| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984, | |
| "eval_steps": 100, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 598.697940826416, | |
| "epoch": 0.0032, | |
| "grad_norm": 0.051513671875, | |
| "kl": 0.0, | |
| "learning_rate": 9.375e-08, | |
| "loss": 0.029, | |
| "reward": 0.8125000149011612, | |
| "reward_std": 0.17311252653598785, | |
| "rewards/accuracy_reward": 0.8125000149011612, | |
| "rewards/format_reward": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 577.0130367279053, | |
| "epoch": 0.016, | |
| "grad_norm": 0.054931640625, | |
| "kl": 8.213493288167228e-05, | |
| "learning_rate": 4.6875e-07, | |
| "loss": 0.033, | |
| "reward": 0.7552083497866988, | |
| "reward_std": 0.15550211956724524, | |
| "rewards/accuracy_reward": 0.7552083497866988, | |
| "rewards/format_reward": 0.0, | |
| "step": 5 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 538.3833518981934, | |
| "epoch": 0.032, | |
| "grad_norm": 0.0791015625, | |
| "kl": 0.00010595652129268274, | |
| "learning_rate": 9.375e-07, | |
| "loss": 0.0076, | |
| "reward": 0.7458333462476731, | |
| "reward_std": 0.11220084913074971, | |
| "rewards/accuracy_reward": 0.7458333462476731, | |
| "rewards/format_reward": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 572.8646041870118, | |
| "epoch": 0.048, | |
| "grad_norm": 0.07373046875, | |
| "kl": 0.00011188817652509896, | |
| "learning_rate": 1.40625e-06, | |
| "loss": 0.0166, | |
| "reward": 0.7125000156462192, | |
| "reward_std": 0.14682335481047631, | |
| "rewards/accuracy_reward": 0.7125000156462192, | |
| "rewards/format_reward": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 579.118766784668, | |
| "epoch": 0.064, | |
| "grad_norm": 0.0693359375, | |
| "kl": 0.00010264360735163791, | |
| "learning_rate": 1.875e-06, | |
| "loss": 0.0194, | |
| "reward": 0.7750000149011612, | |
| "reward_std": 0.12534543462097644, | |
| "rewards/accuracy_reward": 0.7750000149011612, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 535.8145973205567, | |
| "epoch": 0.08, | |
| "grad_norm": 0.06298828125, | |
| "kl": 0.00010584766669126111, | |
| "learning_rate": 2.3437500000000002e-06, | |
| "loss": 0.0068, | |
| "reward": 0.704166678339243, | |
| "reward_std": 0.11830127276480198, | |
| "rewards/accuracy_reward": 0.704166678339243, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 545.3708541870117, | |
| "epoch": 0.096, | |
| "grad_norm": 0.08740234375, | |
| "kl": 0.00010152916220249609, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.0204, | |
| "reward": 0.7583333522081375, | |
| "reward_std": 0.13496793992817402, | |
| "rewards/accuracy_reward": 0.7583333522081375, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 550.9000144958496, | |
| "epoch": 0.112, | |
| "grad_norm": 0.08984375, | |
| "kl": 0.00010129613328899723, | |
| "learning_rate": 2.9991503375003e-06, | |
| "loss": 0.0164, | |
| "reward": 0.7979166805744171, | |
| "reward_std": 0.12246793955564499, | |
| "rewards/accuracy_reward": 0.7979166805744171, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 557.5000205993653, | |
| "epoch": 0.128, | |
| "grad_norm": 0.0673828125, | |
| "kl": 0.00010441835365782026, | |
| "learning_rate": 2.993961440992859e-06, | |
| "loss": 0.0179, | |
| "reward": 0.7791666835546494, | |
| "reward_std": 0.14330127350986005, | |
| "rewards/accuracy_reward": 0.7791666835546494, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 547.3437675476074, | |
| "epoch": 0.144, | |
| "grad_norm": 0.07763671875, | |
| "kl": 0.00010777117104225909, | |
| "learning_rate": 2.984071989079555e-06, | |
| "loss": 0.0222, | |
| "reward": 0.7750000141561031, | |
| "reward_std": 0.15867876969277858, | |
| "rewards/accuracy_reward": 0.7750000141561031, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 541.658351135254, | |
| "epoch": 0.16, | |
| "grad_norm": 0.0634765625, | |
| "kl": 0.00010541909759922418, | |
| "learning_rate": 2.9695130976348534e-06, | |
| "loss": 0.0155, | |
| "reward": 0.7854166805744172, | |
| "reward_std": 0.11988959684967995, | |
| "rewards/accuracy_reward": 0.7854166805744172, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 553.2916847229004, | |
| "epoch": 0.176, | |
| "grad_norm": 0.10693359375, | |
| "kl": 0.00011168952341904515, | |
| "learning_rate": 2.9503305743175096e-06, | |
| "loss": 0.0131, | |
| "reward": 0.7145833455026149, | |
| "reward_std": 0.16448003873229028, | |
| "rewards/accuracy_reward": 0.7145833455026149, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 575.1771041870118, | |
| "epoch": 0.192, | |
| "grad_norm": 0.044189453125, | |
| "kl": 9.437939497729531e-05, | |
| "learning_rate": 2.9265847744427307e-06, | |
| "loss": 0.0192, | |
| "reward": 0.7625000149011611, | |
| "reward_std": 0.14587961547076703, | |
| "rewards/accuracy_reward": 0.7625000149011611, | |
| "rewards/format_reward": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 519.7854316711425, | |
| "epoch": 0.208, | |
| "grad_norm": 0.042724609375, | |
| "kl": 0.00010650227795849787, | |
| "learning_rate": 2.8983504110820214e-06, | |
| "loss": 0.0206, | |
| "reward": 0.7937500163912773, | |
| "reward_std": 0.11636751592159271, | |
| "rewards/accuracy_reward": 0.7937500163912773, | |
| "rewards/format_reward": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 517.5646003723144, | |
| "epoch": 0.224, | |
| "grad_norm": 0.0439453125, | |
| "kl": 0.00010277715955453459, | |
| "learning_rate": 2.865716319988224e-06, | |
| "loss": 0.0221, | |
| "reward": 0.7770833492279052, | |
| "reward_std": 0.12727919183671474, | |
| "rewards/accuracy_reward": 0.7770833492279052, | |
| "rewards/format_reward": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 571.9395950317382, | |
| "epoch": 0.24, | |
| "grad_norm": 0.1064453125, | |
| "kl": 0.00011016046828444814, | |
| "learning_rate": 2.82878518008537e-06, | |
| "loss": 0.0138, | |
| "reward": 0.7729166805744171, | |
| "reward_std": 0.14394585900008677, | |
| "rewards/accuracy_reward": 0.7729166805744171, | |
| "rewards/format_reward": 0.0, | |
| "step": 75 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 547.0708488464355, | |
| "epoch": 0.256, | |
| "grad_norm": 0.07421875, | |
| "kl": 0.00010552951316640246, | |
| "learning_rate": 2.7876731904027993e-06, | |
| "loss": 0.0043, | |
| "reward": 0.7833333518356085, | |
| "reward_std": 0.1494016967713833, | |
| "rewards/accuracy_reward": 0.7833333518356085, | |
| "rewards/format_reward": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 528.5791786193847, | |
| "epoch": 0.272, | |
| "grad_norm": 0.07373046875, | |
| "kl": 0.00010855200152946055, | |
| "learning_rate": 2.7425097044700246e-06, | |
| "loss": 0.0083, | |
| "reward": 0.8020833432674408, | |
| "reward_std": 0.14265668764710426, | |
| "rewards/accuracy_reward": 0.8020833432674408, | |
| "rewards/format_reward": 0.0, | |
| "step": 85 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 578.5750167846679, | |
| "epoch": 0.288, | |
| "grad_norm": 0.09765625, | |
| "kl": 0.00011546339583219378, | |
| "learning_rate": 2.6934368233226715e-06, | |
| "loss": 0.0211, | |
| "reward": 0.7250000156462193, | |
| "reward_std": 0.16125711128115655, | |
| "rewards/accuracy_reward": 0.7250000156462193, | |
| "rewards/format_reward": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 548.235432434082, | |
| "epoch": 0.304, | |
| "grad_norm": 0.1123046875, | |
| "kl": 0.00010799223509820877, | |
| "learning_rate": 2.6406089484000465e-06, | |
| "loss": 0.0284, | |
