hbertv1-tiny-wt-48-emotion
This model is a fine-tuned version of gokuls/model_v1_complete_training_wt_init_48_tiny on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.2695
- Accuracy: 0.8985
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 |
---|---|---|---|---|
1.4321 | 1.0 | 250 | 1.0203 | 0.6475 |
0.8329 | 2.0 | 500 | 0.5954 | 0.814 |
0.5347 | 3.0 | 750 | 0.4146 | 0.8645 |
0.398 | 4.0 | 1000 | 0.3496 | 0.8805 |
0.3418 | 5.0 | 1250 | 0.3091 | 0.889 |
0.2932 | 6.0 | 1500 | 0.2864 | 0.8945 |
0.2646 | 7.0 | 1750 | 0.2782 | 0.8965 |
0.2532 | 8.0 | 2000 | 0.2695 | 0.8985 |
0.2342 | 9.0 | 2250 | 0.2632 | 0.898 |
0.225 | 10.0 | 2500 | 0.2617 | 0.897 |
Framework versions
- Transformers 4.31.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.1
- Tokenizers 0.13.3
- Downloads last month
- 7