<!-- 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_H256_A4_massive
This model is a fine-tuned version of gokuls/HBERTv1_48_L10_H256_A4 on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.8290
- Accuracy: 0.8283
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 |
|---|---|---|---|---|
| 3.5874 | 1.0 | 180 | 2.8714 | 0.2863 |
| 2.3912 | 2.0 | 360 | 1.8365 | 0.5529 |
| 1.6209 | 3.0 | 540 | 1.3243 | 0.6704 |
| 1.1916 | 4.0 | 720 | 1.0857 | 0.7304 |
| 0.9439 | 5.0 | 900 | 0.9409 | 0.7555 |
| 0.7731 | 6.0 | 1080 | 0.9084 | 0.7654 |
| 0.6539 | 7.0 | 1260 | 0.8745 | 0.7831 |
| 0.555 | 8.0 | 1440 | 0.8218 | 0.8072 |
| 0.4776 | 9.0 | 1620 | 0.8185 | 0.8136 |
| 0.4129 | 10.0 | 1800 | 0.8245 | 0.8111 |
| 0.3706 | 11.0 | 1980 | 0.8291 | 0.8160 |
| 0.3217 | 12.0 | 2160 | 0.8129 | 0.8229 |
| 0.2874 | 13.0 | 2340 | 0.8226 | 0.8210 |
| 0.2541 | 14.0 | 2520 | 0.8290 | 0.8283 |
| 0.2383 | 15.0 | 2700 | 0.8303 | 0.8273 |
Framework versions
- Transformers 4.34.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.14.0
