hbertv1-emotion_48_emb_compress
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48 on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.4493
- Accuracy: 0.874
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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.4218 | 1.0 | 250 | 1.1098 | 0.5885 |
| 0.9116 | 2.0 | 500 | 0.7865 | 0.743 |
| 0.5915 | 3.0 | 750 | 0.6149 | 0.805 |
| 0.4435 | 4.0 | 1000 | 0.4932 | 0.841 |
| 0.3626 | 5.0 | 1250 | 0.4634 | 0.855 |
| 0.3031 | 6.0 | 1500 | 0.4514 | 0.8545 |
| 0.2457 | 7.0 | 1750 | 0.4395 | 0.865 |
| 0.2039 | 8.0 | 2000 | 0.4368 | 0.861 |
| 0.1664 | 9.0 | 2250 | 0.4276 | 0.871 |
| 0.1402 | 10.0 | 2500 | 0.4493 | 0.874 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3