minilm_finetuned_emotions
This model is a fine-tuned version of microsoft/MiniLM-L12-H384-uncased on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.4118
- F1: 0.9103
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: 2e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|---|---|---|---|---|
| 1.3758 | 1.0 | 250 | 1.0211 | 0.5910 |
| 0.885 | 2.0 | 500 | 0.7133 | 0.7977 |
| 0.6248 | 3.0 | 750 | 0.5321 | 0.8840 |
| 0.4874 | 4.0 | 1000 | 0.4416 | 0.9013 |
| 0.4193 | 5.0 | 1250 | 0.4118 | 0.9103 |
Framework versions
- Transformers 4.38.1
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2