hbertv1-mini-wt-frz-48-Massive-intent-emb-comp
This model is a fine-tuned version of gokuls/model_v1_complete_training_wt_init_48_mini_emb_comp_frz on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.7199
- Accuracy: 0.8288
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 |
|---|---|---|---|---|
| 2.8181 | 1.0 | 180 | 1.7693 | 0.6085 |
| 1.3856 | 2.0 | 360 | 1.0888 | 0.7211 |
| 0.9216 | 3.0 | 540 | 0.8674 | 0.7782 |
| 0.6977 | 4.0 | 720 | 0.7678 | 0.8028 |
| 0.5417 | 5.0 | 900 | 0.7335 | 0.8106 |
| 0.4402 | 6.0 | 1080 | 0.7076 | 0.8190 |
| 0.3562 | 7.0 | 1260 | 0.6918 | 0.8244 |
| 0.2937 | 8.0 | 1440 | 0.6998 | 0.8210 |
| 0.2331 | 9.0 | 1620 | 0.7244 | 0.8205 |
| 0.1925 | 10.0 | 1800 | 0.7199 | 0.8288 |
| 0.1589 | 11.0 | 1980 | 0.7338 | 0.8278 |
| 0.1321 | 12.0 | 2160 | 0.7561 | 0.8259 |
| 0.1093 | 13.0 | 2340 | 0.7498 | 0.8278 |
| 0.0937 | 14.0 | 2520 | 0.7579 | 0.8278 |
| 0.0852 | 15.0 | 2700 | 0.7542 | 0.8288 |
Framework versions
- Transformers 4.31.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.1
- Tokenizers 0.13.3