metadata
license: apache-2.0
base_model: google-t5/t5-small
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: t5_small_patent
results: []
t5_small_patent
This model is a fine-tuned version of google-t5/t5-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.9833
- Accuracy: 0.657
- F1 Macro: 0.5822
- F1 Micro: 0.657
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0005
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 32
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Macro | F1 Micro |
|---|---|---|---|---|---|---|
| 1.4814 | 0.06 | 50 | 1.4704 | 0.492 | 0.3414 | 0.492 |
| 1.3003 | 0.13 | 100 | 1.2652 | 0.5512 | 0.3876 | 0.5512 |
| 1.2291 | 0.19 | 150 | 1.2304 | 0.563 | 0.4000 | 0.563 |
| 1.142 | 0.26 | 200 | 1.1644 | 0.5894 | 0.4586 | 0.5894 |
| 1.0581 | 0.32 | 250 | 1.1396 | 0.603 | 0.4563 | 0.603 |
| 1.2415 | 0.38 | 300 | 1.1215 | 0.613 | 0.4937 | 0.613 |
| 1.1336 | 0.45 | 350 | 1.1042 | 0.6172 | 0.5292 | 0.6172 |
| 1.045 | 0.51 | 400 | 1.0924 | 0.624 | 0.5271 | 0.624 |
| 1.1204 | 0.58 | 450 | 1.0897 | 0.6184 | 0.5146 | 0.6184 |
| 1.0691 | 0.64 | 500 | 1.0827 | 0.6236 | 0.5169 | 0.6236 |
| 0.9782 | 0.7 | 550 | 1.0664 | 0.6258 | 0.5303 | 0.6258 |
| 1.081 | 0.77 | 600 | 1.0548 | 0.638 | 0.5581 | 0.638 |
| 1.1033 | 0.83 | 650 | 1.0300 | 0.6398 | 0.5593 | 0.6398 |
| 1.0946 | 0.9 | 700 | 1.0620 | 0.632 | 0.5545 | 0.632 |
| 1.0261 | 0.96 | 750 | 1.0328 | 0.6422 | 0.5648 | 0.6422 |
| 0.9153 | 1.02 | 800 | 1.0378 | 0.6438 | 0.5706 | 0.6438 |
| 0.9678 | 1.09 | 850 | 1.0520 | 0.6402 | 0.5483 | 0.6402 |
| 0.9619 | 1.15 | 900 | 1.0483 | 0.6408 | 0.5593 | 0.6408 |
| 0.9972 | 1.21 | 950 | 1.0255 | 0.6496 | 0.5685 | 0.6496 |
| 1.027 | 1.28 | 1000 | 1.0296 | 0.645 | 0.5742 | 0.645 |
| 0.8248 | 1.34 | 1050 | 1.0331 | 0.655 | 0.5812 | 0.655 |
| 0.9405 | 1.41 | 1100 | 1.0208 | 0.6502 | 0.5719 | 0.6502 |
| 0.9735 | 1.47 | 1150 | 1.0389 | 0.6388 | 0.5744 | 0.6388 |
| 0.9566 | 1.53 | 1200 | 0.9963 | 0.658 | 0.5750 | 0.658 |
| 0.9423 | 1.6 | 1250 | 0.9966 | 0.6496 | 0.5832 | 0.6496 |
| 0.9248 | 1.66 | 1300 | 0.9953 | 0.6558 | 0.5857 | 0.6558 |
| 1.008 | 1.73 | 1350 | 0.9940 | 0.6588 | 0.5809 | 0.6588 |
| 0.9098 | 1.79 | 1400 | 0.9833 | 0.657 | 0.5822 | 0.657 |
| 0.8679 | 1.85 | 1450 | 0.9842 | 0.6644 | 0.5899 | 0.6644 |
| 1.1342 | 1.92 | 1500 | 0.9933 | 0.6526 | 0.5762 | 0.6526 |
| 0.9157 | 1.98 | 1550 | 0.9869 | 0.6626 | 0.5924 | 0.6626 |
| 0.8084 | 2.05 | 1600 | 0.9909 | 0.6654 | 0.5893 | 0.6654 |
| 0.7373 | 2.11 | 1650 | 0.9894 | 0.6622 | 0.5965 | 0.6622 |
| 0.9081 | 2.17 | 1700 | 0.9997 | 0.6614 | 0.5880 | 0.6614 |
| 0.8064 | 2.24 | 1750 | 0.9998 | 0.659 | 0.5919 | 0.659 |
| 0.8519 | 2.3 | 1800 | 1.0031 | 0.6584 | 0.5880 | 0.6584 |
| 0.8711 | 2.37 | 1850 | 0.9975 | 0.6666 | 0.5981 | 0.6666 |
| 0.7617 | 2.43 | 1900 | 1.0144 | 0.6584 | 0.5849 | 0.6584 |
| 0.717 | 2.49 | 1950 | 1.0102 | 0.6622 | 0.5903 | 0.6622 |
| 0.857 | 2.56 | 2000 | 1.0059 | 0.6622 | 0.5923 | 0.6622 |
| 0.8623 | 2.62 | 2050 | 1.0025 | 0.664 | 0.5971 | 0.664 |
| 0.782 | 2.69 | 2100 | 1.0013 | 0.6644 | 0.5985 | 0.6644 |
| 0.8018 | 2.75 | 2150 | 1.0044 | 0.6652 | 0.5985 | 0.6652 |
| 0.7901 | 2.81 | 2200 | 0.9987 | 0.6678 | 0.6030 | 0.6678 |
| 0.8835 | 2.88 | 2250 | 1.0015 | 0.6644 | 0.5986 | 0.6644 |
| 0.8679 | 2.94 | 2300 | 0.9994 | 0.6636 | 0.5961 | 0.6636 |
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2