<!-- 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. -->
HBERTv1_48_L10_H64_A2_massive
This model is a fine-tuned version of gokuls/HBERTv1_48_L10_H64_A2 on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 1.9811
- Accuracy: 0.4501
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 |
|---|---|---|---|---|
| 3.9732 | 1.0 | 180 | 3.7522 | 0.1323 |
| 3.5635 | 2.0 | 360 | 3.3729 | 0.0974 |
| 3.2789 | 3.0 | 540 | 3.1242 | 0.2292 |
| 3.0277 | 4.0 | 720 | 2.8648 | 0.2814 |
| 2.7797 | 5.0 | 900 | 2.6533 | 0.2799 |
| 2.5868 | 6.0 | 1080 | 2.4893 | 0.3114 |
| 2.4433 | 7.0 | 1260 | 2.3541 | 0.3433 |
| 2.3172 | 8.0 | 1440 | 2.2641 | 0.3507 |
| 2.219 | 9.0 | 1620 | 2.1869 | 0.3827 |
| 2.1307 | 10.0 | 1800 | 2.1181 | 0.3989 |
| 2.0674 | 11.0 | 1980 | 2.0644 | 0.4097 |
| 2.0109 | 12.0 | 2160 | 2.0236 | 0.4353 |
| 1.9699 | 13.0 | 2340 | 1.9951 | 0.4383 |
| 1.9395 | 14.0 | 2520 | 1.9811 | 0.4501 |
| 1.9234 | 15.0 | 2700 | 1.9706 | 0.4461 |
Framework versions
- Transformers 4.34.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.14.0