| "reward": 0.8083333469927311, | |
| "reward_std": 0.16830127350986004, | |
| "rewards/accuracy_reward": 0.8083333469927311, | |
| "rewards/format_reward": 0.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 2.584192295741087e-06, | |
| "loss": 0.0186, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_clip_ratio": 0.0, | |
| "eval_completion_length": 464.36827672170307, | |
| "eval_kl": 8.895910598325799e-05, | |
| "eval_loss": 0.004563388880342245, | |
| "eval_reward": 0.8485529082799386, | |
| "eval_reward_std": 0.13733045984349565, | |
| "eval_rewards/accuracy_reward": 0.8485529082799386, | |
| "eval_rewards/format_reward": 0.0, | |
| "eval_runtime": 1240.0045, | |
| "eval_samples_per_second": 0.806, | |
| "eval_steps_per_second": 0.068, | |
| "step": 100 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 528.855224609375, | |
| "epoch": 0.336, | |
| "grad_norm": 0.05615234375, | |
| "kl": 0.00010595467788334644, | |
| "learning_rate": 2.5243643730072105e-06, | |
| "loss": 0.0102, | |
| "reward": 0.7875000132247806, | |
| "reward_std": 0.12053418289870024, | |
| "rewards/accuracy_reward": 0.7875000132247806, | |
| "rewards/format_reward": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 569.627101135254, | |
| "epoch": 0.352, | |
| "grad_norm": 0.06982421875, | |
| "kl": 0.00010751374265964842, | |
| "learning_rate": 2.461313420977536e-06, | |
| "loss": 0.0214, | |
| "reward": 0.7270833492279053, | |
| "reward_std": 0.13432335443794727, | |
| "rewards/accuracy_reward": 0.7270833492279053, | |
| "rewards/format_reward": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 546.0083457946778, | |
| "epoch": 0.368, | |
| "grad_norm": 0.059326171875, | |
| "kl": 9.366698777739657e-05, | |
| "learning_rate": 2.3952378212737554e-06, | |
| "loss": 0.0181, | |
| "reward": 0.8020833432674408, | |
| "reward_std": 0.13432335443794727, | |
| "rewards/accuracy_reward": 0.8020833432674408, | |
| "rewards/format_reward": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 554.027099609375, | |
| "epoch": 0.384, | |
| "grad_norm": 0.08642578125, | |
| "kl": 0.00010431327682454139, | |
| "learning_rate": 2.3263454721781537e-06, | |
| "loss": 0.0112, | |
| "reward": 0.8020833477377891, | |
| "reward_std": 0.10803418271243573, | |
| "rewards/accuracy_reward": 0.8020833477377891, | |
| "rewards/format_reward": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 559.7708488464356, | |
| "epoch": 0.4, | |
| "grad_norm": 0.07421875, | |
| "kl": 0.00010436618849780643, | |
| "learning_rate": 2.2548531345087003e-06, | |
| "loss": 0.0268, | |
| "reward": 0.7500000167638063, | |
| "reward_std": 0.115722930803895, | |
| "rewards/accuracy_reward": 0.7500000167638063, | |
| "rewards/format_reward": 0.0, | |
| "step": 125 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 514.7854347229004, | |
| "epoch": 0.416, | |
| "grad_norm": 0.07470703125, | |
| "kl": 0.00010219837422482669, | |
| "learning_rate": 2.18098574960932e-06, | |
| "loss": 0.0252, | |
| "reward": 0.8145833492279053, | |
| "reward_std": 0.1509900216013193, | |
| "rewards/accuracy_reward": 0.8145833492279053, | |
| "rewards/format_reward": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 581.3895988464355, | |
| "epoch": 0.432, | |
| "grad_norm": 0.057861328125, | |
| "kl": 0.0001146518128734897, | |
| "learning_rate": 2.104975731601208e-06, | |
| "loss": 0.0239, | |
| "reward": 0.6833333477377892, | |
| "reward_std": 0.154212948307395, | |
| "rewards/accuracy_reward": 0.6833333477377892, | |
| "rewards/format_reward": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 565.6645988464355, | |
| "epoch": 0.448, | |
| "grad_norm": 0.04931640625, | |
| "kl": 0.00011206413109903224, | |
| "learning_rate": 2.027062236122014e-06, | |
| "loss": 0.0109, | |
| "reward": 0.768750012293458, | |
| "reward_std": 0.15227919220924377, | |
| "rewards/accuracy_reward": 0.768750012293458, | |
| "rewards/format_reward": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 545.1396034240722, | |
| "epoch": 0.464, | |
| "grad_norm": 0.0732421875, | |
| "kl": 0.00010298133001924725, | |
| "learning_rate": 1.9474904078537343e-06, | |
| "loss": 0.0186, | |
| "reward": 0.7250000163912773, | |
| "reward_std": 0.14940169639885426, | |
| "rewards/accuracy_reward": 0.7250000163912773, | |
| "rewards/format_reward": 0.0, | |
| "step": 145 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 541.8229347229004, | |
| "epoch": 0.48, | |
| "grad_norm": 0.06591796875, | |
| "kl": 9.864068970273365e-05, | |
| "learning_rate": 1.866510609206841e-06, | |
| "loss": 0.0136, | |
| "reward": 0.7854166857898235, | |
| "reward_std": 0.16319086775183678, | |
| "rewards/accuracy_reward": 0.7854166857898235, | |
| "rewards/format_reward": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 571.9604377746582, | |
| "epoch": 0.496, | |
| "grad_norm": 0.057861328125, | |
| "kl": 0.00010758389435068238, | |
| "learning_rate": 1.784377632587518e-06, | |
| "loss": 0.0095, | |
| "reward": 0.7041666833683848, | |
| "reward_std": 0.18754628263413906, | |
| "rewards/accuracy_reward": 0.7041666833683848, | |
| "rewards/format_reward": 0.0, | |
| "step": 155 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 602.4896034240722, | |
| "epoch": 0.512, | |
| "grad_norm": 0.05908203125, | |
| "kl": 0.0001044012439706421, | |
| "learning_rate": 1.7013498987264833e-06, | |
| "loss": 0.022, | |
| "reward": 0.6875000156462192, | |
| "reward_std": 0.1670121032744646, | |
| "rewards/accuracy_reward": 0.6875000156462192, | |
| "rewards/format_reward": 0.0, | |
| "step": 160 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 537.7500205993653, | |
| "epoch": 0.528, | |
| "grad_norm": 0.10498046875, | |
| "kl": 0.00010827416845131666, | |
| "learning_rate": 1.6176886435917677e-06, | |
| "loss": 0.0067, | |
| "reward": 0.7562500171363353, | |
| "reward_std": 0.14875711016356946, | |
| "rewards/accuracy_reward": 0.7562500171363353, | |
| "rewards/format_reward": 0.0, | |
| "step": 165 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 561.8604354858398, | |
| "epoch": 0.544, | |
| "grad_norm": 0.0634765625, | |
| "kl": 0.00010869488978642039, | |
| "learning_rate": 1.5336570964437077e-06, | |
| "loss": 0.0288, | |
| "reward": 0.722916679829359, | |
| "reward_std": 0.18080127388238906, | |
| "rewards/accuracy_reward": 0.722916679829359, | |
| "rewards/format_reward": 0.0, | |
| "step": 170 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 553.4791847229004, | |
| "epoch": 0.56, | |
| "grad_norm": 0.046142578125, | |
| "kl": 0.00011422934894653736, | |
| "learning_rate": 1.4495196516183096e-06, | |
| "loss": 0.0057, | |
| "reward": 0.7166666770353913, | |
| "reward_std": 0.13144585788249968, | |
| "rewards/accuracy_reward": 0.7166666770353913, | |
| "rewards/format_reward": 0.0, | |
| "step": 175 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 546.4104309082031, | |
| "epoch": 0.576, | |
| "grad_norm": 0.1083984375, | |
| "kl": 0.00010774649126688018, | |
| "learning_rate": 1.3655410366448499e-06, | |
| "loss": 0.0158, | |
| "reward": 0.7833333492279053, | |
| "reward_std": 0.159967939928174, | |
| "rewards/accuracy_reward": 0.7833333492279053, | |
| "rewards/format_reward": 0.0, | |
| "step": 180 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 602.3583526611328, | |
| "epoch": 0.592, | |
| "grad_norm": 0.0673828125, | |
| "kl": 0.00010071156602862175, | |
| "learning_rate": 1.2819854793151313e-06, | |
| "loss": 0.0106, | |
