| { | |
| "best_metric": 1.5313832759857178, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_bert/bert_base_uncased_scotus/checkpoint-200", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 237, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 7.085313320159912, | |
| "learning_rate": 1.9156118143459917e-05, | |
| "loss": 2.429, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 4.804141521453857, | |
| "learning_rate": 1.8312236286919833e-05, | |
| "loss": 2.2691, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 4.930667877197266, | |
| "learning_rate": 1.746835443037975e-05, | |
| "loss": 2.1744, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.996751308441162, | |
| "learning_rate": 1.662447257383966e-05, | |
| "loss": 2.0977, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.7142813205718994, | |
| "learning_rate": 1.578059071729958e-05, | |
| "loss": 2.0086, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_accuracy": 0.38571428571428573, | |
| "eval_f1_macro": 0.09223731954881484, | |
| "eval_f1_micro": 0.38571428571428573, | |
| "eval_loss": 2.006516218185425, | |
| "eval_runtime": 1.2668, | |
| "eval_samples_per_second": 1105.113, | |
| "eval_steps_per_second": 17.366, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.9603490829467773, | |
| "learning_rate": 1.4936708860759495e-05, | |
| "loss": 1.9617, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.7058732509613037, | |
| "learning_rate": 1.4092827004219412e-05, | |
| "loss": 1.8188, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 3.41971755027771, | |
| "learning_rate": 1.3248945147679326e-05, | |
| "loss": 1.8824, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 4.253640651702881, | |
| "learning_rate": 1.240506329113924e-05, | |
| "loss": 1.6726, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 4.203283786773682, | |
| "learning_rate": 1.1561181434599158e-05, | |
| "loss": 1.5947, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "eval_accuracy": 0.475, | |
| "eval_f1_macro": 0.18707990555027815, | |
| "eval_f1_micro": 0.475, | |
| "eval_loss": 1.6981335878372192, | |
| "eval_runtime": 1.3258, | |
| "eval_samples_per_second": 1055.997, | |
| "eval_steps_per_second": 16.594, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 5.603295803070068, | |
| "learning_rate": 1.0717299578059072e-05, | |
| "loss": 1.6728, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 5.53965425491333, | |
| "learning_rate": 9.87341772151899e-06, | |
| "loss": 1.6157, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 3.6127514839172363, | |
| "learning_rate": 9.029535864978903e-06, | |
| "loss": 1.5524, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 4.825035572052002, | |
| "learning_rate": 8.18565400843882e-06, | |
| "loss": 1.4613, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 4.025143146514893, | |
| "learning_rate": 7.341772151898735e-06, | |
| "loss": 1.5835, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_accuracy": 0.5378571428571428, | |
| "eval_f1_macro": 0.2780332747870082, | |
| "eval_f1_micro": 0.5378571428571428, | |
| "eval_loss": 1.5822380781173706, | |
| "eval_runtime": 1.3354, | |
| "eval_samples_per_second": 1048.397, | |
| "eval_steps_per_second": 16.475, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 4.761880397796631, | |
| "learning_rate": 6.49789029535865e-06, | |
| "loss": 1.4868, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 3.8378217220306396, | |
| "learning_rate": 5.654008438818566e-06, | |
| "loss": 1.4197, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 3.8176848888397217, | |
| "learning_rate": 4.8101265822784815e-06, | |
| "loss": 1.4556, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 4.766996383666992, | |
| "learning_rate": 3.9662447257383965e-06, | |
| "loss": 1.3993, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 3.517928123474121, | |
| "learning_rate": 3.1223628691983127e-06, | |
| "loss": 1.3943, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_accuracy": 0.5514285714285714, | |
| "eval_f1_macro": 0.2849023186790848, | |
| "eval_f1_micro": 0.5514285714285714, | |
| "eval_loss": 1.5313832759857178, | |
| "eval_runtime": 1.3455, | |
| "eval_samples_per_second": 1040.508, | |
| "eval_steps_per_second": 16.351, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 4.90807580947876, | |
| "learning_rate": 2.278481012658228e-06, | |
| "loss": 1.3998, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 3.307119607925415, | |
| "learning_rate": 1.4345991561181436e-06, | |
| "loss": 1.3483, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 4.938708782196045, | |
| "learning_rate": 5.907172995780591e-07, | |
| "loss": 1.4034, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 237, | |
| "total_flos": 997815706714112.0, | |
| "train_loss": 1.6904215390169168, | |
| "train_runtime": 53.9817, | |
| "train_samples_per_second": 277.872, | |
| "train_steps_per_second": 4.39 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 237, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 997815706714112.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |