| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 1452, | |
| "global_step": 10164, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2721, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0539, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0079, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9686, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.668149394347241, | |
| "eval_loss": 1.5662202835083008, | |
| "eval_runtime": 137.7572, | |
| "eval_samples_per_second": 36.296, | |
| "eval_steps_per_second": 2.272, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 9.04, | |
| "eval_f1": 13.087046780017364, | |
| "eval_qa_bleu": 5.297766450359973, | |
| "eval_qa_exact_match": 0.085, | |
| "eval_recite_bleu": 14.716037077738477, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9076, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5211, | |
| "step": 1746 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5039, | |
| "step": 2037 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4834, | |
| "step": 2328 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4604, | |
| "step": 2619 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6914514131897712, | |
| "eval_loss": 1.206884741783142, | |
| "eval_runtime": 139.7305, | |
| "eval_samples_per_second": 35.783, | |
| "eval_steps_per_second": 2.24, | |
| "step": 2904 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 11.64, | |
| "eval_f1": 16.917593370997537, | |
| "eval_qa_bleu": 7.627957176996819, | |
| "eval_qa_exact_match": 0.1106, | |
| "eval_recite_bleu": 18.356743056908567, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 2904 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4368, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0705, | |
| "step": 3201 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0685, | |
| "step": 3492 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0554, | |
| "step": 3783 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0522, | |
| "step": 4074 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7148896366083446, | |
| "eval_loss": 0.9045500159263611, | |
| "eval_runtime": 141.6455, | |
| "eval_samples_per_second": 35.299, | |
| "eval_steps_per_second": 2.21, | |
| "step": 4356 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 12.84, | |
| "eval_f1": 18.555356874635468, | |
| "eval_qa_bleu": 8.395879921702353, | |
| "eval_qa_exact_match": 0.1188, | |
| "eval_recite_bleu": 20.824221134285214, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 4356 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0375, | |
| "step": 4365 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7373, | |
| "step": 4656 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7497, | |
| "step": 4947 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7479, | |
| "step": 5238 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7455, | |
| "step": 5529 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7340508748317631, | |
| "eval_loss": 0.6843858957290649, | |
| "eval_runtime": 143.5169, | |
| "eval_samples_per_second": 34.839, | |
| "eval_steps_per_second": 2.181, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 13.44, | |
| "eval_f1": 19.418847592695993, | |
| "eval_qa_bleu": 8.731752840499793, | |
| "eval_qa_exact_match": 0.1256, | |
| "eval_recite_bleu": 24.11683390270685, | |
| "eval_recite_exact_match": 0.0002, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7226, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5117, | |
| "step": 6111 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5208, | |
| "step": 6402 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5271, | |
| "step": 6693 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5307, | |
| "step": 6984 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7474745625841185, | |
| "eval_loss": 0.5420497059822083, | |
| "eval_runtime": 144.0475, | |
| "eval_samples_per_second": 34.711, | |
| "eval_steps_per_second": 2.173, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 15.08, | |
| "eval_f1": 21.4491845879152, | |
| "eval_qa_bleu": 10.89471390420015, | |
| "eval_qa_exact_match": 0.1434, | |
| "eval_recite_bleu": 27.62075386163555, | |
| "eval_recite_exact_match": 0.0014, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5165, | |
| "step": 7275 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3624, | |
| "step": 7566 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3748, | |
| "step": 7857 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.383, | |
| "step": 8148 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3796, | |
| "step": 8439 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7560010767160161, | |
| "eval_loss": 0.4608670473098755, | |
| "eval_runtime": 144.2314, | |
| "eval_samples_per_second": 34.667, | |
| "eval_steps_per_second": 2.17, | |
| "step": 8712 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 16.22, | |
| "eval_f1": 22.5212141898349, | |
| "eval_qa_bleu": 11.517856911949949, | |
| "eval_qa_exact_match": 0.1524, | |
| "eval_recite_bleu": 32.1802083615024, | |
| "eval_recite_exact_match": 0.0044, | |
| "step": 8712 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3803, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 6.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.27, | |
| "step": 9021 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "learning_rate": 3e-05, | |
| "loss": 0.28, | |
| "step": 9312 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2889, | |
| "step": 9603 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2912, | |
| "step": 9894 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7603886944818304, | |
| "eval_loss": 0.43113213777542114, | |
| "eval_runtime": 144.1381, | |
| "eval_samples_per_second": 34.689, | |
| "eval_steps_per_second": 2.172, | |
| "step": 10164 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 17.64, | |
| "eval_f1": 24.23915265275915, | |
| "eval_qa_bleu": 12.404229060978194, | |
| "eval_qa_exact_match": 0.1648, | |
| "eval_recite_bleu": 37.23348279932964, | |
| "eval_recite_exact_match": 0.0102, | |
| "step": 10164 | |
| } | |
| ], | |
| "logging_steps": 291, | |
| "max_steps": 29040, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 4.641681443905536e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |