| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 18484, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 1.945899156026834e-05, | |
| "loss": 2.5051, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 1.8917983120536682e-05, | |
| "loss": 2.3264, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 1.837697468080502e-05, | |
| "loss": 2.2657, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 1.7835966241073363e-05, | |
| "loss": 2.2257, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 1.7294957801341702e-05, | |
| "loss": 2.1931, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6753949361610044e-05, | |
| "loss": 2.1735, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.6212940921878382e-05, | |
| "loss": 2.1473, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 1.5671932482146724e-05, | |
| "loss": 2.1401, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5130924042415063e-05, | |
| "loss": 2.1171, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.8773, | |
| "eval_gen_len": 16.4704, | |
| "eval_loss": 2.013514280319214, | |
| "eval_precision": 0.8836, | |
| "eval_recall": 0.8717, | |
| "eval_rouge1": 0.3138, | |
| "eval_rouge2": 0.1556, | |
| "eval_rougeL": 0.2853, | |
| "eval_rougeLsum": 0.2853, | |
| "eval_runtime": 1599.8646, | |
| "eval_samples_per_second": 6.251, | |
| "eval_steps_per_second": 0.391, | |
| "step": 4621 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 1.4589915602683402e-05, | |
| "loss": 2.0523, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 1.4048907162951744e-05, | |
| "loss": 2.0306, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.3507898723220082e-05, | |
| "loss": 2.0262, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 1.2966890283488423e-05, | |
| "loss": 2.0169, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 1.2425881843756765e-05, | |
| "loss": 2.0101, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 1.1884873404025104e-05, | |
| "loss": 1.9955, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 1.1343864964293442e-05, | |
| "loss": 1.9943, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 1.0802856524561784e-05, | |
| "loss": 1.9863, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 1.0261848084830125e-05, | |
| "loss": 1.9804, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.8775, | |
| "eval_gen_len": 16.2207, | |
| "eval_loss": 1.9440101385116577, | |
| "eval_precision": 0.8831, | |
| "eval_recall": 0.8725, | |
| "eval_rouge1": 0.3147, | |
| "eval_rouge2": 0.1581, | |
| "eval_rougeL": 0.2864, | |
| "eval_rougeLsum": 0.2866, | |
| "eval_runtime": 1574.2726, | |
| "eval_samples_per_second": 6.352, | |
| "eval_steps_per_second": 0.397, | |
| "step": 9242 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 9.720839645098465e-06, | |
| "loss": 1.9455, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 9.179831205366805e-06, | |
| "loss": 1.9171, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 8.638822765635144e-06, | |
| "loss": 1.9211, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 8.097814325903484e-06, | |
| "loss": 1.9122, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 7.5568058861718255e-06, | |
| "loss": 1.9126, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 7.015797446440165e-06, | |
| "loss": 1.9065, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 6.474789006708505e-06, | |
| "loss": 1.8955, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 5.933780566976846e-06, | |
| "loss": 1.899, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 5.392772127245185e-06, | |
| "loss": 1.8971, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.8792, | |
| "eval_gen_len": 15.4676, | |
| "eval_loss": 1.9157460927963257, | |
| "eval_precision": 0.8857, | |
| "eval_recall": 0.8733, | |
| "eval_rouge1": 0.3209, | |
| "eval_rouge2": 0.1638, | |
| "eval_rougeL": 0.2925, | |
| "eval_rougeLsum": 0.2926, | |
| "eval_runtime": 1567.1343, | |
| "eval_samples_per_second": 6.381, | |
| "eval_steps_per_second": 0.399, | |
| "step": 13863 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 4.8517636875135255e-06, | |
| "loss": 1.8862, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 4.310755247781866e-06, | |
| "loss": 1.8611, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 3.769746808050206e-06, | |
| "loss": 1.8474, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 3.2287383683185457e-06, | |
| "loss": 1.8523, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 2.687729928586886e-06, | |
| "loss": 1.8462, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 2.1467214888552264e-06, | |
| "loss": 1.8455, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 1.6057130491235664e-06, | |
| "loss": 1.8456, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 1.0647046093919065e-06, | |
| "loss": 1.8425, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 5.236961696602468e-07, | |
| "loss": 1.8449, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_f1": 0.8795, | |
| "eval_gen_len": 15.7946, | |
| "eval_loss": 1.9021437168121338, | |
| "eval_precision": 0.8858, | |
| "eval_recall": 0.8739, | |
| "eval_rouge1": 0.3236, | |
| "eval_rouge2": 0.1651, | |
| "eval_rougeL": 0.2953, | |
| "eval_rougeLsum": 0.2953, | |
| "eval_runtime": 1613.4446, | |
| "eval_samples_per_second": 6.198, | |
| "eval_steps_per_second": 0.387, | |
| "step": 18484 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 18484, | |
| "total_flos": 3.845275732313899e+18, | |
| "train_loss": 1.997600625588558, | |
| "train_runtime": 109033.1022, | |
| "train_samples_per_second": 16.274, | |
| "train_steps_per_second": 0.17 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 18484, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 3.845275732313899e+18, | |
| "train_batch_size": 24, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |