{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.15060240963855423, "eval_steps": 500, "global_step": 200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1461.0, "completions/mean_length": 539.939453125, "completions/mean_terminated_length": 534.0687866210938, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.0007530120481927711, "frac_reward_zero_std": 0.0, "grad_norm": 0.24425463378429413, "learning_rate": 1e-06, "loss": 0.0483, "num_tokens": 322497.0, "reward": 18.535240173339844, "reward_std": 4.5626091957092285, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.535240173339844, "rewards/skywork_reward/std": 7.240255355834961, "step": 1 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1313.0, "completions/mean_length": 539.921875, "completions/mean_terminated_length": 537.9725952148438, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.0015060240963855422, "frac_reward_zero_std": 0.0, "grad_norm": 0.2527785897254944, "learning_rate": 1e-06, "loss": 0.0676, "num_tokens": 642601.0, "reward": 19.66384506225586, "reward_std": 4.405518531799316, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.66384506225586, "rewards/skywork_reward/std": 7.860788822174072, "step": 2 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1096.0, "completions/max_terminated_length": 1096.0, "completions/mean_length": 509.203125, "completions/mean_terminated_length": 509.203125, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.002259036144578313, "frac_reward_zero_std": 0.0, "grad_norm": 0.23311284184455872, "learning_rate": 1e-06, "loss": 0.036, "num_tokens": 948897.0, "reward": 19.736499786376953, "reward_std": 4.317551136016846, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.736499786376953, "rewards/skywork_reward/std": 8.241270065307617, "step": 3 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1332.0, "completions/max_terminated_length": 1332.0, "completions/mean_length": 481.34765625, "completions/mean_terminated_length": 481.34765625, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.0030120481927710845, "frac_reward_zero_std": 0.0, "grad_norm": 0.25114020705223083, "learning_rate": 1e-06, "loss": 0.037, "num_tokens": 1240675.0, "reward": 19.980758666992188, "reward_std": 4.497593879699707, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.980758666992188, "rewards/skywork_reward/std": 7.561793804168701, "step": 4 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1367.0, "completions/mean_length": 550.08984375, "completions/mean_terminated_length": 548.1604614257812, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.0037650602409638554, "frac_reward_zero_std": 0.0, "grad_norm": 0.27846774458885193, "learning_rate": 1e-06, "loss": 0.06, "num_tokens": 1567393.0, "reward": 15.45170783996582, "reward_std": 4.615124702453613, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 15.45170783996582, "rewards/skywork_reward/std": 7.330972194671631, "step": 5 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1435.0, "completions/mean_length": 466.416015625, "completions/mean_terminated_length": 462.2215881347656, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.004518072289156626, "frac_reward_zero_std": 0.0, "grad_norm": 0.26218029856681824, "learning_rate": 1e-06, "loss": 0.0554, "num_tokens": 1848294.0, "reward": 15.858068466186523, "reward_std": 4.51039981842041, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 15.858068466186523, "rewards/skywork_reward/std": 8.340972900390625, "step": 6 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1397.0, "completions/mean_length": 546.90625, "completions/mean_terminated_length": 544.9706420898438, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.005271084337349397, "frac_reward_zero_std": 0.0, "grad_norm": 0.2502232789993286, "learning_rate": 1e-06, "loss": 0.0183, "num_tokens": 2176214.0, "reward": 20.738807678222656, "reward_std": 4.506719589233398, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.738807678222656, "rewards/skywork_reward/std": 7.784599781036377, "step": 7 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1344.0, "completions/mean_length": 507.919921875, "completions/mean_terminated_length": 505.90802001953125, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.006024096385542169, "frac_reward_zero_std": 0.0, "grad_norm": 0.24587783217430115, "learning_rate": 1e-06, "loss": 0.0108, "num_tokens": 2483485.0, "reward": 18.190261840820312, "reward_std": 4.080215930938721, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.190261840820312, "rewards/skywork_reward/std": 7.144257068634033, "step": 8 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1392.0, "completions/mean_length": 517.755859375, "completions/mean_terminated_length": 513.7627563476562, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.00677710843373494, "frac_reward_zero_std": 0.0, "grad_norm": 0.25550204515457153, "learning_rate": 1e-06, "loss": 0.0365, "num_tokens": 2791968.0, "reward": 18.123830795288086, "reward_std": 3.959782838821411, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.123830795288086, "rewards/skywork_reward/std": 8.163817405700684, "step": 9 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1385.0, "completions/mean_length": 469.869140625, "completions/mean_terminated_length": 467.78277587890625, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.007530120481927711, "frac_reward_zero_std": 0.0, "grad_norm": 0.2836010456085205, "learning_rate": 1e-06, "loss": 0.0308, "num_tokens": 3079901.0, "reward": 19.304351806640625, "reward_std": 4.372314453125, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.304351806640625, "rewards/skywork_reward/std": 7.716484069824219, "step": 10 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1485.0, "completions/max_terminated_length": 1485.0, "completions/mean_length": 500.615234375, "completions/mean_terminated_length": 500.615234375, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.008283132530120483, "frac_reward_zero_std": 0.0, "grad_norm": 0.246844083070755, "learning_rate": 1e-06, "loss": 0.0386, "num_tokens": 3382520.0, "reward": 18.477981567382812, "reward_std": 4.108589172363281, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.477981567382812, "rewards/skywork_reward/std": 7.084549427032471, "step": 11 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1090.0, "completions/max_terminated_length": 1090.0, "completions/mean_length": 478.2578125, "completions/mean_terminated_length": 478.2578125, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "epoch": 0.009036144578313253, "frac_reward_zero_std": 0.0, "grad_norm": 0.2760355472564697, "learning_rate": 1e-06, "loss": 0.0437, "num_tokens": 3669020.0, "reward": 18.244699478149414, "reward_std": 4.524354457855225, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.244699478149414, "rewards/skywork_reward/std": 7.753305435180664, "step": 12 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1359.0, "completions/mean_length": 509.525390625, "completions/mean_terminated_length": 505.5000305175781, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.009789156626506024, "frac_reward_zero_std": 0.0, "grad_norm": 0.2421659678220749, "learning_rate": 1e-06, "loss": 0.0431, "num_tokens": 3973577.0, "reward": 18.84097671508789, "reward_std": 3.6833090782165527, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.84097671508789, "rewards/skywork_reward/std": 7.0365753173828125, "step": 13 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 986.0, "completions/max_terminated_length": 986.0, "completions/mean_length": 502.75, "completions/mean_terminated_length": 502.75, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "epoch": 0.010542168674698794, "frac_reward_zero_std": 0.0, "grad_norm": 0.24300047755241394, "learning_rate": 1e-06, "loss": 0.0095, "num_tokens": 4275273.0, "reward": 19.73338508605957, "reward_std": 3.7505850791931152, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.73338508605957, "rewards/skywork_reward/std": 7.851949691772461, "step": 14 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1212.0, "completions/mean_length": 495.01171875, "completions/mean_terminated_length": 492.97454833984375, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.011295180722891566, "frac_reward_zero_std": 0.0, "grad_norm": 0.24820354580879211, "learning_rate": 1e-06, "loss": 0.041, "num_tokens": 4572607.0, "reward": 20.985137939453125, "reward_std": 3.9907798767089844, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.985137939453125, "rewards/skywork_reward/std": 7.5082783699035645, "step": 15 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1452.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 503.646484375, "completions/mean_terminated_length": 503.646484375, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.012048192771084338, "frac_reward_zero_std": 0.0, "grad_norm": 0.25032636523246765, "learning_rate": 1e-06, "loss": 0.0271, "num_tokens": 4879242.0, "reward": 19.479820251464844, "reward_std": 3.9832570552825928, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.479820251464844, "rewards/skywork_reward/std": 7.7815022468566895, "step": 16 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1505.0, "completions/mean_length": 548.275390625, "completions/mean_terminated_length": 546.3424682617188, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.012801204819277108, "frac_reward_zero_std": 0.0, "grad_norm": 0.23289093375205994, "learning_rate": 1e-06, "loss": 0.0344, "num_tokens": 5206471.0, "reward": 18.09130859375, "reward_std": 4.67989444732666, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.09130859375, "rewards/skywork_reward/std": 8.046331405639648, "step": 17 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1333.0, "completions/max_terminated_length": 1333.0, "completions/mean_length": 500.513671875, "completions/mean_terminated_length": 500.513671875, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.01355421686746988, "frac_reward_zero_std": 0.0, "grad_norm": 0.2743029296398163, "learning_rate": 1e-06, "loss": 0.0608, "num_tokens": 5508110.0, "reward": 19.004474639892578, "reward_std": 4.812801361083984, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.004474639892578, "rewards/skywork_reward/std": 7.56382942199707, "step": 18 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1300.0, "completions/mean_length": 493.69921875, "completions/mean_terminated_length": 491.65948486328125, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.01430722891566265, "frac_reward_zero_std": 0.0, "grad_norm": 0.24759620428085327, "learning_rate": 1e-06, "loss": 0.0399, "num_tokens": 5802532.0, "reward": 20.18354034423828, "reward_std": 4.097359657287598, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.18354034423828, "rewards/skywork_reward/std": 6.95109224319458, "step": 19 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 540.8828125, "completions/mean_terminated_length": 536.9804077148438, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.015060240963855422, "frac_reward_zero_std": 0.0, "grad_norm": 0.22581446170806885, "learning_rate": 1e-06, "loss": 0.0349, "num_tokens": 6127912.0, "reward": 18.810096740722656, "reward_std": 4.296095848083496, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.810096740722656, "rewards/skywork_reward/std": 7.788106918334961, "step": 20 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1486.0, "completions/mean_length": 568.685546875, "completions/mean_terminated_length": 562.9843139648438, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.01581325301204819, "frac_reward_zero_std": 0.0, "grad_norm": 0.25190842151641846, "learning_rate": 1e-06, "loss": 0.031, "num_tokens": 6462007.0, "reward": 16.9500675201416, "reward_std": 3.961681842803955, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 16.9500675201416, "rewards/skywork_reward/std": 7.946998119354248, "step": 21 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1516.0, "completions/mean_length": 536.982421875, "completions/mean_terminated_length": 535.0274047851562, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.016566265060240965, "frac_reward_zero_std": 0.0, "grad_norm": 0.23905393481254578, "learning_rate": 1e-06, "loss": 0.0237, "num_tokens": 6777998.0, "reward": 19.38282012939453, "reward_std": 3.568166494369507, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.38282012939453, "rewards/skywork_reward/std": 8.19947624206543, "step": 22 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1442.0, "completions/mean_length": 564.58203125, "completions/mean_terminated_length": 562.6810302734375, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.017319277108433735, "frac_reward_zero_std": 0.0, "grad_norm": 0.24202950298786163, "learning_rate": 1e-06, "loss": 0.0379, "num_tokens": 7113960.0, "reward": 17.542194366455078, "reward_std": 3.6086037158966064, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 17.542194366455078, "rewards/skywork_reward/std": 6.972787857055664, "step": 23 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.009765625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1526.0, "completions/mean_length": 594.02734375, "completions/mean_terminated_length": 584.7376708984375, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.018072289156626505, "frac_reward_zero_std": 0.0, "grad_norm": 0.23193013668060303, "learning_rate": 1e-06, "loss": 0.0177, "num_tokens": 7464662.0, "reward": 19.02389907836914, "reward_std": 4.250513076782227, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.02389907836914, "rewards/skywork_reward/std": 6.219736576080322, "step": 24 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1258.0, "completions/mean_length": 549.28125, "completions/mean_terminated_length": 547.3502807617188, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.01882530120481928, "frac_reward_zero_std": 0.0, "grad_norm": 0.24706406891345978, "learning_rate": 1e-06, "loss": 0.0544, "num_tokens": 7789846.0, "reward": 18.385107040405273, "reward_std": 4.854748249053955, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.385107040405273, "rewards/skywork_reward/std": 7.241014003753662, "step": 25 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.009765625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1387.0, "completions/mean_length": 531.689453125, "completions/mean_terminated_length": 521.7850341796875, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.01957831325301205, "frac_reward_zero_std": 0.0, "grad_norm": 0.2525196373462677, "learning_rate": 1e-06, "loss": 0.0523, "num_tokens": 8111927.0, "reward": 19.272109985351562, "reward_std": 3.762904167175293, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.272109985351562, "rewards/skywork_reward/std": 7.09940767288208, "step": 26 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1308.0, "completions/mean_length": 547.365234375, "completions/mean_terminated_length": 545.4305419921875, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.02033132530120482, "frac_reward_zero_std": 0.0, "grad_norm": 0.2493782937526703, "learning_rate": 1e-06, "loss": 0.0507, "num_tokens": 8435202.0, "reward": 17.78737449645996, "reward_std": 4.015490531921387, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 17.78737449645996, "rewards/skywork_reward/std": 7.691453456878662, "step": 27 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1301.0, "completions/max_terminated_length": 1301.0, "completions/mean_length": 478.685546875, "completions/mean_terminated_length": 478.685546875, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.02108433734939759, "frac_reward_zero_std": 0.0, "grad_norm": 0.26527339220046997, "learning_rate": 1e-06, "loss": 0.0339, "num_tokens": 8727633.0, "reward": 18.366004943847656, "reward_std": 3.85819411277771, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.366004943847656, "rewards/skywork_reward/std": 6.539942264556885, "step": 28 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1473.0, "completions/max_terminated_length": 1473.0, "completions/mean_length": 509.15234375, "completions/mean_terminated_length": 509.15234375, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.021837349397590362, "frac_reward_zero_std": 0.0, "grad_norm": 0.2530669867992401, "learning_rate": 1e-06, "loss": 0.0226, "num_tokens": 9036511.0, "reward": 18.36314582824707, "reward_std": 3.6422839164733887, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.36314582824707, "rewards/skywork_reward/std": 7.377774715423584, "step": 29 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 523.609375, "completions/mean_terminated_length": 521.628173828125, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.022590361445783132, "frac_reward_zero_std": 0.0, "grad_norm": 0.24128130078315735, "learning_rate": 1e-06, "loss": 0.0412, "num_tokens": 9345751.0, "reward": 20.08295440673828, "reward_std": 3.875840902328491, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.08295440673828, "rewards/skywork_reward/std": 7.567595958709717, "step": 30 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1503.0, "completions/mean_length": 516.138671875, "completions/mean_terminated_length": 512.1392211914062, "completions/min_length": 169.0, "completions/min_terminated_length": 169.0, "epoch": 0.023343373493975902, "frac_reward_zero_std": 0.0, "grad_norm": 0.2687889337539673, "learning_rate": 1e-06, "loss": 0.0406, "num_tokens": 9651646.0, "reward": 18.440399169921875, "reward_std": 4.034783363342285, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.440399169921875, "rewards/skywork_reward/std": 9.241204261779785, "step": 31 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1351.0, "completions/max_terminated_length": 1351.0, "completions/mean_length": 509.87109375, "completions/mean_terminated_length": 509.87109375, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.024096385542168676, "frac_reward_zero_std": 0.0, "grad_norm": 0.2565000653266907, "learning_rate": 1e-06, "loss": 0.0341, "num_tokens": 9958156.0, "reward": 21.218032836914062, "reward_std": 3.8647594451904297, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.218032836914062, "rewards/skywork_reward/std": 7.692581653594971, "step": 32 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1272.0, "completions/mean_length": 483.21484375, "completions/mean_terminated_length": 481.15460205078125, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.024849397590361446, "frac_reward_zero_std": 0.0, "grad_norm": 0.2624817490577698, "learning_rate": 1e-06, "loss": 0.0431, "num_tokens": 10249610.0, "reward": 19.79848861694336, "reward_std": 4.353243827819824, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.79848861694336, "rewards/skywork_reward/std": 7.000552654266357, "step": 33 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 479.90234375, "completions/mean_terminated_length": 479.90234375, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.025602409638554216, "frac_reward_zero_std": 0.0, "grad_norm": 0.2779456377029419, "learning_rate": 1e-06, "loss": 0.0414, "num_tokens": 10539784.0, "reward": 18.208389282226562, "reward_std": 3.7550158500671387, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.208389282226562, "rewards/skywork_reward/std": 7.102542400360107, "step": 34 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1438.0, "completions/max_terminated_length": 1438.0, "completions/mean_length": 538.0859375, "completions/mean_terminated_length": 538.0859375, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.02635542168674699, "frac_reward_zero_std": 0.0, "grad_norm": 0.27090051770210266, "learning_rate": 1e-06, "loss": 0.0616, "num_tokens": 10861060.0, "reward": 18.551515579223633, "reward_std": 4.775388717651367, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.551515579223633, "rewards/skywork_reward/std": 8.144341468811035, "step": 35 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1414.0, "completions/mean_length": 523.318359375, "completions/mean_terminated_length": 521.3366088867188, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.02710843373493976, "frac_reward_zero_std": 0.0, "grad_norm": 0.24933283030986786, "learning_rate": 1e-06, "loss": 0.0219, "num_tokens": 11174711.0, "reward": 19.23906707763672, "reward_std": 3.579415798187256, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.23906707763672, "rewards/skywork_reward/std": 6.7693352699279785, "step": 36 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1471.0, "completions/max_terminated_length": 1471.0, "completions/mean_length": 540.38671875, "completions/mean_terminated_length": 540.38671875, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.02786144578313253, "frac_reward_zero_std": 0.0, "grad_norm": 0.2600509524345398, "learning_rate": 1e-06, "loss": 0.0552, "num_tokens": 11499629.0, "reward": 17.48117446899414, "reward_std": 4.680095672607422, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 17.48117446899414, "rewards/skywork_reward/std": 7.522308826446533, "step": 37 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1324.0, "completions/mean_length": 527.3359375, "completions/mean_terminated_length": 525.362060546875, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.0286144578313253, "frac_reward_zero_std": 0.0, "grad_norm": 0.2569137215614319, "learning_rate": 1e-06, "loss": 0.0256, "num_tokens": 11812185.0, "reward": 19.502559661865234, "reward_std": 4.2003655433654785, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.502559661865234, "rewards/skywork_reward/std": 7.574954986572266, "step": 38 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1497.0, "completions/mean_length": 546.02734375, "completions/mean_terminated_length": 540.1925659179688, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "epoch": 0.029367469879518073, "frac_reward_zero_std": 0.0, "grad_norm": 0.25866448879241943, "learning_rate": 1e-06, "loss": 0.0702, "num_tokens": 12136471.0, "reward": 19.45880889892578, "reward_std": 4.22914457321167, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.45880889892578, "rewards/skywork_reward/std": 7.417150497436523, "step": 39 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1300.0, "completions/mean_length": 549.212890625, "completions/mean_terminated_length": 543.3968505859375, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.030120481927710843, "frac_reward_zero_std": 0.0, "grad_norm": 0.26242557168006897, "learning_rate": 1e-06, "loss": 0.0572, "num_tokens": 12468676.0, "reward": 18.719432830810547, "reward_std": 4.223837852478027, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.719432830810547, "rewards/skywork_reward/std": 7.981883525848389, "step": 40 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1474.0, "completions/mean_length": 592.703125, "completions/mean_terminated_length": 587.1434326171875, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.030873493975903613, "frac_reward_zero_std": 0.0, "grad_norm": 0.23195688426494598, "learning_rate": 1e-06, "loss": 0.0277, "num_tokens": 12819564.0, "reward": 19.721298217773438, "reward_std": 4.098422527313232, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.721298217773438, "rewards/skywork_reward/std": 8.307689666748047, "step": 41 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1385.0, "completions/mean_length": 505.2734375, "completions/mean_terminated_length": 501.2314147949219, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.03162650602409638, "frac_reward_zero_std": 0.0, "grad_norm": 0.26936817169189453, "learning_rate": 1e-06, "loss": 0.0703, "num_tokens": 13120760.0, "reward": 19.512004852294922, "reward_std": 4.078932762145996, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.512004852294922, "rewards/skywork_reward/std": 7.664077281951904, "step": 42 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1478.0, "completions/mean_length": 525.787109375, "completions/mean_terminated_length": 523.8101806640625, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.03237951807228916, "frac_reward_zero_std": 0.0, "grad_norm": 0.25450485944747925, "learning_rate": 1e-06, "loss": 0.0453, "num_tokens": 13435275.0, "reward": 19.4080810546875, "reward_std": 4.121971130371094, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.4080810546875, "rewards/skywork_reward/std": 8.246720314025879, "step": 43 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.025390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1514.0, "completions/mean_length": 598.4140625, "completions/mean_terminated_length": 573.9879760742188, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.03313253012048193, "frac_reward_zero_std": 0.0, "grad_norm": 0.22897975146770477, "learning_rate": 1e-06, "loss": 0.0552, "num_tokens": 13792623.0, "reward": 19.808815002441406, "reward_std": 4.206421375274658, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.808815002441406, "rewards/skywork_reward/std": 8.275568008422852, "step": 44 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.02734375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1531.0, "completions/mean_length": 540.21875, "completions/mean_terminated_length": 512.224853515625, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.0338855421686747, "frac_reward_zero_std": 0.0, "grad_norm": 0.26464077830314636, "learning_rate": 1e-06, "loss": 0.084, "num_tokens": 14114543.0, "reward": 19.398923873901367, "reward_std": 4.753540992736816, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.398923873901367, "rewards/skywork_reward/std": 8.964569091796875, "step": 45 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1331.0, "completions/mean_length": 550.970703125, "completions/mean_terminated_length": 543.2145385742188, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.03463855421686747, "frac_reward_zero_std": 0.0, "grad_norm": 0.26099711656570435, "learning_rate": 1e-06, "loss": 0.0685, "num_tokens": 14438880.0, "reward": 20.744400024414062, "reward_std": 4.4266676902771, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.744400024414062, "rewards/skywork_reward/std": 9.021385192871094, "step": 46 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1476.0, "completions/mean_length": 555.751953125, "completions/mean_terminated_length": 553.8336791992188, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.035391566265060244, "frac_reward_zero_std": 0.0, "grad_norm": 0.23918218910694122, "learning_rate": 1e-06, "loss": 0.0153, "num_tokens": 14773041.0, "reward": 18.19009017944336, "reward_std": 3.2478275299072266, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.19009017944336, "rewards/skywork_reward/std": 8.022202491760254, "step": 47 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1381.0, "completions/mean_length": 504.818359375, "completions/mean_terminated_length": 502.8003845214844, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.03614457831325301, "frac_reward_zero_std": 0.0, "grad_norm": 0.2639124095439911, "learning_rate": 1e-06, "loss": 0.0118, "num_tokens": 15076308.0, "reward": 18.888355255126953, "reward_std": 3.525162696838379, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.888355255126953, "rewards/skywork_reward/std": 7.233704090118408, "step": 48 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1477.0, "completions/mean_length": 558.080078125, "completions/mean_terminated_length": 552.3163452148438, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.036897590361445784, "frac_reward_zero_std": 0.0, "grad_norm": 0.2637381851673126, "learning_rate": 1e-06, "loss": 0.061, "num_tokens": 15408557.0, "reward": 18.518508911132812, "reward_std": 4.11885929107666, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.518508911132812, "rewards/skywork_reward/std": 7.327213287353516, "step": 49 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1511.0, "completions/mean_length": 519.8359375, "completions/mean_terminated_length": 503.7063903808594, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.03765060240963856, "frac_reward_zero_std": 0.0, "grad_norm": 0.2628968060016632, "learning_rate": 1e-06, "loss": 0.052, "num_tokens": 15719481.0, "reward": 19.84031105041504, "reward_std": 3.317889928817749, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.84031105041504, "rewards/skywork_reward/std": 7.0079755783081055, "step": 50 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1354.0, "completions/max_terminated_length": 1354.0, "completions/mean_length": 460.279296875, "completions/mean_terminated_length": 460.279296875, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.038403614457831324, "frac_reward_zero_std": 0.0, "grad_norm": 0.279048889875412, "learning_rate": 1e-06, "loss": 0.0238, "num_tokens": 16001848.0, "reward": 20.461505889892578, "reward_std": 3.4711790084838867, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.461505889892578, "rewards/skywork_reward/std": 6.913259506225586, "step": 51 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1500.0, "completions/max_terminated_length": 1500.0, "completions/mean_length": 550.14453125, "completions/mean_terminated_length": 550.14453125, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.0391566265060241, "frac_reward_zero_std": 0.0, "grad_norm": 0.24522554874420166, "learning_rate": 1e-06, "loss": 0.0237, "num_tokens": 16331186.0, "reward": 19.905433654785156, "reward_std": 3.6912999153137207, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.905433654785156, "rewards/skywork_reward/std": 7.08914852142334, "step": 52 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 586.515625, "completions/mean_terminated_length": 582.7921752929688, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.039909638554216864, "frac_reward_zero_std": 0.0, "grad_norm": 0.2638297379016876, "learning_rate": 1e-06, "loss": 0.0588, "num_tokens": 16675610.0, "reward": 19.24047088623047, "reward_std": 4.73069953918457, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.24047088623047, "rewards/skywork_reward/std": 8.525131225585938, "step": 53 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1396.0, "completions/max_terminated_length": 1396.0, "completions/mean_length": 524.859375, "completions/mean_terminated_length": 524.859375, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.04066265060240964, "frac_reward_zero_std": 0.0, "grad_norm": 0.27795007824897766, "learning_rate": 1e-06, "loss": 0.033, "num_tokens": 16987650.0, "reward": 18.526611328125, "reward_std": 4.1653361320495605, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.526611328125, "rewards/skywork_reward/std": 6.746675968170166, "step": 54 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1436.0, "completions/mean_length": 551.83203125, "completions/mean_terminated_length": 549.9060668945312, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.04141566265060241, "frac_reward_zero_std": 0.0, "grad_norm": 0.26270273327827454, "learning_rate": 1e-06, "loss": 0.0523, "num_tokens": 17316684.0, "reward": 18.824560165405273, "reward_std": 3.540390968322754, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.824560165405273, "rewards/skywork_reward/std": 8.527042388916016, "step": 55 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1397.0, "completions/mean_length": 575.982421875, "completions/mean_terminated_length": 572.2176513671875, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.04216867469879518, "frac_reward_zero_std": 0.0, "grad_norm": 0.24960987269878387, "learning_rate": 1e-06, "loss": 0.0122, "num_tokens": 17659123.0, "reward": 19.136398315429688, "reward_std": 3.8432905673980713, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.136398315429688, "rewards/skywork_reward/std": 8.282490730285645, "step": 56 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1534.0, "completions/mean_length": 542.234375, "completions/mean_terminated_length": 540.2896118164062, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.04292168674698795, "frac_reward_zero_std": 0.0, "grad_norm": 0.25888824462890625, "learning_rate": 1e-06, "loss": 0.0212, "num_tokens": 17983035.0, "reward": 19.597389221191406, "reward_std": 3.953815221786499, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.597389221191406, "rewards/skywork_reward/std": 8.315513610839844, "step": 57 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1502.0, "completions/max_terminated_length": 1502.0, "completions/mean_length": 521.236328125, "completions/mean_terminated_length": 521.236328125, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.043674698795180725, "frac_reward_zero_std": 0.0, "grad_norm": 0.2533751130104065, "learning_rate": 1e-06, "loss": 0.0332, "num_tokens": 18294164.0, "reward": 20.750991821289062, "reward_std": 3.6470861434936523, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.750991821289062, "rewards/skywork_reward/std": 6.336122035980225, "step": 58 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1522.0, "completions/mean_length": 568.337890625, "completions/mean_terminated_length": 562.6345825195312, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.04442771084337349, "frac_reward_zero_std": 0.0, "grad_norm": 0.2571529150009155, "learning_rate": 1e-06, "loss": 0.0495, "num_tokens": 18633617.0, "reward": 17.738445281982422, "reward_std": 3.885385751724243, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 17.738445281982422, "rewards/skywork_reward/std": 7.068282127380371, "step": 59 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1509.0, "completions/mean_length": 504.2734375, "completions/mean_terminated_length": 502.25439453125, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.045180722891566265, "frac_reward_zero_std": 0.0, "grad_norm": 0.2818680703639984, "learning_rate": 1e-06, "loss": 0.0518, "num_tokens": 18939565.0, "reward": 19.56861114501953, "reward_std": 3.992223024368286, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.56861114501953, "rewards/skywork_reward/std": 7.308934688568115, "step": 60 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1499.0, "completions/mean_length": 550.580078125, "completions/mean_terminated_length": 546.7156982421875, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.04593373493975904, "frac_reward_zero_std": 0.0, "grad_norm": 0.2757847309112549, "learning_rate": 1e-06, "loss": 0.0732, "num_tokens": 19263446.0, "reward": 18.18868637084961, "reward_std": 4.381465911865234, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.18868637084961, "rewards/skywork_reward/std": 7.37313175201416, "step": 61 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1189.0, "completions/max_terminated_length": 1189.0, "completions/mean_length": 484.359375, "completions/mean_terminated_length": 484.359375, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.046686746987951805, "frac_reward_zero_std": 0.0, "grad_norm": 0.2803126275539398, "learning_rate": 1e-06, "loss": 0.0489, "num_tokens": 19560014.0, "reward": 20.43779754638672, "reward_std": 4.052783489227295, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.43779754638672, "rewards/skywork_reward/std": 6.660309314727783, "step": 62 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1449.0, "completions/max_terminated_length": 1449.0, "completions/mean_length": 572.158203125, "completions/mean_terminated_length": 572.158203125, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.04743975903614458, "frac_reward_zero_std": 0.0, "grad_norm": 0.25800344347953796, "learning_rate": 1e-06, "loss": 0.0317, "num_tokens": 19902143.0, "reward": 20.61004638671875, "reward_std": 3.9054417610168457, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.61004638671875, "rewards/skywork_reward/std": 7.143301010131836, "step": 63 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1434.0, "completions/max_terminated_length": 1434.0, "completions/mean_length": 531.67578125, "completions/mean_terminated_length": 531.67578125, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.04819277108433735, "frac_reward_zero_std": 0.0, "grad_norm": 0.2537233829498291, "learning_rate": 1e-06, "loss": 0.013, "num_tokens": 20217081.0, "reward": 22.289291381835938, "reward_std": 3.0658791065216064, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.289291381835938, "rewards/skywork_reward/std": 7.240992069244385, "step": 64 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 574.796875, "completions/mean_terminated_length": 559.5397338867188, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.04894578313253012, "frac_reward_zero_std": 0.0, "grad_norm": 0.2763891816139221, "learning_rate": 1e-06, "loss": 0.0475, "num_tokens": 20557233.0, "reward": 21.378662109375, "reward_std": 4.32180118560791, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.378662109375, "rewards/skywork_reward/std": 8.400932312011719, "step": 65 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1176.0, "completions/max_terminated_length": 1176.0, "completions/mean_length": 519.951171875, "completions/mean_terminated_length": 519.951171875, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.04969879518072289, "frac_reward_zero_std": 0.0, "grad_norm": 0.2865884304046631, "learning_rate": 1e-06, "loss": 0.0077, "num_tokens": 20867336.0, "reward": 22.28472900390625, "reward_std": 3.520258665084839, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.28472900390625, "rewards/skywork_reward/std": 7.874880313873291, "step": 66 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1073.0, "completions/max_terminated_length": 1073.0, "completions/mean_length": 483.09765625, "completions/mean_terminated_length": 483.09765625, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.050451807228915665, "frac_reward_zero_std": 0.0, "grad_norm": 0.30400651693344116, "learning_rate": 1e-06, "loss": 0.0363, "num_tokens": 21158026.0, "reward": 20.857650756835938, "reward_std": 3.738304615020752, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.857650756835938, "rewards/skywork_reward/std": 7.2754058837890625, "step": 67 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1195.0, "completions/max_terminated_length": 1195.0, "completions/mean_length": 529.998046875, "completions/mean_terminated_length": 529.998046875, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.05120481927710843, "frac_reward_zero_std": 0.0, "grad_norm": 0.2766510546207428, "learning_rate": 1e-06, "loss": 0.0479, "num_tokens": 21474633.0, "reward": 20.8536434173584, "reward_std": 3.9283032417297363, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.8536434173584, "rewards/skywork_reward/std": 7.341203689575195, "step": 68 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1532.0, "completions/mean_length": 554.416015625, "completions/mean_terminated_length": 550.5667114257812, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.051957831325301206, "frac_reward_zero_std": 0.0, "grad_norm": 0.2611565887928009, "learning_rate": 1e-06, "loss": 0.0488, "num_tokens": 21798510.0, "reward": 19.972572326660156, "reward_std": 3.5066475868225098, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.972572326660156, "rewards/skywork_reward/std": 7.777656555175781, "step": 69 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1522.0, "completions/mean_length": 555.4453125, "completions/mean_terminated_length": 549.666015625, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.05271084337349398, "frac_reward_zero_std": 0.0, "grad_norm": 0.24628393352031708, "learning_rate": 1e-06, "loss": 0.0382, "num_tokens": 22124706.0, "reward": 21.486955642700195, "reward_std": 3.622591257095337, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.486955642700195, "rewards/skywork_reward/std": 7.51347541809082, "step": 70 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1233.0, "completions/mean_length": 549.400390625, "completions/mean_terminated_length": 547.4696655273438, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.053463855421686746, "frac_reward_zero_std": 0.0, "grad_norm": 0.26588577032089233, "learning_rate": 1e-06, "loss": 0.0084, "num_tokens": 22455471.0, "reward": 20.11474609375, "reward_std": 3.2167859077453613, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.11474609375, "rewards/skywork_reward/std": 6.900139808654785, "step": 71 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1421.0, "completions/mean_length": 515.904296875, "completions/mean_terminated_length": 513.9080200195312, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.05421686746987952, "frac_reward_zero_std": 0.0, "grad_norm": 0.28270772099494934, "learning_rate": 1e-06, "loss": 0.0185, "num_tokens": 22765118.0, "reward": 21.868011474609375, "reward_std": 3.350029468536377, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.868011474609375, "rewards/skywork_reward/std": 6.04210090637207, "step": 72 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1175.0, "completions/max_terminated_length": 1175.0, "completions/mean_length": 571.068359375, "completions/mean_terminated_length": 571.068359375, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.054969879518072286, "frac_reward_zero_std": 0.0, "grad_norm": 0.263434499502182, "learning_rate": 1e-06, "loss": 0.0321, "num_tokens": 23103025.0, "reward": 19.462799072265625, "reward_std": 3.916111707687378, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.462799072265625, "rewards/skywork_reward/std": 8.177480697631836, "step": 73 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1504.0, "completions/mean_length": 615.650390625, "completions/mean_terminated_length": 612.0411987304688, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.05572289156626506, "frac_reward_zero_std": 0.0, "grad_norm": 0.2504398822784424, "learning_rate": 1e-06, "loss": 0.024, "num_tokens": 23460030.0, "reward": 22.602195739746094, "reward_std": 4.115928649902344, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.602195739746094, "rewards/skywork_reward/std": 7.8947224617004395, "step": 74 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1517.0, "completions/max_terminated_length": 1517.0, "completions/mean_length": 550.38671875, "completions/mean_terminated_length": 550.38671875, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.05647590361445783, "frac_reward_zero_std": 0.0, "grad_norm": 0.27360963821411133, "learning_rate": 1e-06, "loss": 0.0359, "num_tokens": 23787700.0, "reward": 19.700706481933594, "reward_std": 3.7519848346710205, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.700706481933594, "rewards/skywork_reward/std": 6.8410749435424805, "step": 75 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1494.0, "completions/max_terminated_length": 1494.0, "completions/mean_length": 555.943359375, "completions/mean_terminated_length": 555.943359375, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.0572289156626506, "frac_reward_zero_std": 0.0, "grad_norm": 0.2818851172924042, "learning_rate": 1e-06, "loss": 0.048, "num_tokens": 24116295.0, "reward": 20.025314331054688, "reward_std": 3.705512285232544, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.025314331054688, "rewards/skywork_reward/std": 6.649584770202637, "step": 76 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1384.0, "completions/mean_length": 556.890625, "completions/mean_terminated_length": 554.9745483398438, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.05798192771084337, "frac_reward_zero_std": 0.0, "grad_norm": 0.29094648361206055, "learning_rate": 1e-06, "loss": 0.0675, "num_tokens": 24448271.0, "reward": 19.403905868530273, "reward_std": 3.9135756492614746, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.403905868530273, "rewards/skywork_reward/std": 7.241198539733887, "step": 77 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1534.0, "completions/max_terminated_length": 1534.0, "completions/mean_length": 556.380859375, "completions/mean_terminated_length": 556.380859375, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.058734939759036146, "frac_reward_zero_std": 0.0, "grad_norm": 0.29406511783599854, "learning_rate": 1e-06, "loss": 0.0525, "num_tokens": 24778050.0, "reward": 20.131879806518555, "reward_std": 3.6868181228637695, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.131879806518555, "rewards/skywork_reward/std": 6.814955711364746, "step": 78 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.013671875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1524.0, "completions/mean_length": 654.173828125, "completions/mean_terminated_length": 641.9505004882812, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.05948795180722891, "frac_reward_zero_std": 0.0, "grad_norm": 4.713456153869629, "learning_rate": 1e-06, "loss": 0.0458, "num_tokens": 25163019.0, "reward": 19.8343563079834, "reward_std": 3.734424114227295, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.8343563079834, "rewards/skywork_reward/std": 7.607394695281982, "step": 79 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1445.0, "completions/mean_length": 608.091796875, "completions/mean_terminated_length": 606.2759399414062, "completions/min_length": 258.0, "completions/min_terminated_length": 258.0, "epoch": 0.060240963855421686, "frac_reward_zero_std": 0.0, "grad_norm": 0.2803974449634552, "learning_rate": 1e-06, "loss": 0.0548, "num_tokens": 25519930.0, "reward": 20.908302307128906, "reward_std": 4.699563980102539, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.908302307128906, "rewards/skywork_reward/std": 8.098281860351562, "step": 80 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1437.0, "completions/mean_length": 515.775390625, "completions/mean_terminated_length": 513.7788696289062, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.06099397590361446, "frac_reward_zero_std": 0.0, "grad_norm": 0.3042682707309723, "learning_rate": 1e-06, "loss": 0.0298, "num_tokens": 25827463.0, "reward": 21.09960174560547, "reward_std": 4.133601188659668, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.09960174560547, "rewards/skywork_reward/std": 7.604403972625732, "step": 81 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1306.0, "completions/max_terminated_length": 1306.0, "completions/mean_length": 542.271484375, "completions/mean_terminated_length": 542.271484375, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.061746987951807226, "frac_reward_zero_std": 0.0, "grad_norm": 0.292880117893219, "learning_rate": 1e-06, "loss": 0.0317, "num_tokens": 26153682.0, "reward": 19.809471130371094, "reward_std": 3.9524645805358887, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.809471130371094, "rewards/skywork_reward/std": 5.885915279388428, "step": 82 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 550.228515625, "completions/mean_terminated_length": 544.41845703125, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.0625, "frac_reward_zero_std": 0.0, "grad_norm": 0.30422526597976685, "learning_rate": 1e-06, "loss": 0.031, "num_tokens": 26487431.0, "reward": 20.889541625976562, "reward_std": 3.701859474182129, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.889541625976562, "rewards/skywork_reward/std": 6.698598384857178, "step": 83 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1312.0, "completions/mean_length": 571.2109375, "completions/mean_terminated_length": 567.427490234375, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.06325301204819277, "frac_reward_zero_std": 0.0, "grad_norm": 0.304601788520813, "learning_rate": 1e-06, "loss": 0.0356, "num_tokens": 26825955.0, "reward": 19.412330627441406, "reward_std": 3.9354281425476074, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.412330627441406, "rewards/skywork_reward/std": 7.623223304748535, "step": 84 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1232.0, "completions/max_terminated_length": 1232.0, "completions/mean_length": 523.302734375, "completions/mean_terminated_length": 523.302734375, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.06400602409638555, "frac_reward_zero_std": 0.0, "grad_norm": 0.2882124185562134, "learning_rate": 1e-06, "loss": 0.029, "num_tokens": 27138190.0, "reward": 21.05793571472168, "reward_std": 3.831148862838745, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.05793571472168, "rewards/skywork_reward/std": 7.37190580368042, "step": 85 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1453.0, "completions/mean_length": 548.119140625, "completions/mean_terminated_length": 546.1859130859375, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.06475903614457831, "frac_reward_zero_std": 0.0, "grad_norm": 0.2833157181739807, "learning_rate": 1e-06, "loss": 0.0274, "num_tokens": 27464027.0, "reward": 21.267467498779297, "reward_std": 3.6207642555236816, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.267467498779297, "rewards/skywork_reward/std": 7.28432559967041, "step": 86 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.01171875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1466.0, "completions/mean_length": 555.1640625, "completions/mean_terminated_length": 543.5336303710938, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.06551204819277108, "frac_reward_zero_std": 0.0, "grad_norm": 0.2782001793384552, "learning_rate": 1e-06, "loss": 0.0658, "num_tokens": 27796975.0, "reward": 21.733179092407227, "reward_std": 3.8758647441864014, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.733179092407227, "rewards/skywork_reward/std": 7.234414577484131, "step": 87 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1435.0, "completions/max_terminated_length": 1435.0, "completions/mean_length": 540.154296875, "completions/mean_terminated_length": 540.154296875, "completions/min_length": 68.0, "completions/min_terminated_length": 68.0, "epoch": 0.06626506024096386, "frac_reward_zero_std": 0.0, "grad_norm": 0.3396439254283905, "learning_rate": 1e-06, "loss": 0.0201, "num_tokens": 28117070.0, "reward": 21.417911529541016, "reward_std": 3.201744556427002, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.417911529541016, "rewards/skywork_reward/std": 6.518486022949219, "step": 88 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0234375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1520.0, "completions/mean_length": 639.146484375, "completions/mean_terminated_length": 617.6220092773438, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.06701807228915663, "frac_reward_zero_std": 0.0, "grad_norm": 0.2612081468105316, "learning_rate": 1e-06, "loss": 0.0429, "num_tokens": 28487609.0, "reward": 20.419099807739258, "reward_std": 3.7477428913116455, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.419099807739258, "rewards/skywork_reward/std": 7.6364030838012695, "step": 89 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1367.0, "completions/mean_length": 546.509765625, "completions/mean_terminated_length": 544.5733642578125, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.0677710843373494, "frac_reward_zero_std": 0.0, "grad_norm": 0.31919634342193604, "learning_rate": 1e-06, "loss": 0.0386, "num_tokens": 28814782.0, "reward": 21.126543045043945, "reward_std": 3.912198066711426, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.126543045043945, "rewards/skywork_reward/std": 7.0028533935546875, "step": 90 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1289.0, "completions/mean_length": 560.15625, "completions/mean_terminated_length": 556.3294677734375, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.06852409638554217, "frac_reward_zero_std": 0.0, "grad_norm": 0.3072337508201599, "learning_rate": 1e-06, "loss": 0.0445, "num_tokens": 29144750.0, "reward": 19.54640007019043, "reward_std": 3.560420513153076, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.54640007019043, "rewards/skywork_reward/std": 7.461215019226074, "step": 91 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1483.0, "completions/mean_length": 601.64453125, "completions/mean_terminated_length": 599.8160400390625, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.06927710843373494, "frac_reward_zero_std": 0.0, "grad_norm": 0.2847882807254791, "learning_rate": 1e-06, "loss": 0.0469, "num_tokens": 29494344.0, "reward": 19.150531768798828, "reward_std": 3.979829788208008, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.150531768798828, "rewards/skywork_reward/std": 7.566142559051514, "step": 92 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 602.068359375, "completions/mean_terminated_length": 600.24072265625, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.07003012048192771, "frac_reward_zero_std": 0.0, "grad_norm": 0.26867735385894775, "learning_rate": 1e-06, "loss": 0.0211, "num_tokens": 29846395.0, "reward": 20.15349578857422, "reward_std": 3.5188255310058594, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.15349578857422, "rewards/skywork_reward/std": 6.31872034072876, "step": 93 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1406.0, "completions/mean_length": 609.859375, "completions/mean_terminated_length": 608.0469360351562, "completions/min_length": 287.0, "completions/min_terminated_length": 287.0, "epoch": 0.07078313253012049, "frac_reward_zero_std": 0.0, "grad_norm": 0.28086456656455994, "learning_rate": 1e-06, "loss": 0.0549, "num_tokens": 30203763.0, "reward": 22.653549194335938, "reward_std": 3.634199857711792, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.653549194335938, "rewards/skywork_reward/std": 7.612020969390869, "step": 94 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1342.0, "completions/max_terminated_length": 1342.0, "completions/mean_length": 564.158203125, "completions/mean_terminated_length": 564.158203125, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.07153614457831325, "frac_reward_zero_std": 0.0, "grad_norm": 0.3087657690048218, "learning_rate": 1e-06, "loss": 0.0215, "num_tokens": 30535428.0, "reward": 20.03140640258789, "reward_std": 3.1922054290771484, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.03140640258789, "rewards/skywork_reward/std": 6.323237419128418, "step": 95 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.021484375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1477.0, "completions/mean_length": 591.076171875, "completions/mean_terminated_length": 570.329345703125, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.07228915662650602, "frac_reward_zero_std": 0.0, "grad_norm": 0.2983706295490265, "learning_rate": 1e-06, "loss": 0.0296, "num_tokens": 30883611.0, "reward": 17.560230255126953, "reward_std": 3.3740811347961426, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 17.560230255126953, "rewards/skywork_reward/std": 7.882108211517334, "step": 96 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.013671875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1092.0, "completions/mean_length": 549.53515625, "completions/mean_terminated_length": 535.8613891601562, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.0730421686746988, "frac_reward_zero_std": 0.0, "grad_norm": 0.3238271474838257, "learning_rate": 1e-06, "loss": 0.0384, "num_tokens": 31210173.0, "reward": 20.257064819335938, "reward_std": 3.4314146041870117, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.257064819335938, "rewards/skywork_reward/std": 7.943423271179199, "step": 97 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1285.0, "completions/max_terminated_length": 1285.0, "completions/mean_length": 505.283203125, "completions/mean_terminated_length": 505.283203125, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.07379518072289157, "frac_reward_zero_std": 0.0, "grad_norm": 0.33978375792503357, "learning_rate": 1e-06, "loss": 0.0432, "num_tokens": 31517790.0, "reward": 22.60009765625, "reward_std": 3.893589496612549, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.60009765625, "rewards/skywork_reward/std": 6.1815080642700195, "step": 98 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1375.0, "completions/max_terminated_length": 1375.0, "completions/mean_length": 537.619140625, "completions/mean_terminated_length": 537.619140625, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.07454819277108433, "frac_reward_zero_std": 0.0, "grad_norm": 0.3200412094593048, "learning_rate": 1e-06, "loss": 0.0187, "num_tokens": 31837723.0, "reward": 19.1778564453125, "reward_std": 3.4576809406280518, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.1778564453125, "rewards/skywork_reward/std": 8.491567611694336, "step": 99 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1507.0, "completions/mean_length": 560.501953125, "completions/mean_terminated_length": 552.8208618164062, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.07530120481927711, "frac_reward_zero_std": 0.0, "grad_norm": 0.30854177474975586, "learning_rate": 1e-06, "loss": 0.0291, "num_tokens": 32171164.0, "reward": 20.515422821044922, "reward_std": 3.6772897243499756, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.515422821044922, "rewards/skywork_reward/std": 6.499829292297363, "step": 100 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1443.0, "completions/mean_length": 553.505859375, "completions/mean_terminated_length": 551.5831909179688, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.07605421686746988, "frac_reward_zero_std": 0.0, "grad_norm": 0.3389620780944824, "learning_rate": 1e-06, "loss": 0.0467, "num_tokens": 32497007.0, "reward": 19.85546875, "reward_std": 4.108412265777588, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.85546875, "rewards/skywork_reward/std": 6.893716812133789, "step": 101 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.01171875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1468.0, "completions/mean_length": 544.18359375, "completions/mean_terminated_length": 532.4229736328125, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.07680722891566265, "frac_reward_zero_std": 0.0, "grad_norm": 0.31205376982688904, "learning_rate": 1e-06, "loss": 0.0342, "num_tokens": 32825373.0, "reward": 21.427410125732422, "reward_std": 4.496371269226074, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.427410125732422, "rewards/skywork_reward/std": 8.456439018249512, "step": 102 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1345.0, "completions/max_terminated_length": 1345.0, "completions/mean_length": 530.46875, "completions/mean_terminated_length": 530.46875, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.07756024096385543, "frac_reward_zero_std": 0.0, "grad_norm": 0.3013162314891815, "learning_rate": 1e-06, "loss": 0.0147, "num_tokens": 33142141.0, "reward": 20.615509033203125, "reward_std": 2.9499316215515137, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.615509033203125, "rewards/skywork_reward/std": 5.820370674133301, "step": 103 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1393.0, "completions/mean_length": 543.732421875, "completions/mean_terminated_length": 541.7905883789062, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.0783132530120482, "frac_reward_zero_std": 0.0, "grad_norm": 0.31271547079086304, "learning_rate": 1e-06, "loss": 0.0393, "num_tokens": 33467588.0, "reward": 20.930374145507812, "reward_std": 3.4611310958862305, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.930374145507812, "rewards/skywork_reward/std": 6.854768753051758, "step": 104 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1375.0, "completions/max_terminated_length": 1375.0, "completions/mean_length": 561.560546875, "completions/mean_terminated_length": 561.560546875, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.07906626506024096, "frac_reward_zero_std": 0.0, "grad_norm": 0.30660298466682434, "learning_rate": 1e-06, "loss": 0.0255, "num_tokens": 33798691.0, "reward": 17.608062744140625, "reward_std": 3.2770185470581055, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 17.608062744140625, "rewards/skywork_reward/std": 7.616006851196289, "step": 105 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.017578125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1508.0, "completions/mean_length": 541.017578125, "completions/mean_terminated_length": 523.2146606445312, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.07981927710843373, "frac_reward_zero_std": 0.0, "grad_norm": 0.32308229804039, "learning_rate": 1e-06, "loss": 0.0226, "num_tokens": 34123004.0, "reward": 20.36962890625, "reward_std": 3.120267152786255, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.36962890625, "rewards/skywork_reward/std": 7.793717861175537, "step": 106 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1401.0, "completions/mean_length": 521.126953125, "completions/mean_terminated_length": 519.140869140625, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.08057228915662651, "frac_reward_zero_std": 0.0, "grad_norm": 0.31717127561569214, "learning_rate": 1e-06, "loss": 0.0358, "num_tokens": 34435741.0, "reward": 21.01158905029297, "reward_std": 3.5439646244049072, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.01158905029297, "rewards/skywork_reward/std": 8.22716999053955, "step": 107 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0234375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1532.0, "completions/mean_length": 624.4296875, "completions/mean_terminated_length": 602.552001953125, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.08132530120481928, "frac_reward_zero_std": 0.0, "grad_norm": 0.2963525652885437, "learning_rate": 1e-06, "loss": 0.0347, "num_tokens": 34802345.0, "reward": 20.850997924804688, "reward_std": 3.9396917819976807, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.850997924804688, "rewards/skywork_reward/std": 7.685678958892822, "step": 108 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1478.0, "completions/max_terminated_length": 1478.0, "completions/mean_length": 587.759765625, "completions/mean_terminated_length": 587.759765625, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.08207831325301204, "frac_reward_zero_std": 0.0, "grad_norm": 0.3052363991737366, "learning_rate": 1e-06, "loss": 0.0248, "num_tokens": 35147342.0, "reward": 19.248363494873047, "reward_std": 3.9026362895965576, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.248363494873047, "rewards/skywork_reward/std": 7.5164618492126465, "step": 109 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 592.96875, "completions/mean_terminated_length": 592.96875, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.08283132530120482, "frac_reward_zero_std": 0.0, "grad_norm": 0.2893350124359131, "learning_rate": 1e-06, "loss": 0.0428, "num_tokens": 35498750.0, "reward": 20.73992919921875, "reward_std": 4.127569675445557, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.73992919921875, "rewards/skywork_reward/std": 7.064544677734375, "step": 110 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1275.0, "completions/max_terminated_length": 1275.0, "completions/mean_length": 580.20703125, "completions/mean_terminated_length": 580.20703125, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "epoch": 0.08358433734939759, "frac_reward_zero_std": 0.0, "grad_norm": 0.2851979732513428, "learning_rate": 1e-06, "loss": 0.0143, "num_tokens": 35838488.0, "reward": 22.53789520263672, "reward_std": 3.5906736850738525, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.53789520263672, "rewards/skywork_reward/std": 7.760654449462891, "step": 111 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1381.0, "completions/mean_length": 562.802734375, "completions/mean_terminated_length": 560.8982543945312, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.08433734939759036, "frac_reward_zero_std": 0.0, "grad_norm": 0.3198689818382263, "learning_rate": 1e-06, "loss": 0.0329, "num_tokens": 36174899.0, "reward": 18.075515747070312, "reward_std": 3.7063498497009277, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.075515747070312, "rewards/skywork_reward/std": 8.073354721069336, "step": 112 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1242.0, "completions/mean_length": 566.4921875, "completions/mean_terminated_length": 564.5949096679688, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.08509036144578314, "frac_reward_zero_std": 0.0, "grad_norm": 0.2963120639324188, "learning_rate": 1e-06, "loss": 0.0291, "num_tokens": 36510591.0, "reward": 19.600627899169922, "reward_std": 3.769819736480713, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.600627899169922, "rewards/skywork_reward/std": 8.610282897949219, "step": 113 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1418.0, "completions/mean_length": 620.0, "completions/mean_terminated_length": 618.2074584960938, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.0858433734939759, "frac_reward_zero_std": 0.0, "grad_norm": 0.28681400418281555, "learning_rate": 1e-06, "loss": 0.0203, "num_tokens": 36873903.0, "reward": 19.95730209350586, "reward_std": 3.9464974403381348, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.95730209350586, "rewards/skywork_reward/std": 7.085601806640625, "step": 114 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1423.0, "completions/mean_length": 568.66796875, "completions/mean_terminated_length": 566.7749633789062, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.08659638554216867, "frac_reward_zero_std": 0.0, "grad_norm": 0.3024079501628876, "learning_rate": 1e-06, "loss": 0.0374, "num_tokens": 37214469.0, "reward": 20.049148559570312, "reward_std": 3.8342490196228027, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.049148559570312, "rewards/skywork_reward/std": 7.704470634460449, "step": 115 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1490.0, "completions/mean_length": 600.234375, "completions/mean_terminated_length": 596.5647583007812, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.08734939759036145, "frac_reward_zero_std": 0.0, "grad_norm": 0.3003012239933014, "learning_rate": 1e-06, "loss": 0.0454, "num_tokens": 37569709.0, "reward": 19.68072509765625, "reward_std": 4.329319953918457, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.68072509765625, "rewards/skywork_reward/std": 7.497686386108398, "step": 116 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1498.0, "completions/mean_length": 607.294921875, "completions/mean_terminated_length": 605.4774780273438, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.08810240963855422, "frac_reward_zero_std": 0.0, "grad_norm": 0.27138060331344604, "learning_rate": 1e-06, "loss": 0.0173, "num_tokens": 37926180.0, "reward": 22.437545776367188, "reward_std": 3.5447933673858643, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.437545776367188, "rewards/skywork_reward/std": 6.564459323883057, "step": 117 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1526.0, "completions/max_terminated_length": 1526.0, "completions/mean_length": 588.505859375, "completions/mean_terminated_length": 588.505859375, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.08885542168674698, "frac_reward_zero_std": 0.0, "grad_norm": 0.27087512612342834, "learning_rate": 1e-06, "loss": 0.031, "num_tokens": 38270599.0, "reward": 21.066497802734375, "reward_std": 3.951112985610962, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.066497802734375, "rewards/skywork_reward/std": 7.548466205596924, "step": 118 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1512.0, "completions/max_terminated_length": 1512.0, "completions/mean_length": 568.57421875, "completions/mean_terminated_length": 568.57421875, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.08960843373493976, "frac_reward_zero_std": 0.0, "grad_norm": 0.2935068607330322, "learning_rate": 1e-06, "loss": 0.0364, "num_tokens": 38607181.0, "reward": 18.969436645507812, "reward_std": 3.821528434753418, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 18.969436645507812, "rewards/skywork_reward/std": 8.197724342346191, "step": 119 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1475.0, "completions/mean_length": 571.787109375, "completions/mean_terminated_length": 566.1041259765625, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.09036144578313253, "frac_reward_zero_std": 0.0, "grad_norm": 0.32413604855537415, "learning_rate": 1e-06, "loss": 0.0357, "num_tokens": 38945056.0, "reward": 19.515613555908203, "reward_std": 3.548654556274414, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.515613555908203, "rewards/skywork_reward/std": 7.051810264587402, "step": 120 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1199.0, "completions/max_terminated_length": 1199.0, "completions/mean_length": 541.04296875, "completions/mean_terminated_length": 541.04296875, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.0911144578313253, "frac_reward_zero_std": 0.0, "grad_norm": 0.33912891149520874, "learning_rate": 1e-06, "loss": 0.0302, "num_tokens": 39266406.0, "reward": 21.350250244140625, "reward_std": 3.481703758239746, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.350250244140625, "rewards/skywork_reward/std": 7.860495567321777, "step": 121 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.013671875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1444.0, "completions/mean_length": 608.25390625, "completions/mean_terminated_length": 595.3941040039062, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.09186746987951808, "frac_reward_zero_std": 0.0, "grad_norm": 0.27568143606185913, "learning_rate": 1e-06, "loss": 0.0365, "num_tokens": 39626184.0, "reward": 20.61075210571289, "reward_std": 3.709376335144043, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.61075210571289, "rewards/skywork_reward/std": 7.836774826049805, "step": 122 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1520.0, "completions/mean_length": 671.375, "completions/mean_terminated_length": 664.5669555664062, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.09262048192771084, "frac_reward_zero_std": 0.0, "grad_norm": 0.2713894844055176, "learning_rate": 1e-06, "loss": 0.0227, "num_tokens": 40013720.0, "reward": 19.10358428955078, "reward_std": 3.289210081100464, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.10358428955078, "rewards/skywork_reward/std": 8.462961196899414, "step": 123 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1436.0, "completions/mean_length": 618.333984375, "completions/mean_terminated_length": 614.7353515625, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.09337349397590361, "frac_reward_zero_std": 0.0, "grad_norm": 0.28417181968688965, "learning_rate": 1e-06, "loss": 0.045, "num_tokens": 40377171.0, "reward": 19.55655288696289, "reward_std": 4.263092994689941, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.55655288696289, "rewards/skywork_reward/std": 7.342254161834717, "step": 124 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1383.0, "completions/max_terminated_length": 1383.0, "completions/mean_length": 596.92578125, "completions/mean_terminated_length": 596.92578125, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.09412650602409639, "frac_reward_zero_std": 0.0, "grad_norm": 0.28085947036743164, "learning_rate": 1e-06, "loss": 0.0195, "num_tokens": 40728669.0, "reward": 21.083993911743164, "reward_std": 3.448134422302246, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.083993911743164, "rewards/skywork_reward/std": 6.892836570739746, "step": 125 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1493.0, "completions/mean_length": 575.734375, "completions/mean_terminated_length": 573.8551635742188, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.09487951807228916, "frac_reward_zero_std": 0.0, "grad_norm": 0.30826082825660706, "learning_rate": 1e-06, "loss": 0.0245, "num_tokens": 41069797.0, "reward": 21.492095947265625, "reward_std": 3.321489095687866, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.492095947265625, "rewards/skywork_reward/std": 7.797940254211426, "step": 126 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1433.0, "completions/mean_length": 586.591796875, "completions/mean_terminated_length": 580.99609375, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.09563253012048192, "frac_reward_zero_std": 0.0, "grad_norm": 0.31485578417778015, "learning_rate": 1e-06, "loss": 0.0292, "num_tokens": 41417188.0, "reward": 19.127901077270508, "reward_std": 3.9224164485931396, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.127901077270508, "rewards/skywork_reward/std": 7.79984712600708, "step": 127 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1366.0, "completions/max_terminated_length": 1366.0, "completions/mean_length": 515.953125, "completions/mean_terminated_length": 515.953125, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.0963855421686747, "frac_reward_zero_std": 0.0, "grad_norm": 0.3710417151451111, "learning_rate": 1e-06, "loss": 0.0218, "num_tokens": 41725820.0, "reward": 21.463088989257812, "reward_std": 3.2953317165374756, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.463088989257812, "rewards/skywork_reward/std": 7.579995155334473, "step": 128 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1239.0, "completions/mean_length": 565.0234375, "completions/mean_terminated_length": 563.123291015625, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.09713855421686747, "frac_reward_zero_std": 0.0, "grad_norm": 0.3211367428302765, "learning_rate": 1e-06, "loss": 0.0347, "num_tokens": 42057704.0, "reward": 21.086503982543945, "reward_std": 3.46575927734375, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.086503982543945, "rewards/skywork_reward/std": 7.8865861892700195, "step": 129 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1267.0, "completions/mean_length": 560.736328125, "completions/mean_terminated_length": 553.0570678710938, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.09789156626506024, "frac_reward_zero_std": 0.0, "grad_norm": 0.4249287545681, "learning_rate": 1e-06, "loss": 0.0553, "num_tokens": 42388433.0, "reward": 19.455886840820312, "reward_std": 3.8543548583984375, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.455886840820312, "rewards/skywork_reward/std": 7.918321132659912, "step": 130 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1415.0, "completions/max_terminated_length": 1415.0, "completions/mean_length": 542.6484375, "completions/mean_terminated_length": 542.6484375, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.09864457831325302, "frac_reward_zero_std": 0.0, "grad_norm": 0.3672284483909607, "learning_rate": 1e-06, "loss": 0.0236, "num_tokens": 42711533.0, "reward": 21.042701721191406, "reward_std": 4.084582805633545, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.042701721191406, "rewards/skywork_reward/std": 8.580524444580078, "step": 131 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1478.0, "completions/mean_length": 606.06640625, "completions/mean_terminated_length": 600.5855102539062, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.09939759036144578, "frac_reward_zero_std": 0.0, "grad_norm": 0.3537883758544922, "learning_rate": 1e-06, "loss": 0.0434, "num_tokens": 43065375.0, "reward": 22.385543823242188, "reward_std": 4.083456516265869, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.385543823242188, "rewards/skywork_reward/std": 7.7553229331970215, "step": 132 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1357.0, "completions/max_terminated_length": 1357.0, "completions/mean_length": 638.109375, "completions/mean_terminated_length": 638.109375, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.10015060240963855, "frac_reward_zero_std": 0.0, "grad_norm": 0.31206345558166504, "learning_rate": 1e-06, "loss": 0.0194, "num_tokens": 43435015.0, "reward": 23.201324462890625, "reward_std": 3.3933279514312744, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.201324462890625, "rewards/skywork_reward/std": 8.763078689575195, "step": 133 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1519.0, "completions/mean_length": 683.912109375, "completions/mean_terminated_length": 678.8900146484375, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.10090361445783133, "frac_reward_zero_std": 0.0, "grad_norm": 0.3065231144428253, "learning_rate": 1e-06, "loss": 0.0145, "num_tokens": 43832490.0, "reward": 21.31167221069336, "reward_std": 3.8207640647888184, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.31167221069336, "rewards/skywork_reward/std": 6.737093448638916, "step": 134 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1257.0, "completions/max_terminated_length": 1257.0, "completions/mean_length": 599.5703125, "completions/mean_terminated_length": 599.5703125, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.1016566265060241, "frac_reward_zero_std": 0.0, "grad_norm": 0.3484726548194885, "learning_rate": 1e-06, "loss": 0.0294, "num_tokens": 44187758.0, "reward": 22.86431884765625, "reward_std": 3.3447577953338623, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.86431884765625, "rewards/skywork_reward/std": 7.047182083129883, "step": 135 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1471.0, "completions/max_terminated_length": 1471.0, "completions/mean_length": 613.40625, "completions/mean_terminated_length": 613.40625, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.10240963855421686, "frac_reward_zero_std": 0.0, "grad_norm": 0.3716602623462677, "learning_rate": 1e-06, "loss": 0.0281, "num_tokens": 44547310.0, "reward": 22.30841064453125, "reward_std": 3.4224910736083984, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.30841064453125, "rewards/skywork_reward/std": 7.069167137145996, "step": 136 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1517.0, "completions/mean_length": 641.091796875, "completions/mean_terminated_length": 635.8173217773438, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.10316265060240964, "frac_reward_zero_std": 0.0, "grad_norm": 0.34256511926651, "learning_rate": 1e-06, "loss": 0.0295, "num_tokens": 44925341.0, "reward": 19.721071243286133, "reward_std": 3.7587203979492188, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.721071243286133, "rewards/skywork_reward/std": 7.847194194793701, "step": 137 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1499.0, "completions/max_terminated_length": 1499.0, "completions/mean_length": 639.66796875, "completions/mean_terminated_length": 639.66796875, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.10391566265060241, "frac_reward_zero_std": 0.0, "grad_norm": 0.33065295219421387, "learning_rate": 1e-06, "loss": 0.0388, "num_tokens": 45295539.0, "reward": 21.96673583984375, "reward_std": 3.900968074798584, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.96673583984375, "rewards/skywork_reward/std": 7.551266670227051, "step": 138 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1523.0, "completions/max_terminated_length": 1523.0, "completions/mean_length": 610.98828125, "completions/mean_terminated_length": 610.98828125, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.10466867469879518, "frac_reward_zero_std": 0.0, "grad_norm": 0.3167376220226288, "learning_rate": 1e-06, "loss": 0.0114, "num_tokens": 45658877.0, "reward": 19.854530334472656, "reward_std": 4.059847354888916, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.854530334472656, "rewards/skywork_reward/std": 6.8408002853393555, "step": 139 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1457.0, "completions/mean_length": 653.609375, "completions/mean_terminated_length": 651.882568359375, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.10542168674698796, "frac_reward_zero_std": 0.0, "grad_norm": 0.29037731885910034, "learning_rate": 1e-06, "loss": 0.0306, "num_tokens": 46037205.0, "reward": 20.99395751953125, "reward_std": 3.6438076496124268, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.99395751953125, "rewards/skywork_reward/std": 6.79690408706665, "step": 140 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1523.0, "completions/mean_length": 633.763671875, "completions/mean_terminated_length": 631.998046875, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.10617469879518072, "frac_reward_zero_std": 0.0, "grad_norm": 0.2938140332698822, "learning_rate": 1e-06, "loss": 0.0272, "num_tokens": 46405644.0, "reward": 21.723541259765625, "reward_std": 4.110295295715332, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.723541259765625, "rewards/skywork_reward/std": 6.7619757652282715, "step": 141 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.009765625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1354.0, "completions/mean_length": 589.501953125, "completions/mean_terminated_length": 580.1676635742188, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.10692771084337349, "frac_reward_zero_std": 0.0, "grad_norm": 0.3247605264186859, "learning_rate": 1e-06, "loss": 0.0524, "num_tokens": 46749773.0, "reward": 19.817337036132812, "reward_std": 3.662755012512207, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.817337036132812, "rewards/skywork_reward/std": 7.711267948150635, "step": 142 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1359.0, "completions/max_terminated_length": 1359.0, "completions/mean_length": 625.251953125, "completions/mean_terminated_length": 625.251953125, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.10768072289156627, "frac_reward_zero_std": 0.0, "grad_norm": 0.3041718900203705, "learning_rate": 1e-06, "loss": 0.0231, "num_tokens": 47112942.0, "reward": 21.378835678100586, "reward_std": 3.913424491882324, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.378835678100586, "rewards/skywork_reward/std": 7.104243755340576, "step": 143 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.017578125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1508.0, "completions/mean_length": 646.98828125, "completions/mean_terminated_length": 631.0814819335938, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.10843373493975904, "frac_reward_zero_std": 0.0, "grad_norm": 0.2911466062068939, "learning_rate": 1e-06, "loss": 0.0378, "num_tokens": 47491000.0, "reward": 21.82143211364746, "reward_std": 3.859783887863159, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.82143211364746, "rewards/skywork_reward/std": 9.108434677124023, "step": 144 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1434.0, "completions/mean_length": 588.384765625, "completions/mean_terminated_length": 586.5303344726562, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.1091867469879518, "frac_reward_zero_std": 0.0, "grad_norm": 0.30668869614601135, "learning_rate": 1e-06, "loss": 0.0326, "num_tokens": 47839837.0, "reward": 21.99334716796875, "reward_std": 3.588073968887329, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.99334716796875, "rewards/skywork_reward/std": 8.704425811767578, "step": 145 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1261.0, "completions/mean_length": 647.630859375, "completions/mean_terminated_length": 645.892333984375, "completions/min_length": 271.0, "completions/min_terminated_length": 271.0, "epoch": 0.10993975903614457, "frac_reward_zero_std": 0.0, "grad_norm": 0.29260244965553284, "learning_rate": 1e-06, "loss": 0.0329, "num_tokens": 48214544.0, "reward": 21.543075561523438, "reward_std": 3.8497776985168457, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.543075561523438, "rewards/skywork_reward/std": 8.165260314941406, "step": 146 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1532.0, "completions/mean_length": 726.556640625, "completions/mean_terminated_length": 720.1830444335938, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.11069277108433735, "frac_reward_zero_std": 0.0, "grad_norm": 0.27006274461746216, "learning_rate": 1e-06, "loss": 0.0695, "num_tokens": 48635277.0, "reward": 19.606515884399414, "reward_std": 4.228018760681152, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.606515884399414, "rewards/skywork_reward/std": 6.9952826499938965, "step": 147 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 619.673828125, "completions/mean_terminated_length": 614.2731323242188, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.11144578313253012, "frac_reward_zero_std": 0.0, "grad_norm": 0.2880130410194397, "learning_rate": 1e-06, "loss": 0.0294, "num_tokens": 48994470.0, "reward": 20.1602783203125, "reward_std": 3.2607784271240234, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.1602783203125, "rewards/skywork_reward/std": 6.346843719482422, "step": 148 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1392.0, "completions/mean_length": 629.67578125, "completions/mean_terminated_length": 627.9021606445312, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.11219879518072289, "frac_reward_zero_std": 0.0, "grad_norm": 0.2898614704608917, "learning_rate": 1e-06, "loss": 0.0405, "num_tokens": 49363600.0, "reward": 22.240509033203125, "reward_std": 3.9023704528808594, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.240509033203125, "rewards/skywork_reward/std": 7.541024208068848, "step": 149 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1434.0, "completions/max_terminated_length": 1434.0, "completions/mean_length": 629.8828125, "completions/mean_terminated_length": 629.8828125, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.11295180722891567, "frac_reward_zero_std": 0.0, "grad_norm": 0.2920468747615814, "learning_rate": 1e-06, "loss": 0.0138, "num_tokens": 49729284.0, "reward": 21.534242630004883, "reward_std": 3.1994595527648926, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.534242630004883, "rewards/skywork_reward/std": 7.222407341003418, "step": 150 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.025390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1490.0, "completions/mean_length": 638.59765625, "completions/mean_terminated_length": 615.2184448242188, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.11370481927710843, "frac_reward_zero_std": 0.0, "grad_norm": 0.2872624397277832, "learning_rate": 1e-06, "loss": 0.0157, "num_tokens": 50105238.0, "reward": 19.462825775146484, "reward_std": 3.6386303901672363, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.462825775146484, "rewards/skywork_reward/std": 7.454095363616943, "step": 151 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1464.0, "completions/mean_length": 575.103515625, "completions/mean_terminated_length": 567.5374145507812, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.1144578313253012, "frac_reward_zero_std": 0.0, "grad_norm": 0.3207628130912781, "learning_rate": 1e-06, "loss": 0.0549, "num_tokens": 50444907.0, "reward": 20.47686767578125, "reward_std": 3.3109498023986816, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.47686767578125, "rewards/skywork_reward/std": 7.746679306030273, "step": 152 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1445.0, "completions/mean_length": 629.671875, "completions/mean_terminated_length": 626.11767578125, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.11521084337349398, "frac_reward_zero_std": 0.0, "grad_norm": 0.2737574875354767, "learning_rate": 1e-06, "loss": 0.0135, "num_tokens": 50814211.0, "reward": 20.94318199157715, "reward_std": 4.387291431427002, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.94318199157715, "rewards/skywork_reward/std": 8.372903823852539, "step": 153 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.01171875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1522.0, "completions/mean_length": 580.333984375, "completions/mean_terminated_length": 569.0020141601562, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.11596385542168675, "frac_reward_zero_std": 0.0, "grad_norm": 0.32104018330574036, "learning_rate": 1e-06, "loss": 0.0363, "num_tokens": 51156158.0, "reward": 21.5543212890625, "reward_std": 3.9272093772888184, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.5543212890625, "rewards/skywork_reward/std": 6.690492153167725, "step": 154 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.009765625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1456.0, "completions/mean_length": 579.384765625, "completions/mean_terminated_length": 569.95068359375, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.11671686746987951, "frac_reward_zero_std": 0.0, "grad_norm": 0.30905836820602417, "learning_rate": 1e-06, "loss": 0.0484, "num_tokens": 51498627.0, "reward": 21.084388732910156, "reward_std": 3.953840494155884, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.084388732910156, "rewards/skywork_reward/std": 8.337218284606934, "step": 155 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1343.0, "completions/max_terminated_length": 1343.0, "completions/mean_length": 585.482421875, "completions/mean_terminated_length": 585.482421875, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.11746987951807229, "frac_reward_zero_std": 0.0, "grad_norm": 0.2910420000553131, "learning_rate": 1e-06, "loss": 0.0174, "num_tokens": 51845034.0, "reward": 24.57891845703125, "reward_std": 3.3543262481689453, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 24.57891845703125, "rewards/skywork_reward/std": 6.263175010681152, "step": 156 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.025390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1421.0, "completions/mean_length": 614.875, "completions/mean_terminated_length": 590.8777465820312, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.11822289156626506, "frac_reward_zero_std": 0.0, "grad_norm": 0.30173835158348083, "learning_rate": 1e-06, "loss": 0.0252, "num_tokens": 52204538.0, "reward": 21.792831420898438, "reward_std": 2.7370691299438477, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.792831420898438, "rewards/skywork_reward/std": 7.606500148773193, "step": 157 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1515.0, "completions/mean_length": 575.78125, "completions/mean_terminated_length": 572.0157470703125, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.11897590361445783, "frac_reward_zero_std": 0.0, "grad_norm": 0.3065696656703949, "learning_rate": 1e-06, "loss": 0.0432, "num_tokens": 52541514.0, "reward": 21.47296142578125, "reward_std": 3.3334720134735107, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.47296142578125, "rewards/skywork_reward/std": 8.22110652923584, "step": 158 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1460.0, "completions/max_terminated_length": 1460.0, "completions/mean_length": 588.296875, "completions/mean_terminated_length": 588.296875, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.1197289156626506, "frac_reward_zero_std": 0.0, "grad_norm": 0.30373716354370117, "learning_rate": 1e-06, "loss": 0.023, "num_tokens": 52887346.0, "reward": 21.733245849609375, "reward_std": 3.1107327938079834, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.733245849609375, "rewards/skywork_reward/std": 7.991820335388184, "step": 159 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1281.0, "completions/mean_length": 575.61328125, "completions/mean_terminated_length": 573.7338256835938, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.12048192771084337, "frac_reward_zero_std": 0.0, "grad_norm": 0.3277619183063507, "learning_rate": 1e-06, "loss": 0.027, "num_tokens": 53228412.0, "reward": 19.673015594482422, "reward_std": 3.533682107925415, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 19.673015594482422, "rewards/skywork_reward/std": 7.22337532043457, "step": 160 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1520.0, "completions/mean_length": 553.216796875, "completions/mean_terminated_length": 551.2935180664062, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.12123493975903614, "frac_reward_zero_std": 0.0, "grad_norm": 0.33466610312461853, "learning_rate": 1e-06, "loss": 0.029, "num_tokens": 53559739.0, "reward": 20.984954833984375, "reward_std": 3.4584217071533203, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.984954833984375, "rewards/skywork_reward/std": 6.437849521636963, "step": 161 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1488.0, "completions/mean_length": 622.49609375, "completions/mean_terminated_length": 618.9137573242188, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.12198795180722892, "frac_reward_zero_std": 0.0, "grad_norm": 0.3004685342311859, "learning_rate": 1e-06, "loss": 0.017, "num_tokens": 53922953.0, "reward": 22.345840454101562, "reward_std": 3.4734017848968506, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.345840454101562, "rewards/skywork_reward/std": 7.882354736328125, "step": 162 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1463.0, "completions/mean_length": 556.958984375, "completions/mean_terminated_length": 553.11962890625, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.12274096385542169, "frac_reward_zero_std": 0.0, "grad_norm": 0.33252498507499695, "learning_rate": 1e-06, "loss": 0.0592, "num_tokens": 54249444.0, "reward": 20.580535888671875, "reward_std": 3.609018087387085, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.580535888671875, "rewards/skywork_reward/std": 7.147970199584961, "step": 163 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1486.0, "completions/mean_length": 598.171875, "completions/mean_terminated_length": 596.3366088867188, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.12349397590361445, "frac_reward_zero_std": 0.0, "grad_norm": 0.3217158913612366, "learning_rate": 1e-06, "loss": 0.0387, "num_tokens": 54602780.0, "reward": 20.154247283935547, "reward_std": 3.670168399810791, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.154247283935547, "rewards/skywork_reward/std": 6.9621381759643555, "step": 164 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1384.0, "completions/mean_length": 615.55078125, "completions/mean_terminated_length": 613.74951171875, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.12424698795180723, "frac_reward_zero_std": 0.0, "grad_norm": 0.324698805809021, "learning_rate": 1e-06, "loss": 0.062, "num_tokens": 54967254.0, "reward": 22.460899353027344, "reward_std": 4.602289199829102, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.460899353027344, "rewards/skywork_reward/std": 7.923734664916992, "step": 165 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1451.0, "completions/mean_length": 608.3203125, "completions/mean_terminated_length": 606.5048828125, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.125, "frac_reward_zero_std": 0.0, "grad_norm": 0.3245730400085449, "learning_rate": 1e-06, "loss": 0.0314, "num_tokens": 55327114.0, "reward": 20.21398162841797, "reward_std": 3.487682342529297, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.21398162841797, "rewards/skywork_reward/std": 7.384416103363037, "step": 166 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1434.0, "completions/mean_length": 548.36328125, "completions/mean_terminated_length": 546.4305419921875, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.12575301204819278, "frac_reward_zero_std": 0.0, "grad_norm": 0.35269981622695923, "learning_rate": 1e-06, "loss": 0.0122, "num_tokens": 55651652.0, "reward": 22.794578552246094, "reward_std": 3.213481903076172, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.794578552246094, "rewards/skywork_reward/std": 7.596462726593018, "step": 167 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1524.0, "completions/mean_length": 584.244140625, "completions/mean_terminated_length": 576.75, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.12650602409638553, "frac_reward_zero_std": 0.0, "grad_norm": 0.33272233605384827, "learning_rate": 1e-06, "loss": 0.0165, "num_tokens": 55995473.0, "reward": 21.558271408081055, "reward_std": 3.5698699951171875, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.558271408081055, "rewards/skywork_reward/std": 7.89260721206665, "step": 168 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1353.0, "completions/max_terminated_length": 1353.0, "completions/mean_length": 607.416015625, "completions/mean_terminated_length": 607.416015625, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.1272590361445783, "frac_reward_zero_std": 0.0, "grad_norm": 0.35381069779396057, "learning_rate": 1e-06, "loss": 0.0291, "num_tokens": 56355750.0, "reward": 21.860408782958984, "reward_std": 4.002436637878418, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.860408782958984, "rewards/skywork_reward/std": 6.837855339050293, "step": 169 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1370.0, "completions/mean_length": 570.771484375, "completions/mean_terminated_length": 568.882568359375, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.1280120481927711, "frac_reward_zero_std": 0.0, "grad_norm": 0.34828981757164, "learning_rate": 1e-06, "loss": 0.0188, "num_tokens": 56692929.0, "reward": 20.509864807128906, "reward_std": 3.7775073051452637, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.509864807128906, "rewards/skywork_reward/std": 7.960306167602539, "step": 170 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1454.0, "completions/mean_length": 624.5546875, "completions/mean_terminated_length": 622.7710571289062, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.12876506024096385, "frac_reward_zero_std": 0.0, "grad_norm": 0.3327150046825409, "learning_rate": 1e-06, "loss": 0.0299, "num_tokens": 57063869.0, "reward": 22.369293212890625, "reward_std": 3.6728134155273438, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.369293212890625, "rewards/skywork_reward/std": 5.960691928863525, "step": 171 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1187.0, "completions/max_terminated_length": 1187.0, "completions/mean_length": 547.8203125, "completions/mean_terminated_length": 547.8203125, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.12951807228915663, "frac_reward_zero_std": 0.0, "grad_norm": 0.38631579279899597, "learning_rate": 1e-06, "loss": 0.013, "num_tokens": 57390065.0, "reward": 21.2592716217041, "reward_std": 3.4279208183288574, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.2592716217041, "rewards/skywork_reward/std": 7.6550774574279785, "step": 172 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1250.0, "completions/max_terminated_length": 1250.0, "completions/mean_length": 591.0859375, "completions/mean_terminated_length": 591.0859375, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.1302710843373494, "frac_reward_zero_std": 0.0, "grad_norm": 0.3411237299442291, "learning_rate": 1e-06, "loss": 0.0128, "num_tokens": 57743421.0, "reward": 21.829246520996094, "reward_std": 3.742777109146118, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.829246520996094, "rewards/skywork_reward/std": 7.164978981018066, "step": 173 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 577.67578125, "completions/mean_terminated_length": 575.8004150390625, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.13102409638554216, "frac_reward_zero_std": 0.0, "grad_norm": 0.34287434816360474, "learning_rate": 1e-06, "loss": 0.0272, "num_tokens": 58083831.0, "reward": 22.2662353515625, "reward_std": 3.2713894844055176, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.2662353515625, "rewards/skywork_reward/std": 7.178630352020264, "step": 174 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1258.0, "completions/mean_length": 565.248046875, "completions/mean_terminated_length": 561.4412231445312, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "epoch": 0.13177710843373494, "frac_reward_zero_std": 0.0, "grad_norm": 0.35165417194366455, "learning_rate": 1e-06, "loss": 0.0374, "num_tokens": 58418918.0, "reward": 22.354202270507812, "reward_std": 4.054803848266602, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.354202270507812, "rewards/skywork_reward/std": 7.76880407333374, "step": 175 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1072.0, "completions/max_terminated_length": 1072.0, "completions/mean_length": 513.833984375, "completions/mean_terminated_length": 513.833984375, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.13253012048192772, "frac_reward_zero_std": 0.0, "grad_norm": 0.3769795298576355, "learning_rate": 1e-06, "loss": 0.0094, "num_tokens": 58727089.0, "reward": 24.008331298828125, "reward_std": 3.7061939239501953, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 24.008331298828125, "rewards/skywork_reward/std": 7.291410446166992, "step": 176 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1463.0, "completions/mean_length": 594.93359375, "completions/mean_terminated_length": 593.0919799804688, "completions/min_length": 238.0, "completions/min_terminated_length": 238.0, "epoch": 0.13328313253012047, "frac_reward_zero_std": 0.0, "grad_norm": 0.3345366418361664, "learning_rate": 1e-06, "loss": 0.0389, "num_tokens": 59079759.0, "reward": 22.873056411743164, "reward_std": 3.8607959747314453, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.873056411743164, "rewards/skywork_reward/std": 6.990558624267578, "step": 177 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1447.0, "completions/max_terminated_length": 1447.0, "completions/mean_length": 588.767578125, "completions/mean_terminated_length": 588.767578125, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.13403614457831325, "frac_reward_zero_std": 0.0, "grad_norm": 0.34772616624832153, "learning_rate": 1e-06, "loss": 0.0202, "num_tokens": 59424152.0, "reward": 20.871658325195312, "reward_std": 3.6873836517333984, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.871658325195312, "rewards/skywork_reward/std": 7.474151611328125, "step": 178 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1480.0, "completions/mean_length": 597.91015625, "completions/mean_terminated_length": 592.3811645507812, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.13478915662650603, "frac_reward_zero_std": 0.0, "grad_norm": 0.3342570662498474, "learning_rate": 1e-06, "loss": 0.0287, "num_tokens": 59775834.0, "reward": 22.359867095947266, "reward_std": 3.218701124191284, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.359867095947266, "rewards/skywork_reward/std": 8.73936939239502, "step": 179 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1486.0, "completions/max_terminated_length": 1486.0, "completions/mean_length": 583.130859375, "completions/mean_terminated_length": 583.130859375, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.1355421686746988, "frac_reward_zero_std": 0.0, "grad_norm": 0.32025235891342163, "learning_rate": 1e-06, "loss": 0.0153, "num_tokens": 60119293.0, "reward": 21.634063720703125, "reward_std": 3.2388057708740234, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.634063720703125, "rewards/skywork_reward/std": 6.3536248207092285, "step": 180 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1263.0, "completions/mean_length": 599.6484375, "completions/mean_terminated_length": 597.8160400390625, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.13629518072289157, "frac_reward_zero_std": 0.0, "grad_norm": 0.3474103808403015, "learning_rate": 1e-06, "loss": 0.0315, "num_tokens": 60467305.0, "reward": 22.63400650024414, "reward_std": 3.6428089141845703, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.63400650024414, "rewards/skywork_reward/std": 8.048898696899414, "step": 181 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1180.0, "completions/max_terminated_length": 1180.0, "completions/mean_length": 602.431640625, "completions/mean_terminated_length": 602.431640625, "completions/min_length": 287.0, "completions/min_terminated_length": 287.0, "epoch": 0.13704819277108435, "frac_reward_zero_std": 0.0, "grad_norm": 0.3456987142562866, "learning_rate": 1e-06, "loss": 0.0283, "num_tokens": 60822518.0, "reward": 22.980743408203125, "reward_std": 3.3244166374206543, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.980743408203125, "rewards/skywork_reward/std": 7.1537909507751465, "step": 182 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1241.0, "completions/max_terminated_length": 1241.0, "completions/mean_length": 617.828125, "completions/mean_terminated_length": 617.828125, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.1378012048192771, "frac_reward_zero_std": 0.0, "grad_norm": 0.32726573944091797, "learning_rate": 1e-06, "loss": 0.0093, "num_tokens": 61183902.0, "reward": 23.039154052734375, "reward_std": 2.817411422729492, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.039154052734375, "rewards/skywork_reward/std": 6.791592597961426, "step": 183 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.001953125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1528.0, "completions/mean_length": 601.18359375, "completions/mean_terminated_length": 599.3541870117188, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.13855421686746988, "frac_reward_zero_std": 0.0, "grad_norm": 0.3330095112323761, "learning_rate": 1e-06, "loss": 0.0182, "num_tokens": 61539884.0, "reward": 23.63543701171875, "reward_std": 3.1759846210479736, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.63543701171875, "rewards/skywork_reward/std": 7.07273006439209, "step": 184 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1424.0, "completions/max_terminated_length": 1424.0, "completions/mean_length": 651.955078125, "completions/mean_terminated_length": 651.955078125, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.13930722891566266, "frac_reward_zero_std": 0.0, "grad_norm": 0.39070987701416016, "learning_rate": 1e-06, "loss": 0.0259, "num_tokens": 61919077.0, "reward": 21.221710205078125, "reward_std": 3.5847105979919434, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.221710205078125, "rewards/skywork_reward/std": 7.764612197875977, "step": 185 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1329.0, "completions/max_terminated_length": 1329.0, "completions/mean_length": 621.73828125, "completions/mean_terminated_length": 621.73828125, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.14006024096385541, "frac_reward_zero_std": 0.0, "grad_norm": 0.36137887835502625, "learning_rate": 1e-06, "loss": 0.0158, "num_tokens": 62282095.0, "reward": 22.08428955078125, "reward_std": 3.342141628265381, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.08428955078125, "rewards/skywork_reward/std": 6.694947719573975, "step": 186 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1294.0, "completions/max_terminated_length": 1294.0, "completions/mean_length": 649.474609375, "completions/mean_terminated_length": 649.474609375, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "epoch": 0.1408132530120482, "frac_reward_zero_std": 0.0, "grad_norm": 0.3418244421482086, "learning_rate": 1e-06, "loss": 0.0059, "num_tokens": 62658690.0, "reward": 23.334083557128906, "reward_std": 3.1341238021850586, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.334083557128906, "rewards/skywork_reward/std": 6.964704513549805, "step": 187 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1408.0, "completions/max_terminated_length": 1408.0, "completions/mean_length": 609.8046875, "completions/mean_terminated_length": 609.8046875, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.14156626506024098, "frac_reward_zero_std": 0.0, "grad_norm": 0.3598421812057495, "learning_rate": 1e-06, "loss": 0.0167, "num_tokens": 63015342.0, "reward": 23.847320556640625, "reward_std": 3.479365825653076, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.847320556640625, "rewards/skywork_reward/std": 7.224637508392334, "step": 188 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.005859375, "completions/max_length": 1536.0, "completions/max_terminated_length": 1336.0, "completions/mean_length": 631.234375, "completions/mean_terminated_length": 625.9017944335938, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.14231927710843373, "frac_reward_zero_std": 0.0, "grad_norm": 0.3684713542461395, "learning_rate": 1e-06, "loss": 0.0391, "num_tokens": 63385654.0, "reward": 20.486244201660156, "reward_std": 3.928035020828247, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 20.486244201660156, "rewards/skywork_reward/std": 7.389331817626953, "step": 189 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1231.0, "completions/max_terminated_length": 1231.0, "completions/mean_length": 559.046875, "completions/mean_terminated_length": 559.046875, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.1430722891566265, "frac_reward_zero_std": 0.0, "grad_norm": 0.43111714720726013, "learning_rate": 1e-06, "loss": 0.0227, "num_tokens": 63715182.0, "reward": 24.990936279296875, "reward_std": 3.196828842163086, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 24.990936279296875, "rewards/skywork_reward/std": 6.8701629638671875, "step": 190 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1423.0, "completions/max_terminated_length": 1423.0, "completions/mean_length": 673.578125, "completions/mean_terminated_length": 673.578125, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.1438253012048193, "frac_reward_zero_std": 0.0, "grad_norm": 0.33778226375579834, "learning_rate": 1e-06, "loss": 0.0162, "num_tokens": 64108118.0, "reward": 22.113998413085938, "reward_std": 3.290834903717041, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.113998413085938, "rewards/skywork_reward/std": 7.376873970031738, "step": 191 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.013671875, "completions/max_length": 1536.0, "completions/max_terminated_length": 1530.0, "completions/mean_length": 710.322265625, "completions/mean_terminated_length": 698.8772583007812, "completions/min_length": 360.0, "completions/min_terminated_length": 360.0, "epoch": 0.14457831325301204, "frac_reward_zero_std": 0.0, "grad_norm": 0.33378493785858154, "learning_rate": 1e-06, "loss": 0.0461, "num_tokens": 64517355.0, "reward": 22.64098358154297, "reward_std": 3.9250552654266357, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.64098358154297, "rewards/skywork_reward/std": 7.483473777770996, "step": 192 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1432.0, "completions/max_terminated_length": 1432.0, "completions/mean_length": 609.9453125, "completions/mean_terminated_length": 609.9453125, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.14533132530120482, "frac_reward_zero_std": 0.0, "grad_norm": 0.3949362337589264, "learning_rate": 1e-06, "loss": 0.0283, "num_tokens": 64875503.0, "reward": 22.26373291015625, "reward_std": 3.6811184883117676, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.26373291015625, "rewards/skywork_reward/std": 7.222413539886475, "step": 193 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1306.0, "completions/max_terminated_length": 1306.0, "completions/mean_length": 611.69140625, "completions/mean_terminated_length": 611.69140625, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.1460843373493976, "frac_reward_zero_std": 0.0, "grad_norm": 0.37980496883392334, "learning_rate": 1e-06, "loss": 0.0178, "num_tokens": 65240289.0, "reward": 23.238971710205078, "reward_std": 3.542300224304199, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.238971710205078, "rewards/skywork_reward/std": 7.373708724975586, "step": 194 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.00390625, "completions/max_length": 1536.0, "completions/max_terminated_length": 1278.0, "completions/mean_length": 675.423828125, "completions/mean_terminated_length": 672.049072265625, "completions/min_length": 333.0, "completions/min_terminated_length": 333.0, "epoch": 0.14683734939759036, "frac_reward_zero_std": 0.0, "grad_norm": 10.594128608703613, "learning_rate": 1e-06, "loss": 0.0214, "num_tokens": 65631722.0, "reward": 22.663818359375, "reward_std": 3.3215575218200684, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 22.663818359375, "rewards/skywork_reward/std": 7.0012664794921875, "step": 195 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0078125, "completions/max_length": 1536.0, "completions/max_terminated_length": 1510.0, "completions/mean_length": 643.921875, "completions/mean_terminated_length": 636.8976440429688, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.14759036144578314, "frac_reward_zero_std": 0.0, "grad_norm": 0.38080358505249023, "learning_rate": 1e-06, "loss": 0.025, "num_tokens": 66005026.0, "reward": 23.194847106933594, "reward_std": 3.9582090377807617, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.194847106933594, "rewards/skywork_reward/std": 6.189321041107178, "step": 196 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1524.0, "completions/max_terminated_length": 1524.0, "completions/mean_length": 705.38671875, "completions/mean_terminated_length": 705.38671875, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "epoch": 0.14834337349397592, "frac_reward_zero_std": 0.0, "grad_norm": 0.3392493724822998, "learning_rate": 1e-06, "loss": 0.0361, "num_tokens": 66416936.0, "reward": 23.629547119140625, "reward_std": 3.766573905944824, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.629547119140625, "rewards/skywork_reward/std": 6.338648796081543, "step": 197 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1299.0, "completions/max_terminated_length": 1299.0, "completions/mean_length": 635.73828125, "completions/mean_terminated_length": 635.73828125, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.14909638554216867, "frac_reward_zero_std": 0.0, "grad_norm": 0.37394610047340393, "learning_rate": 1e-06, "loss": 0.0242, "num_tokens": 66789730.0, "reward": 23.806228637695312, "reward_std": 3.5995078086853027, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.806228637695312, "rewards/skywork_reward/std": 7.185088157653809, "step": 198 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1284.0, "completions/max_terminated_length": 1284.0, "completions/mean_length": 688.634765625, "completions/mean_terminated_length": 688.634765625, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "epoch": 0.14984939759036145, "frac_reward_zero_std": 0.0, "grad_norm": 0.31833726167678833, "learning_rate": 1e-06, "loss": 0.0197, "num_tokens": 67186647.0, "reward": 23.605316162109375, "reward_std": 3.3630738258361816, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 23.605316162109375, "rewards/skywork_reward/std": 8.333732604980469, "step": 199 }, { "clip_ratio": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1263.0, "completions/max_terminated_length": 1263.0, "completions/mean_length": 617.896484375, "completions/mean_terminated_length": 617.896484375, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "epoch": 0.15060240963855423, "frac_reward_zero_std": 0.0, "grad_norm": 0.3671130836009979, "learning_rate": 1e-06, "loss": 0.0301, "num_tokens": 67548866.0, "reward": 21.822662353515625, "reward_std": 3.521603584289551, "rewards/accuracy_reward/mean": 0.0, "rewards/accuracy_reward/std": 0.0, "rewards/skywork_reward/mean": 21.822662353515625, "rewards/skywork_reward/std": 6.942006587982178, "step": 200 } ], "logging_steps": 1, "max_steps": 1328, "num_input_tokens_seen": 67548866, "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": 64, "trial_name": null, "trial_params": null }