| "reward": 0.6895833477377892, | |
| "reward_std": 0.139134606346488, | |
| "rewards/accuracy_reward": 0.6895833477377892, | |
| "rewards/format_reward": 0.0, | |
| "step": 185 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 560.5500137329102, | |
| "epoch": 0.608, | |
| "grad_norm": 0.06640625, | |
| "kl": 0.00011235472247790312, | |
| "learning_rate": 1.199115876325091e-06, | |
| "loss": 0.0011, | |
| "reward": 0.7354166826233268, | |
| "reward_std": 0.17727919295430183, | |
| "rewards/accuracy_reward": 0.7354166826233268, | |
| "rewards/format_reward": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 556.4875118255616, | |
| "epoch": 0.624, | |
| "grad_norm": 0.04736328125, | |
| "kl": 0.00011359924974385649, | |
| "learning_rate": 1.1171929661045361e-06, | |
| "loss": 0.0149, | |
| "reward": 0.7500000115483999, | |
| "reward_std": 0.12405626438558101, | |
| "rewards/accuracy_reward": 0.7500000115483999, | |
| "rewards/format_reward": 0.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.044189453125, | |
| "learning_rate": 1.036474508437579e-06, | |
| "loss": 0.0226, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_clip_ratio": 0.0, | |
| "eval_completion_length": 465.14322685196015, | |
| "eval_kl": 8.934064043752367e-05, | |
| "eval_loss": 0.003860337659716606, | |
| "eval_reward": 0.8478044058212977, | |
| "eval_reward_std": 0.1438588550883139, | |
| "eval_rewards/accuracy_reward": 0.8478044058212977, | |
| "eval_rewards/format_reward": 0.0, | |
| "eval_runtime": 1207.4697, | |
| "eval_samples_per_second": 0.828, | |
| "eval_steps_per_second": 0.07, | |
| "step": 200 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 535.7416812896729, | |
| "epoch": 0.656, | |
| "grad_norm": 0.06982421875, | |
| "kl": 0.00011015057557415276, | |
| "learning_rate": 9.57214473454992e-07, | |
| "loss": 0.0301, | |
| "reward": 0.7635416811332106, | |
| "reward_std": 0.14010148495435715, | |
| "rewards/accuracy_reward": 0.7635416811332106, | |
| "rewards/format_reward": 0.0, | |
| "step": 205 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 556.1916847229004, | |
| "epoch": 0.672, | |
| "grad_norm": 0.04296875, | |
| "kl": 0.00010989538504873053, | |
| "learning_rate": 8.796622425502193e-07, | |
| "loss": 0.0141, | |
| "reward": 0.735416679084301, | |
| "reward_std": 0.130801273137331, | |
| "rewards/accuracy_reward": 0.735416679084301, | |
| "rewards/format_reward": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 600.8562660217285, | |
| "epoch": 0.688, | |
| "grad_norm": 0.060791015625, | |
| "kl": 0.00011158880770381074, | |
| "learning_rate": 8.040618237332491e-07, | |
| "loss": 0.0276, | |
| "reward": 0.6958333497866989, | |
| "reward_std": 0.1936467058956623, | |
| "rewards/accuracy_reward": 0.6958333497866989, | |
| "rewards/format_reward": 0.0, | |
| "step": 215 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 560.8437644958497, | |
| "epoch": 0.704, | |
| "grad_norm": 0.05029296875, | |
| "kl": 0.00010739189892774447, | |
| "learning_rate": 7.30651083891141e-07, | |
| "loss": 0.0285, | |
| "reward": 0.7666666835546494, | |
| "reward_std": 0.12534543462097644, | |
| "rewards/accuracy_reward": 0.7666666835546494, | |
| "rewards/format_reward": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 580.1146057128906, | |
| "epoch": 0.72, | |
| "grad_norm": 0.0537109375, | |
| "kl": 0.00010475763965587249, | |
| "learning_rate": 6.596610003707959e-07, | |
| "loss": 0.0188, | |
| "reward": 0.7625000156462193, | |
| "reward_std": 0.14330127313733101, | |
| "rewards/accuracy_reward": 0.7625000156462193, | |
| "rewards/format_reward": 0.0, | |
| "step": 225 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 600.7646011352539, | |
| "epoch": 0.736, | |
| "grad_norm": 0.07958984375, | |
| "kl": 0.00010898288783209865, | |
| "learning_rate": 5.913149342387704e-07, | |
| "loss": 0.0155, | |
| "reward": 0.7000000149011611, | |
| "reward_std": 0.1673575345426798, | |
| "rewards/accuracy_reward": 0.7000000149011611, | |
| "rewards/format_reward": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 541.1729309082032, | |
| "epoch": 0.752, | |
| "grad_norm": 0.07421875, | |
| "kl": 0.00011250958050368354, | |
| "learning_rate": 5.258279275047247e-07, | |
| "loss": 0.0154, | |
| "reward": 0.7770833522081375, | |
| "reward_std": 0.12504628151655198, | |
| "rewards/accuracy_reward": 0.7770833522081375, | |
| "rewards/format_reward": 0.0, | |
| "step": 235 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 552.583349609375, | |
| "epoch": 0.768, | |
| "grad_norm": 0.0751953125, | |
| "kl": 0.00011499359970912337, | |
| "learning_rate": 4.63406026519703e-07, | |
| "loss": 0.0189, | |
| "reward": 0.7812500141561032, | |
| "reward_std": 0.11284543462097645, | |
| "rewards/accuracy_reward": 0.7812500141561032, | |
| "rewards/format_reward": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 573.3916809082032, | |
| "epoch": 0.784, | |
| "grad_norm": 0.080078125, | |
| "kl": 0.00011338020558468998, | |
| "learning_rate": 4.042456336780838e-07, | |
| "loss": 0.0157, | |
| "reward": 0.7458333499729634, | |
| "reward_std": 0.14106836169958115, | |
| "rewards/accuracy_reward": 0.7458333499729634, | |
| "rewards/format_reward": 0.0, | |
| "step": 245 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 570.3416870117187, | |
| "epoch": 0.8, | |
| "grad_norm": 0.07666015625, | |
| "kl": 0.00010969964605465066, | |
| "learning_rate": 3.4853288946298335e-07, | |
| "loss": 0.0117, | |
| "reward": 0.7062500147148967, | |
| "reward_std": 0.16413460709154606, | |
| "rewards/accuracy_reward": 0.7062500147148967, | |
| "rewards/format_reward": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 555.5708526611328, | |
| "epoch": 0.816, | |
| "grad_norm": 0.0859375, | |
| "kl": 0.00011466140349511989, | |
| "learning_rate": 2.9644308677943315e-07, | |
| "loss": 0.0341, | |
| "reward": 0.7625000171363354, | |
| "reward_std": 0.1901246253401041, | |
| "rewards/accuracy_reward": 0.7625000171363354, | |
| "rewards/format_reward": 0.0, | |
| "step": 255 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 559.4687660217285, | |
| "epoch": 0.832, | |
| "grad_norm": 0.09326171875, | |
| "kl": 0.00011700030408974271, | |
| "learning_rate": 2.48140119418046e-07, | |
| "loss": 0.0038, | |
| "reward": 0.7437500134110451, | |
| "reward_std": 0.16988959908485413, | |
| "rewards/accuracy_reward": 0.7437500134110451, | |
| "rewards/format_reward": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 586.2000167846679, | |
| "epoch": 0.848, | |
| "grad_norm": 0.0673828125, | |
| "kl": 0.00010750685323728249, | |
| "learning_rate": 2.0377596638451812e-07, | |
| "loss": 0.0151, | |
| "reward": 0.7458333522081375, | |
| "reward_std": 0.14459044374525548, | |
| "rewards/accuracy_reward": 0.7458333522081375, | |
| "rewards/format_reward": 0.0, | |
| "step": 265 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 559.5666870117187, | |
| "epoch": 0.864, | |
| "grad_norm": 0.040283203125, | |
| "kl": 0.00010784281039377674, | |
| "learning_rate": 1.634902137174483e-07, | |
| "loss": 0.0087, | |
| "reward": 0.7375000141561031, | |
| "reward_std": 0.11349002048373222, | |
| "rewards/accuracy_reward": 0.7375000141561031, | |
| "rewards/format_reward": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 559.8062644958496, | |
| "epoch": 0.88, | |
| "grad_norm": 0.0296630859375, | |
| "kl": 0.00010248722046526381, | |
| "learning_rate": 1.274096152990203e-07, | |
| "loss": 0.0123, | |
