<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
bert_uncased_L-10_H-768_A-12_massive
This model is a fine-tuned version of google/bert_uncased_L-10_H-768_A-12 on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.6424
- Accuracy: 0.8908
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 33
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.8931 | 1.0 | 180 | 0.7740 | 0.8342 |
| 0.5861 | 2.0 | 360 | 0.5441 | 0.8667 |
| 0.312 | 3.0 | 540 | 0.4858 | 0.8805 |
| 0.1828 | 4.0 | 720 | 0.4977 | 0.8849 |
| 0.1183 | 5.0 | 900 | 0.5322 | 0.8824 |
| 0.077 | 6.0 | 1080 | 0.5617 | 0.8780 |
| 0.0491 | 7.0 | 1260 | 0.5901 | 0.8810 |
| 0.0328 | 8.0 | 1440 | 0.6181 | 0.8775 |
| 0.0216 | 9.0 | 1620 | 0.6174 | 0.8869 |
| 0.0159 | 10.0 | 1800 | 0.6278 | 0.8869 |
| 0.0116 | 11.0 | 1980 | 0.6228 | 0.8888 |
| 0.0083 | 12.0 | 2160 | 0.6352 | 0.8893 |
| 0.0058 | 13.0 | 2340 | 0.6443 | 0.8893 |
| 0.0056 | 14.0 | 2520 | 0.6424 | 0.8908 |
| 0.0047 | 15.0 | 2700 | 0.6462 | 0.8893 |
Framework versions
- Transformers 4.34.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.14.1
