/max_terminated_length": 1919.0, "completions/mean_length": 372.9375, "completions/mean_terminated_length": 372.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02875, "frac_reward_zero_std": 0.0, "grad_norm": 12.129836082458496, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 2440597.0, "reward": 0.35624998807907104, "reward_std": 0.4260106682777405, "rewards/format_reward/mean": 0.59375, "rewards/format_reward/std": 0.3435921370983124, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1689.0, "completions/mean_length": 222.78125, "completions/mean_terminated_length": 193.80953979492188, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03, "frac_reward_zero_std": 0.0, "grad_norm": 18.581729888916016, "learning_rate": 9.884934153917996e-07, "loss": 0.0, "num_tokens": 2543559.0, "reward": 0.39921873807907104, "reward_std": 0.3741254508495331, "rewards/format_reward/mean": 0.7109375, "rewards/format_reward/std": 0.42542871832847595, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1669.0, "completions/mean_length": 219.9375, "completions/mean_terminated_length": 190.92063903808594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03125, "frac_reward_zero_std": 0.25, "grad_norm": 5.54977560043335, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 2672467.0, "reward": 0.265625, "reward_std": 0.2895887494087219, "rewards/format_reward/mean": 0.78125, "rewards/format_reward/std": 0.3549536168575287, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 470.0, "completions/mean_length": 162.234375, "completions/mean_terminated_length": 101.4032211303711, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0325, "frac_reward_zero_std": 0.0, "grad_norm": 20.450454711914062, "learning_rate": 9.847001329696652e-07, "loss": 0.0, "num_tokens": 2745754.0, "reward": 0.5093749761581421, "reward_std": 0.4139128029346466, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.29546841979026794, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 95.015625, "completions/mean_terminated_length": 95.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03375, "frac_reward_zero_std": 0.0, "grad_norm": 14.90458869934082, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 2831627.0, "reward": 0.7265625, "reward_std": 0.38551604747772217, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.28824523091316223, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1642.0, "completions/mean_length": 135.203125, "completions/mean_terminated_length": 73.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.25, "grad_norm": 16.59746551513672, "learning_rate": 9.803768380684242e-07, "loss": 0.0, "num_tokens": 2915136.0, "reward": 0.40312501788139343, "reward_std": 0.2227361798286438, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1399.0, "completions/max_terminated_length": 1399.0, "completions/mean_length": 109.875, "completions/mean_terminated_length": 109.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03625, "frac_reward_zero_std": 0.375, "grad_norm": 8.776859283447266, "learning_rate": 9.780178907671788e-07, "loss": 0.0, "num_tokens": 2999896.0, "reward": 0.15156251192092896, "reward_std": 0.1783457100391388, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.0625, "rewards/mcq_exact_match_reward/std": 0.24397502839565277, "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": 380.0, "completions/max_terminated_length": 380.0, "completions/mean_length": 25.40625, "completions/mean_terminated_length": 25.40625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0375, "frac_reward_zero_std": 0.375, "grad_norm": 13.594465255737305, "learning_rate": 9.755282581475767e-07, "loss": -0.0, "num_tokens": 3087370.0, "reward": 0.5171875357627869, "reward_std": 0.21780578792095184, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.1717960685491562, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 449.0, "completions/mean_length": 96.09375, "completions/mean_terminated_length": 33.129032135009766, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03875, "frac_reward_zero_std": 0.375, "grad_norm": 9.822609901428223, "learning_rate": 9.729086208503173e-07, "loss": 0.0, "num_tokens": 3177416.0, "reward": 0.52734375, "reward_std": 0.2801167666912079, "rewards/format_reward/mean": 0.8984375, "rewards/format_reward/std": 0.2387082874774933, "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": 187.0, "completions/max_terminated_length": 187.0, "completions/mean_length": 17.640625, "completions/mean_terminated_length": 17.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04, "frac_reward_zero_std": 0.625, "grad_norm": 21.135448455810547, "learning_rate": 9.701596950580807e-07, "loss": 0.0, "num_tokens": 3255833.0, "reward": 0.6664062738418579, "reward_std": 0.12476710975170135, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 30.0, "completions/max_terminated_length": 30.0, "completions/mean_length": 12.71875, "completions/mean_terminated_length": 12.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04125, "frac_reward_zero_std": 0.375, "grad_norm": 18.746240615844727, "learning_rate": 9.672822322997304e-07, "loss": 0.0, "num_tokens": 3356175.0, "reward": 0.5960937738418579, "reward_std": 0.2415097951889038, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 372.0, "completions/max_terminated_length": 372.0, "completions/mean_length": 18.125, "completions/mean_terminated_length": 18.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.5, "grad_norm": 10.819052696228027, "learning_rate": 9.642770192448535e-07, "loss": -0.0, "num_tokens": 3463031.0, "reward": 0.34453126788139343, "reward_std": 0.14846181869506836, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 510.0, "completions/max_terminated_length": 510.0, "completions/mean_length": 27.015625, "completions/mean_terminated_length": 27.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04375, "frac_reward_zero_std": 0.375, "grad_norm": 15.864429473876953, "learning_rate": 9.611448774886923e-07, "loss": 0.0, "num_tokens": 3537936.0, "reward": 0.2835937738418579, "reward_std": 0.2292654812335968, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.16194961965084076, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 11.03125, "completions/mean_terminated_length": 11.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.125, "grad_norm": 24.202913284301758, "learning_rate": 9.578866633275286e-07, "loss": 0.0, "num_tokens": 3640722.0, "reward": 0.45078128576278687, "reward_std": 0.37580984830856323, "rewards/format_reward/mean": 0.9140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 11.0625, "completions/mean_terminated_length": 11.0625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04625, "frac_reward_zero_std": 0.625, "grad_norm": 13.204147338867188, "learning_rate": 9.545032675245813e-07, "loss": 0.0, "num_tokens": 3717014.0, "reward": 0.5140625238418579, "reward_std": 0.15902264416217804, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 320.0, "completions/mean_length": 48.71875, "completions/mean_terminated_length": 16.984127044677734, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0475, "frac_reward_zero_std": 0.25, "grad_norm": 12.726435661315918, "learning_rate": 9.509956150664795e-07, "loss": 0.0, "num_tokens": 3796628.0, "reward": 0.48906251788139343, "reward_std": 0.25272947549819946, "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": 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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 222.0, "completions/mean_length": 53.46875, "completions/mean_terminated_length": 21.809524536132812, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04875, "frac_reward_zero_std": 0.5, "grad_norm": 21.27250099182129, "learning_rate": 9.473646649103817e-07, "loss": -0.0, "num_tokens": 3869618.0, "reward": 0.811718761920929, "reward_std": 0.21845673024654388, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.19654127955436707, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 54.0, "completions/max_terminated_length": 54.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.625, "grad_norm": 23.1184139251709, "learning_rate": 9.436114097218058e-07, "loss": 0.0, "num_tokens": 3952802.0, "reward": 0.15703126788139343, "reward_std": 0.14846181869506836, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.0625, "rewards/mcq_exact_match_reward/std": 0.24397502839565277, "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": 312.0, "completions/max_terminated_length": 312.0, "completions/mean_length": 16.09375, "completions/mean_terminated_length": 16.09375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05125, "frac_reward_zero_std": 0.75, "grad_norm": 12.768502235412598, "learning_rate": 9.397368756032444e-07, "loss": 0.0, "num_tokens": 4031320.0, "reward": 0.6101562976837158, "reward_std": 0.1323750764131546, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1512.0, "completions/max_terminated_length": 1512.0, "completions/mean_length": 50.265625, "completions/mean_terminated_length": 50.265625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0525, "frac_reward_zero_std": 0.75, "grad_norm": 23.840425491333008, "learning_rate": 9.357421218136386e-07, "loss": 0.0, "num_tokens": 4117881.0, "reward": 0.5843750238418579, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 419.0, "completions/max_terminated_length": 419.0, "completions/mean_length": 24.75, "completions/mean_terminated_length": 24.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05375, "frac_reward_zero_std": 0.625, "grad_norm": 9.856094360351562, "learning_rate": 9.316282404787869e-07, "loss": 0.0, "num_tokens": 4199609.0, "reward": 0.3890625238418579, "reward_std": 0.17584913969039917, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.20351573824882507, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.055, "frac_reward_zero_std": 0.625, "grad_norm": 13.109899520874023, "learning_rate": 9.273963562927694e-07, "loss": 0.0, "num_tokens": 4271338.0, "reward": 0.3656250238418579, "reward_std": 0.15992169082164764, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 21.0, "completions/max_terminated_length": 21.0, "completions/mean_length": 11.515625, "completions/mean_terminated_length": 11.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05625, "frac_reward_zero_std": 0.375, "grad_norm": 26.21106719970703, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 4351443.0, "reward": 0.5679687261581421, "reward_std": 0.19368097186088562, "rewards/format_reward/mean": 0.8359375, "rewards/format_reward/std": 0.30930912494659424, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 11.25, "completions/mean_terminated_length": 11.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.5, "grad_norm": 21.321683883666992, "learning_rate": 9.185832391312642e-07, "loss": 0.0, "num_tokens": 4459083.0, "reward": 0.546875, "reward_std": 0.24039676785469055, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 11.90625, "completions/mean_terminated_length": 11.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05875, "frac_reward_zero_std": 0.75, "grad_norm": 25.36126708984375, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 4542701.0, "reward": 0.2554687559604645, "reward_std": 0.06007346510887146, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 33.0, "completions/max_terminated_length": 33.0, "completions/mean_length": 14.078125, "completions/mean_terminated_length": 14.078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.5, "grad_norm": 26.997413635253906, "learning_rate": 9.093124073433462e-07, "loss": 0.0, "num_tokens": 4657762.0, "reward": 0.39531251788139343, "reward_std": 0.15809684991836548, "rewards/format_reward/mean": 0.828125, "rewards/format_reward/std": 0.3475395441055298, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.06125, "frac_reward_zero_std": 0.625, "grad_norm": 13.210586547851562, "learning_rate": 9.045084971874737e-07, "loss": 0.0, "num_tokens": 4725234.0, "reward": 0.6000000238418579, "reward_std": 0.17570313811302185, "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": 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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.03125, "completions/mean_terminated_length": 12.03125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0625, "frac_reward_zero_std": 0.375, "grad_norm": 14.75271987915039, "learning_rate": 8.995939984474623e-07, "loss": 0.0, "num_tokens": 4791620.0, "reward": 0.4281250238418579, "reward_std": 0.2414703369140625, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.06375, "frac_reward_zero_std": 0.75, "grad_norm": 16.48665428161621, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 4872452.0, "reward": 0.6000000238418579, "reward_std": 0.0883883461356163, "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": 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": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.0625, "completions/mean_terminated_length": 12.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.065, "frac_reward_zero_std": 0.5, "grad_norm": 15.117328643798828, "learning_rate": 8.894386393810562e-07, "loss": 0.0, "num_tokens": 4941224.0, "reward": 0.2718750238418579, "reward_std": 0.21778544783592224, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.06625, "frac_reward_zero_std": 0.75, "grad_norm": 14.067106246948242, "learning_rate": 8.842005554284295e-07, "loss": 0.0, "num_tokens": 5024888.0, "reward": 0.5843750238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0675, "frac_reward_zero_std": 0.875, "grad_norm": 7.727897644042969, "learning_rate": 8.788574348801674e-07, "loss": 0.0, "num_tokens": 5133184.0, "reward": 0.6781250238418579, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.06875, "frac_reward_zero_std": 0.875, "grad_norm": 5.476644515991211, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 5193313.0, "reward": 0.4906250238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.07, "frac_reward_zero_std": 0.625, "grad_norm": 17.98200225830078, "learning_rate": 8.678619553365658e-07, "loss": 0.0, "num_tokens": 5265001.0, "reward": 0.4593750238418579, "reward_std": 0.1804211586713791, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.07125, "frac_reward_zero_std": 0.875, "grad_norm": 7.063903331756592, "learning_rate": 8.622126023955445e-07, "loss": 0.0, "num_tokens": 5342409.0, "reward": 0.9593750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.859375, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0725, "frac_reward_zero_std": 0.5, "grad_norm": 13.756085395812988, "learning_rate": 8.564642241456986e-07, "loss": 0.0, "num_tokens": 5442433.0, "reward": 0.5531250238418579, "reward_std": 0.22673700749874115, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.07375, "frac_reward_zero_std": 0.75, "grad_norm": 10.853642463684082, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 5526609.0, "reward": 0.7250000238418579, "reward_std": 0.1293872892856598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.075, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 8.446767045592829e-07, "loss": 0.0, "num_tokens": 5633697.0, "reward": 0.8500000238418579, "reward_std": 0.0, "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": 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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.07625, "frac_reward_zero_std": 0.75, "grad_norm": 28.784507751464844, "learning_rate": 8.386407858128706e-07, "loss": 0.0, "num_tokens": 5703498.0, "reward": 0.5375000238418579, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0775, "frac_reward_zero_std": 0.875, "grad_norm": 9.740281105041504, "learning_rate": 8.325122860569241e-07, "loss": 0.0, "num_tokens": 5761730.0, "reward": 0.3343750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 0.0, "completions/max_length": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.07875, "frac_reward_zero_std": 0.75, "grad_norm": 18.035011291503906, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 5838435.0, "reward": 0.7562500238418579, "reward_std": 0.1293872892856598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.08, "frac_reward_zero_std": 0.875, "grad_norm": 7.4544501304626465, "learning_rate": 8.199842702516582e-07, "loss": 0.0, "num_tokens": 5908339.0, "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": 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": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.08125, "frac_reward_zero_std": 0.5, "grad_norm": 21.87096405029297, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 6001347.0, "reward": 0.5218750238418579, "reward_std": 0.23144522309303284, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0825, "frac_reward_zero_std": 0.875, "grad_norm": 5.694606304168701, "learning_rate": 8.071063563448339e-07, "loss": 0.0, "num_tokens": 6077524.0, "reward": 0.6312500238418579, "reward_std": 0.0578637570142746, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.08375, "frac_reward_zero_std": 0.75, "grad_norm": 10.359199523925781, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 6166844.0, "reward": 0.5218750238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.085, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.938926261462365e-07, "loss": 0.0, "num_tokens": 6251940.0, "reward": 0.8500000238418579, "reward_std": 0.0, "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": 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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.08625, "frac_reward_zero_std": 0.625, "grad_norm": 15.241501808166504, "learning_rate": 7.871643313414718e-07, "loss": 0.0, "num_tokens": 6330485.0, "reward": 0.5375000238418579, "reward_std": 0.1462521106004715, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0875, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.803575286758363e-07, "loss": 0.0, "num_tokens": 6404309.0, "reward": 0.7250000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.08875, "frac_reward_zero_std": 0.75, "grad_norm": 16.783206939697266, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 6484293.0, "reward": 0.5062500238418579, "reward_std": 0.1293872892856598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.09, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.665158643639969e-07, "loss": 0.0, "num_tokens": 6556669.0, "reward": 0.8500000238418579, "reward_std": 0.0, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.09125, "frac_reward_zero_std": 0.5, "grad_norm": 30.312788009643555, "learning_rate": 7.594847868906076e-07, "loss": 0.0, "num_tokens": 6630717.0, "reward": 0.7406250238418579, "reward_std": 0.2109457552433014, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0925, "frac_reward_zero_std": 0.75, "grad_norm": 11.27795696258545, "learning_rate": 7.523827688674219e-07, "loss": -0.0, "num_tokens": 6709965.0, "reward": 0.5062500238418579, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.09375, "frac_reward_zero_std": 0.75, "grad_norm": 6.554471015930176, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 6841037.0, "reward": 0.5531250238418579, "reward_std": 0.11100947856903076, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.095, "frac_reward_zero_std": 0.875, "grad_norm": 9.212408065795898, "learning_rate": 7.379736965185368e-07, "loss": 0.0, "num_tokens": 6919005.0, "reward": 0.6937500238418579, "reward_std": 0.0578637570142746, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.09625, "frac_reward_zero_std": 0.625, "grad_norm": 37.92858123779297, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 6996109.0, "reward": 0.5062500238418579, "reward_std": 0.16675157845020294, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0975, "frac_reward_zero_std": 0.625, "grad_norm": 17.949735641479492, "learning_rate": 7.233044034264033e-07, "loss": 0.0, "num_tokens": 7091661.0, "reward": 0.6937500238418579, "reward_std": 0.1872510462999344, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.09875, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 7178109.0, "reward": 0.8500000238418579, "reward_std": 0.0, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1, "frac_reward_zero_std": 0.625, "grad_norm": 20.589990615844727, "learning_rate": 7.083909302476452e-07, "loss": 0.0, "num_tokens": 7252685.0, "reward": 0.8031250238418579, "reward_std": 0.16887323558330536, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.10125, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.008477123264847e-07, "loss": 0.0, "num_tokens": 7330925.0, "reward": 0.8500000238418579, "reward_std": 0.0, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1025, "frac_reward_zero_std": 0.75, "grad_norm": 34.49396514892578, "learning_rate": 6.932495846462261e-07, "loss": 0.0, "num_tokens": 7403133.0, "reward": 0.6781250238418579, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1735.0, "completions/mean_length": 173.828125, "completions/mean_terminated_length": 144.07937622070312, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.10375, "frac_reward_zero_std": 0.875, "grad_norm": 20.851499557495117, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 7509546.0, "reward": 0.5359375476837158, "reward_std": 0.07116301357746124, "rewards/format_reward/mean": 0.828125, "rewards/format_reward/std": 0.38025420904159546, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.105, "frac_reward_zero_std": 0.875, "grad_norm": 8.781059265136719, "learning_rate": 6.778969234612583e-07, "loss": 0.0, "num_tokens": 7580522.0, "reward": 0.8343750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.10625, "frac_reward_zero_std": 0.75, "grad_norm": 7.650911331176758, "learning_rate": 6.701465872208216e-07, "loss": -0.0, "num_tokens": 7690706.0, "reward": 0.6000000238418579, "reward_std": 0.0883883461356163, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1075, "frac_reward_zero_std": 0.75, "grad_norm": 12.82839298248291, "learning_rate": 6.623497346023417e-07, "loss": 0.0, "num_tokens": 7771410.0, "reward": 0.6140625476837158, "reward_std": 0.04861358925700188, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.10875, "frac_reward_zero_std": 0.5, "grad_norm": 26.970367431640625, "learning_rate": 6.545084971874736e-07, "loss": -0.0, "num_tokens": 7840058.0, "reward": 0.5531250238418579, "reward_std": 0.19939783215522766, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11, "frac_reward_zero_std": 0.75, "grad_norm": 21.771242141723633, "learning_rate": 6.466250186922324e-07, "loss": 0.0, "num_tokens": 7942954.0, "reward": 1.006250023841858, "reward_std": 0.1293872892856598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.90625, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11125, "frac_reward_zero_std": 0.75, "grad_norm": 15.462240219116211, "learning_rate": 6.387014543809223e-07, "loss": 0.0, "num_tokens": 8039274.0, "reward": 0.7406250238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1125, "frac_reward_zero_std": 0.875, "grad_norm": 9.880819320678711, "learning_rate": 6.307399704769098e-07, "loss": -0.0, "num_tokens": 8116786.0, "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": 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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11375, "frac_reward_zero_std": 0.75, "grad_norm": 15.82265853881836, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 8196418.0, "reward": 0.4593750238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 17.0, "completions/max_terminated_length": 17.0, "completions/mean_length": 12.078125, "completions/mean_terminated_length": 12.078125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.115, "frac_reward_zero_std": 0.75, "grad_norm": 11.207847595214844, "learning_rate": 6.147119600233758e-07, "loss": 0.0, "num_tokens": 8268607.0, "reward": 0.9421875476837158, "reward_std": 0.09280776977539062, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.84375, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11625, "frac_reward_zero_std": 0.875, "grad_norm": 6.866385459899902, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 8372607.0, "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": 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": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1175, "frac_reward_zero_std": 0.875, "grad_norm": 15.375765800476074, "learning_rate": 5.985585137257401e-07, "loss": 0.0, "num_tokens": 8444655.0, "reward": 0.8187500238418579, "reward_std": 0.0578637570142746, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11875, "frac_reward_zero_std": 0.875, "grad_norm": 15.74787425994873, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 8522247.0, "reward": 0.6781250238418579, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.12, "frac_reward_zero_std": 0.75, "grad_norm": 19.906932830810547, "learning_rate": 5.82297295140367e-07, "loss": 0.0, "num_tokens": 8618319.0, "reward": 0.7718750238418579, "reward_std": 0.11100947856903076, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.12125, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 8705847.0, "reward": 0.7250000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1225, "frac_reward_zero_std": 0.875, "grad_norm": 4.607022762298584, "learning_rate": 5.659460856710345e-07, "loss": 0.0, "num_tokens": 8818503.0, "reward": 0.4593750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.12375, "frac_reward_zero_std": 0.875, "grad_norm": 5.152970790863037, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 8897183.0, "reward": 0.8343750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.125, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 5.495227651252315e-07, "loss": 0.0, "num_tokens": 8980063.0, "reward": 0.7250000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 100 } ], "logging_steps": 1, "max_steps": 200, "num_input_tokens_seen": 8980063, "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }