<!-- 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-Massive-intent-48-emb-comp-gelu
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48_gelu on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.9566
- Accuracy: 0.8028
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 |
|---|---|---|---|---|
| 3.4622 | 1.0 | 180 | 3.0181 | 0.2169 |
| 2.7526 | 2.0 | 360 | 2.4760 | 0.3168 |
| 2.188 | 3.0 | 540 | 1.9627 | 0.4368 |
| 1.7069 | 4.0 | 720 | 1.5603 | 0.5568 |
| 1.3045 | 5.0 | 900 | 1.3354 | 0.6345 |
| 1.0621 | 6.0 | 1080 | 1.1726 | 0.6862 |
| 0.8745 | 7.0 | 1260 | 1.0703 | 0.7226 |
| 0.7286 | 8.0 | 1440 | 0.9905 | 0.7516 |
| 0.6005 | 9.0 | 1620 | 0.9881 | 0.7644 |
| 0.5021 | 10.0 | 1800 | 0.9661 | 0.7732 |
| 0.4208 | 11.0 | 1980 | 0.9621 | 0.7787 |
| 0.3524 | 12.0 | 2160 | 0.9480 | 0.7939 |
| 0.282 | 13.0 | 2340 | 0.9614 | 0.7924 |
| 0.2327 | 14.0 | 2520 | 0.9525 | 0.7969 |
| 0.1912 | 15.0 | 2700 | 0.9566 | 0.8028 |
Framework versions
- Transformers 4.31.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.1
- Tokenizers 0.13.3
