| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9996190476190476, | |
| "eval_steps": 100, | |
| "global_step": 451, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.022164502164502164, | |
| "grad_norm": 2.910351514816284, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 1.5193, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04432900432900433, | |
| "grad_norm": 2.435637950897217, | |
| "learning_rate": 4.997674683630404e-06, | |
| "loss": 1.5122, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0664935064935065, | |
| "grad_norm": 2.5778419971466064, | |
| "learning_rate": 4.983480078908139e-06, | |
| "loss": 1.5007, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08865800865800866, | |
| "grad_norm": 3.0028574466705322, | |
| "learning_rate": 4.956455951420386e-06, | |
| "loss": 1.5252, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11082251082251082, | |
| "grad_norm": 2.561404228210449, | |
| "learning_rate": 4.916741906337227e-06, | |
| "loss": 1.4841, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.132987012987013, | |
| "grad_norm": 2.3713862895965576, | |
| "learning_rate": 4.864543104251587e-06, | |
| "loss": 1.4968, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15515151515151515, | |
| "grad_norm": 2.773104429244995, | |
| "learning_rate": 4.800129201330785e-06, | |
| "loss": 1.5239, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17731601731601732, | |
| "grad_norm": 2.6305017471313477, | |
| "learning_rate": 4.723832956287465e-06, | |
| "loss": 1.5138, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19948051948051948, | |
| "grad_norm": 2.8621950149536133, | |
| "learning_rate": 4.636048511366222e-06, | |
| "loss": 1.4801, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22164502164502164, | |
| "grad_norm": 2.79162335395813, | |
| "learning_rate": 4.537229356226238e-06, | |
| "loss": 1.509, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22164502164502164, | |
| "eval_dpo_accuracy": 0.4307142857142857, | |
| "eval_reward_accuracy": 0.8981428571428571, | |
| "eval_runtime": 77.5047, | |
| "eval_samples_per_second": 90.317, | |
| "eval_steps_per_second": 11.29, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2438095238095238, | |
| "grad_norm": 2.651031494140625, | |
| "learning_rate": 4.427885985238417e-06, | |
| "loss": 1.4879, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.265974025974026, | |
| "grad_norm": 2.7951595783233643, | |
| "learning_rate": 4.308583260299282e-06, | |
| "loss": 1.5208, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.28813852813852814, | |
| "grad_norm": 2.712519407272339, | |
| "learning_rate": 4.179937492785189e-06, | |
| "loss": 1.5048, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3103030303030303, | |
| "grad_norm": 2.770768642425537, | |
| "learning_rate": 4.0426132597213485e-06, | |
| "loss": 1.5152, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.33246753246753247, | |
| "grad_norm": 2.7763257026672363, | |
| "learning_rate": 3.897319970613119e-06, | |
| "loss": 1.5328, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.35463203463203463, | |
| "grad_norm": 2.4849092960357666, | |
| "learning_rate": 3.7448082026751594e-06, | |
| "loss": 1.5193, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3767965367965368, | |
| "grad_norm": 2.8126847743988037, | |
| "learning_rate": 3.5858658233904335e-06, | |
| "loss": 1.4932, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.39896103896103896, | |
| "grad_norm": 2.7220044136047363, | |
| "learning_rate": 3.4213139204297306e-06, | |
| "loss": 1.5076, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4211255411255411, | |
| "grad_norm": 2.623095750808716, | |
| "learning_rate": 3.2520025599574878e-06, | |
| "loss": 1.5336, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4432900432900433, | |
| "grad_norm": 2.739753484725952, | |
| "learning_rate": 3.078806395236313e-06, | |
| "loss": 1.5066, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4432900432900433, | |
| "eval_dpo_accuracy": 0.43028571428571427, | |
| "eval_reward_accuracy": 0.8991428571428571, | |
| "eval_runtime": 77.4491, | |
| "eval_samples_per_second": 90.382, | |
| "eval_steps_per_second": 11.298, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.46545454545454545, | |
| "grad_norm": 2.728768825531006, | |
| "learning_rate": 2.9026201482158826e-06, | |
| "loss": 1.4946, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4876190476190476, | |
| "grad_norm": 2.8276684284210205, | |
| "learning_rate": 2.724353987448107e-06, | |
| "loss": 1.4973, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5097835497835498, | |
