| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 3, | |
| "global_step": 39, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 6.9555134773254395, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1647, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "eval_loss": 2.2015950679779053, | |
| "eval_runtime": 0.5143, | |
| "eval_samples_per_second": 42.778, | |
| "eval_steps_per_second": 1.944, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 7.17453670501709, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1358, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 7.099859714508057, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1505, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "eval_loss": 2.116824150085449, | |
| "eval_runtime": 0.5016, | |
| "eval_samples_per_second": 43.86, | |
| "eval_steps_per_second": 1.994, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 7.153745651245117, | |
| "learning_rate": 4e-05, | |
| "loss": 2.1955, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 5.896260738372803, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9378, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 5.061944484710693, | |
| "learning_rate": 6e-05, | |
| "loss": 1.7332, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "eval_loss": 1.5603915452957153, | |
| "eval_runtime": 0.4962, | |
| "eval_samples_per_second": 44.34, | |
| "eval_steps_per_second": 2.015, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 4.622457981109619, | |
| "learning_rate": 7e-05, | |
| "loss": 1.5975, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 4.024200439453125, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3523, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 3.341926097869873, | |
| "learning_rate": 9e-05, | |
| "loss": 1.0807, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "eval_loss": 0.878767728805542, | |
| "eval_runtime": 0.5037, | |
| "eval_samples_per_second": 43.68, | |
| "eval_steps_per_second": 1.985, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 3.572037696838379, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8712, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 2.6981351375579834, | |
| "learning_rate": 9.986018985905901e-05, | |
| "loss": 0.694, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 2.876271963119507, | |
| "learning_rate": 9.944154131125642e-05, | |
| "loss": 0.5284, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "eval_loss": 0.4853472411632538, | |
| "eval_runtime": 0.5088, | |
| "eval_samples_per_second": 43.236, | |
| "eval_steps_per_second": 1.965, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.3492231369018555, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.4715, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 1.5317707061767578, | |
| "learning_rate": 9.777864028930705e-05, | |
| "loss": 0.3566, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 1.37990140914917, | |
| "learning_rate": 9.654368743221022e-05, | |
| "loss": 0.3215, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "eval_loss": 0.291117787361145, | |
| "eval_runtime": 0.5024, | |
| "eval_samples_per_second": 43.789, | |
| "eval_steps_per_second": 1.99, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 1.4043002128601074, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.2691, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 1.0550378561019897, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.2266, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.9474859833717346, | |
| "learning_rate": 9.131193871579975e-05, | |
| "loss": 0.2114, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "eval_loss": 0.19578741490840912, | |
| "eval_runtime": 0.503, | |
| "eval_samples_per_second": 43.739, | |
| "eval_steps_per_second": 1.988, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.984850287437439, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1919, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.8843994140625, | |
| "learning_rate": 8.665259359149132e-05, | |
| "loss": 0.1432, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.8318038582801819, | |
| "learning_rate": 8.400863688854597e-05, | |
| "loss": 0.1493, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "eval_loss": 0.1373557448387146, | |
| "eval_runtime": 0.5325, | |
| "eval_samples_per_second": 41.316, | |
| "eval_steps_per_second": 1.878, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.7481142282485962, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1359, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.777579665184021, | |
| "learning_rate": 7.81660029031811e-05, | |
| "loss": 0.1135, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.6446991562843323, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.1081, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "eval_loss": 0.10664429515600204, | |
| "eval_runtime": 0.4983, | |
| "eval_samples_per_second": 44.146, | |
| "eval_steps_per_second": 2.007, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.699760913848877, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.104, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7046389579772949, | |
| "learning_rate": 6.826705121831976e-05, | |
| "loss": 0.0943, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.49757829308509827, | |
| "learning_rate": 6.473775872054521e-05, | |
| "loss": 0.0751, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "eval_loss": 0.0820903405547142, | |
| "eval_runtime": 0.5036, | |
| "eval_samples_per_second": 43.688, | |
| "eval_steps_per_second": 1.986, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.482883483171463, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.0692, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.5144394040107727, | |
| "learning_rate": 5.745211330880872e-05, | |
| "loss": 0.0641, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.739709734916687, | |
| "learning_rate": 5.373650467932122e-05, | |
| "loss": 0.0782, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "eval_loss": 0.06886172294616699, | |
| "eval_runtime": 0.5049, | |
| "eval_samples_per_second": 43.576, | |
| "eval_steps_per_second": 1.981, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.5622302889823914, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0559, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.5018132328987122, | |
| "learning_rate": 4.626349532067879e-05, | |
| "loss": 0.0586, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.40318235754966736, | |
| "learning_rate": 4.254788669119127e-05, | |
| "loss": 0.0524, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "eval_loss": 0.06022738292813301, | |
| "eval_runtime": 0.5037, | |
| "eval_samples_per_second": 43.679, | |
| "eval_steps_per_second": 1.985, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.41394540667533875, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0414, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.49381062388420105, | |
| "learning_rate": 3.5262241279454785e-05, | |
| "loss": 0.0495, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.6479844450950623, | |
| "learning_rate": 3.173294878168025e-05, | |
| "loss": 0.0538, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "eval_loss": 0.05229267105460167, | |
| "eval_runtime": 0.4971, | |
| "eval_samples_per_second": 44.255, | |
| "eval_steps_per_second": 2.012, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.47744765877723694, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0491, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.4440220296382904, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.046, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.44091033935546875, | |
| "learning_rate": 2.1833997096818898e-05, | |
| "loss": 0.0442, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.04639214277267456, | |
| "eval_runtime": 0.5026, | |
| "eval_samples_per_second": 43.771, | |
| "eval_steps_per_second": 1.99, | |
| "step": 39 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 52, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.180338678543155e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |