<!-- 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. -->
bert-base-emotion_24
This model is a fine-tuned version of gokuls/bert_base_24 on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.5612
- Accuracy: 0.8611
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 |
|---|---|---|---|---|
| 0.736 | 1.0 | 250 | 0.4242 | 0.8565 |
| 0.3013 | 2.0 | 500 | 0.3314 | 0.8845 |
| 0.2014 | 3.0 | 750 | 0.3442 | 0.8905 |
| 0.1392 | 4.0 | 1000 | 0.3276 | 0.8915 |
| 0.1072 | 5.0 | 1250 | 0.3833 | 0.89 |
| 0.0783 | 6.0 | 1500 | 0.4205 | 0.8895 |
| 0.0559 | 7.0 | 1750 | 0.5287 | 0.8865 |
| 0.0378 | 8.0 | 2000 | 0.5459 | 0.8865 |
| 0.027 | 9.0 | 2250 | 0.5612 | 0.8925 |
| 0.02 | 10.0 | 2500 | 0.5601 | 0.8915 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3