| "grad_norm": 2.889288902282715, | |
| "learning_rate": 2.5449288262059803e-06, | |
| "loss": 1.4954, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.531948051948052, | |
| "grad_norm": 2.7917494773864746, | |
| "learning_rate": 2.365271565095797e-06, | |
| "loss": 1.5146, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5541125541125541, | |
| "grad_norm": 2.705782890319824, | |
| "learning_rate": 2.186310303739121e-06, | |
| "loss": 1.522, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5762770562770563, | |
| "grad_norm": 2.7684695720672607, | |
| "learning_rate": 2.008969546260717e-06, | |
| "loss": 1.4967, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5984415584415584, | |
| "grad_norm": 2.9542436599731445, | |
| "learning_rate": 1.8341654253506098e-06, | |
| "loss": 1.4849, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6206060606060606, | |
| "grad_norm": 2.868518590927124, | |
| "learning_rate": 1.6628009695725348e-06, | |
| "loss": 1.5097, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6427705627705628, | |
| "grad_norm": 2.629920244216919, | |
| "learning_rate": 1.495761438367577e-06, | |
| "loss": 1.4932, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "grad_norm": 2.820434331893921, | |
| "learning_rate": 1.3339097488521364e-06, | |
| "loss": 1.5209, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "eval_dpo_accuracy": 0.4298571428571429, | |
| "eval_reward_accuracy": 0.8994285714285715, | |
| "eval_runtime": 77.4739, | |
| "eval_samples_per_second": 90.353, | |
| "eval_steps_per_second": 11.294, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6870995670995671, | |
| "grad_norm": 2.696439743041992, | |
| "learning_rate": 1.1780820180351262e-06, | |
| "loss": 1.4872, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7092640692640693, | |
| "grad_norm": 2.8368752002716064, | |
| "learning_rate": 1.029083243483064e-06, | |
| "loss": 1.5035, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7314285714285714, | |
| "grad_norm": 2.790517568588257, | |
| "learning_rate": 8.876831447465015e-07, | |
| "loss": 1.4983, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7535930735930736, | |
| "grad_norm": 2.788957118988037, | |
| "learning_rate": 7.546121870307743e-07, | |
| "loss": 1.51, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7757575757575758, | |
| "grad_norm": 2.6294381618499756, | |
| "learning_rate": 6.305578076525481e-07, | |
| "loss": 1.4995, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7979220779220779, | |
| "grad_norm": 2.827662467956543, | |
| "learning_rate": 5.161608647761013e-07, | |
| "loss": 1.4912, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8200865800865801, | |
| "grad_norm": 2.721107006072998, | |
| "learning_rate": 4.1201232677496846e-07, | |
| "loss": 1.5299, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8422510822510823, | |
| "grad_norm": 2.850194215774536, | |
| "learning_rate": 3.186502193215221e-07, | |
| "loss": 1.517, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8644155844155844, | |
| "grad_norm": 2.7145802974700928, | |
| "learning_rate": 2.3655684597566563e-07, | |
| "loss": 1.4859, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8865800865800866, | |
| "grad_norm": 2.9116082191467285, | |
| "learning_rate": 1.6615629663092492e-07, | |
| "loss": 1.4921, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8865800865800866, | |
| "eval_dpo_accuracy": 0.43, | |
| "eval_reward_accuracy": 0.899, | |
| "eval_runtime": 77.552, | |
| "eval_samples_per_second": 90.262, | |
| "eval_steps_per_second": 11.283, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9087445887445887, | |
| "grad_norm": 2.624758243560791, | |
| "learning_rate": 1.078122566891654e-07, | |
| "loss": 1.5212, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9309090909090909, | |
| "grad_norm": 2.613771677017212, | |
| "learning_rate": 6.182612828162154e-08, | |
| "loss": 1.4931, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9530735930735931, | |
| "grad_norm": 2.632521390914917, | |
| "learning_rate": 2.843547324191259e-08, | |
| "loss": 1.512, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9752380952380952, | |
| "grad_norm": 2.524284601211548, | |
| "learning_rate": 7.81278587455303e-09, | |
| "loss": 1.4976, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "grad_norm": 2.6475911140441895, | |
| "learning_rate": 6.460185877465997e-11, | |
| "loss": 1.5115, | |
| "step": 450 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 451, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |