visobert-hatespeech-multitask-vithsd
This model is a fine-tuned version of uitnlp/visobert on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4785
- Hamming Loss: 0.1092
- Subset Accuracy: 0.024
- Per Label Accuracy: 0.8908
- F1 Macro: 0.1908
- F1 Weighted: 0.6536
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Hamming Loss | Subset Accuracy | Per Label Accuracy | F1 Macro | F1 Weighted |
|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 32 | 0.4785 | 0.1092 | 0.024 | 0.8908 | 0.1908 | 0.6536 |
| 0.5533 | 2.0 | 64 | 0.3700 | 0.1048 | 0.0 | 0.8952 | 0.1725 | 0.6231 |
| 0.5533 | 3.0 | 96 | 0.3305 | 0.104 | 0.0 | 0.8960 | 0.1837 | 0.6493 |
| 0.3732 | 4.0 | 128 | 0.3147 | 0.104 | 0.0 | 0.8960 | 0.1837 | 0.6493 |
| 0.3349 | 5.0 | 160 | 0.3102 | 0.104 | 0.0 | 0.8960 | 0.1837 | 0.6493 |
Framework versions
- Transformers 4.57.6
- Pytorch 2.9.0+cu126
- Datasets 4.0.0
- Tokenizers 0.22.2
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for bie-nhd/visobert-vithsd
Base model
uitnlp/visobert