<!-- 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. -->
hbertv2-Massive-intent-48-emb-comp-gelu
This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new_emb_compress_48_gelu on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 1.0025
- Accuracy: 0.8421
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 |
|---|---|---|---|---|
| 1.9202 | 1.0 | 180 | 1.0767 | 0.7068 |
| 0.9104 | 2.0 | 360 | 0.9209 | 0.7482 |
| 0.6425 | 3.0 | 540 | 0.8343 | 0.7821 |
| 0.4854 | 4.0 | 720 | 0.8159 | 0.7954 |
| 0.3682 | 5.0 | 900 | 0.8154 | 0.8077 |
| 0.272 | 6.0 | 1080 | 0.8417 | 0.7993 |
| 0.204 | 7.0 | 1260 | 0.7931 | 0.8155 |
| 0.1363 | 8.0 | 1440 | 0.8740 | 0.8195 |
| 0.1016 | 9.0 | 1620 | 0.8993 | 0.8205 |
| 0.0689 | 10.0 | 1800 | 0.9309 | 0.8210 |
| 0.0478 | 11.0 | 1980 | 0.9877 | 0.8318 |
| 0.0254 | 12.0 | 2160 | 1.0041 | 0.8293 |
| 0.0133 | 13.0 | 2340 | 0.9982 | 0.8396 |
| 0.0068 | 14.0 | 2520 | 1.0049 | 0.8406 |
| 0.005 | 15.0 | 2700 | 1.0025 | 0.8421 |
Framework versions
- Transformers 4.31.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.1
- Tokenizers 0.13.3