| "reward": 0.7833333499729633, | |
| "reward_std": 0.12182335332036018, | |
| "rewards/accuracy_reward": 0.7833333499729633, | |
| "rewards/format_reward": 0.0, | |
| "step": 275 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 560.8646018981933, | |
| "epoch": 0.896, | |
| "grad_norm": 0.08349609375, | |
| "kl": 0.00010658866949597723, | |
| "learning_rate": 9.564769404039419e-08, | |
| "loss": 0.024, | |
| "reward": 0.7583333499729633, | |
| "reward_std": 0.14459044449031352, | |
| "rewards/accuracy_reward": 0.7583333499729633, | |
| "rewards/format_reward": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 572.097932434082, | |
| "epoch": 0.912, | |
| "grad_norm": 0.0947265625, | |
| "kl": 0.00011029940260414151, | |
| "learning_rate": 6.830438469662892e-08, | |
| "loss": 0.0198, | |
| "reward": 0.7250000156462193, | |
| "reward_std": 0.14845795668661593, | |
| "rewards/accuracy_reward": 0.7250000156462193, | |
| "rewards/format_reward": 0.0, | |
| "step": 285 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 540.3437705993653, | |
| "epoch": 0.928, | |
| "grad_norm": 0.0517578125, | |
| "kl": 0.00010695143555494724, | |
| "learning_rate": 4.546571943496969e-08, | |
| "loss": 0.0248, | |
| "reward": 0.818750013411045, | |
| "reward_std": 0.13784543573856353, | |
| "rewards/accuracy_reward": 0.818750013411045, | |
| "rewards/format_reward": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 581.6541847229004, | |
| "epoch": 0.944, | |
| "grad_norm": 0.068359375, | |
| "kl": 0.00010272065319441026, | |
| "learning_rate": 2.72035571458224e-08, | |
| "loss": 0.0252, | |
| "reward": 0.7541666850447655, | |
| "reward_std": 0.16477919220924378, | |
| "rewards/accuracy_reward": 0.7541666850447655, | |
| "rewards/format_reward": 0.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 1.357535734809795e-08, | |
| "loss": 0.0261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_clip_ratio": 0.0, | |
| "eval_completion_length": 464.5865904756649, | |
| "eval_kl": 8.974588041372066e-05, | |
| "eval_loss": 0.004327021539211273, | |
| "eval_reward": 0.8458083985808367, | |
| "eval_reward_std": 0.14224328386212537, | |
| "eval_rewards/accuracy_reward": 0.8458083985808367, | |
| "eval_rewards/format_reward": 0.0, | |
| "eval_runtime": 1260.0209, | |
| "eval_samples_per_second": 0.794, | |
| "eval_steps_per_second": 0.067, | |
| "step": 300 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 563.8583469390869, | |
| "epoch": 0.976, | |
| "grad_norm": 0.279296875, | |
| "kl": 0.00011092701979578123, | |
| "learning_rate": 4.623999400308054e-09, | |
| "loss": 0.0154, | |
| "reward": 0.7604166809469461, | |
| "reward_std": 0.14394585862755777, | |
| "rewards/accuracy_reward": 0.7604166809469461, | |
| "rewards/format_reward": 0.0, | |
| "step": 305 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 554.9770988464355, | |
| "epoch": 0.992, | |
| "grad_norm": 0.07373046875, | |
| "kl": 0.00010652115543052787, | |
| "learning_rate": 3.77647586240204e-10, | |
| "loss": 0.0119, | |
| "reward": 0.7562500186264515, | |
| "reward_std": 0.1330341823399067, | |
| "rewards/accuracy_reward": 0.7562500186264515, | |
| "rewards/format_reward": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 575.7395992279053, | |
| "epoch": 0.9984, | |
| "kl": 0.00010837154786713654, | |
| "reward": 0.7760416828095913, | |
| "reward_std": 0.17472399026155472, | |
| "rewards/accuracy_reward": 0.7760416828095913, | |
| "rewards/format_reward": 0.0, | |
| "step": 312, | |
| "total_flos": 0.0, | |
| "train_loss": 0.017263087375352208, | |
| "train_runtime": 14335.1534, | |
| "train_samples_per_second": 0.523, | |
| "train_steps_per_second": 0.022 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |