File size: 1,147 Bytes
4b916c6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 |
Writing logs to ./outputs/2024-04-30-11-40-48-251003/train_log.txt. Wrote original training args to ./outputs/2024-04-30-11-40-48-251003/training_args.json. ***** Running training ***** Num examples = 4800 Num epochs = 3 Num clean epochs = 3 Instantaneous batch size per device = 8 Total train batch size (w. parallel, distributed & accumulation) = 8 Gradient accumulation steps = 1 Total optimization steps = 1800 ========================================================== Epoch 1 Running clean epoch 1/3 Train accuracy: 82.60% Eval accuracy: 89.90% Best score found. Saved model to ./outputs/2024-04-30-11-40-48-251003/best_model/ ========================================================== Epoch 2 Running clean epoch 2/3 Train accuracy: 94.21% Eval accuracy: 93.37% Best score found. Saved model to ./outputs/2024-04-30-11-40-48-251003/best_model/ ========================================================== Epoch 3 Running clean epoch 3/3 Train accuracy: 97.81% Eval accuracy: 93.70% Best score found. Saved model to ./outputs/2024-04-30-11-40-48-251003/best_model/ Wrote README to ./outputs/2024-04-30-11-40-48-251003/README.md. |