| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.997333333333334, | |
| "eval_steps": 500, | |
| "global_step": 1687, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3988, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6660784313725491, | |
| "eval_loss": 1.2107256650924683, | |
| "eval_runtime": 7.728, | |
| "eval_samples_per_second": 64.7, | |
| "eval_steps_per_second": 8.152, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 19.4, | |
| "eval_f1": 27.911959489312448, | |
| "eval_qa_bleu": 2.842200543476061, | |
| "eval_qa_exact_match": 0.152, | |
| "eval_recite_bleu": 12.149078895446083, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 5e-05, | |
| "loss": 1.238, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1977, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6676078431372549, | |
| "eval_loss": 1.1987448930740356, | |
| "eval_runtime": 7.4628, | |
| "eval_samples_per_second": 66.999, | |
| "eval_steps_per_second": 8.442, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 21.4, | |
| "eval_f1": 27.979969946667268, | |
| "eval_qa_bleu": 3.4273838915653356, | |
| "eval_qa_exact_match": 0.158, | |
| "eval_recite_bleu": 11.474877690710539, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1628, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1458, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6675764705882353, | |
| "eval_loss": 1.1964411735534668, | |
| "eval_runtime": 6.6677, | |
| "eval_samples_per_second": 74.988, | |
| "eval_steps_per_second": 9.448, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 23.4, | |
| "eval_f1": 31.525761083856903, | |
| "eval_qa_bleu": 2.0606924929592334, | |
| "eval_qa_exact_match": 0.17, | |
| "eval_recite_bleu": 12.027903845545522, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1179, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0707, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6665960784313726, | |
| "eval_loss": 1.208784580230713, | |
| "eval_runtime": 6.6546, | |
| "eval_samples_per_second": 75.136, | |
| "eval_steps_per_second": 9.467, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 23.4, | |
| "eval_f1": 33.18893657873498, | |
| "eval_qa_bleu": 3.8183981024520506, | |
| "eval_qa_exact_match": 0.176, | |
| "eval_recite_bleu": 14.071857337313705, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0179, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0066, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6651764705882353, | |
| "eval_loss": 1.2342489957809448, | |
| "eval_runtime": 6.699, | |
| "eval_samples_per_second": 74.638, | |
| "eval_steps_per_second": 9.404, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 24.0, | |
| "eval_f1": 32.88692081962977, | |
| "eval_qa_bleu": 3.6747877275980905, | |
| "eval_qa_exact_match": 0.174, | |
| "eval_recite_bleu": 14.36158051041567, | |
| "eval_recite_exact_match": 0.006, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9595, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9353, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6628392156862745, | |
| "eval_loss": 1.2645816802978516, | |
| "eval_runtime": 7.4539, | |
| "eval_samples_per_second": 67.079, | |
| "eval_steps_per_second": 8.452, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 23.4, | |
| "eval_f1": 33.50380952380953, | |
| "eval_qa_bleu": 13.968612067397348, | |
| "eval_qa_exact_match": 0.168, | |
| "eval_recite_bleu": 15.026821124315997, | |
| "eval_recite_exact_match": 0.012, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8774, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8629, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6617333333333333, | |
| "eval_loss": 1.2990769147872925, | |
| "eval_runtime": 6.6963, | |
| "eval_samples_per_second": 74.668, | |
| "eval_steps_per_second": 9.408, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 22.0, | |
| "eval_f1": 31.8904761904762, | |
| "eval_qa_bleu": 15.632357578213835, | |
| "eval_qa_exact_match": 0.166, | |
| "eval_recite_bleu": 14.318212267740977, | |
| "eval_recite_exact_match": 0.016, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7933, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.659607843137255, | |
| "eval_loss": 1.3466852903366089, | |
| "eval_runtime": 6.6563, | |
| "eval_samples_per_second": 75.117, | |
| "eval_steps_per_second": 9.465, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 22.6, | |
| "eval_f1": 31.16493506493508, | |
| "eval_qa_bleu": 21.45561748665761, | |
| "eval_qa_exact_match": 0.17, | |
| "eval_recite_bleu": 14.23612921374107, | |
| "eval_recite_exact_match": 0.012, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7185, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6574980392156863, | |
| "eval_loss": 1.391683578491211, | |
| "eval_runtime": 7.7056, | |
| "eval_samples_per_second": 64.888, | |
| "eval_steps_per_second": 8.176, | |
| "step": 1687 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 21.8, | |
| "eval_f1": 31.150317460317464, | |
| "eval_qa_bleu": 18.411626756160196, | |
| "eval_qa_exact_match": 0.17, | |
| "eval_recite_bleu": 15.144735177405446, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 1687 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 9350, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 3.910718003913687e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |