albert-twitter-hate
This model is a fine-tuned version of albert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5210
- Accuracy: 0.741
- Precision: 0.6591
- Recall: 0.8150
- F1: 0.7288
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 282 | 0.5499 | 0.721 | 0.6418 | 0.7845 | 0.7060 |
| 0.4566 | 2.0 | 564 | 0.5210 | 0.741 | 0.6591 | 0.8150 | 0.7288 |
| 0.4566 | 3.0 | 846 | 0.5642 | 0.749 | 0.6833 | 0.7681 | 0.7233 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.7.1+cu126
- Datasets 2.19.1
- Tokenizers 0.19.1