87
This model is a fine-tuned version of dandelin/vilt-b32-finetuned-nlvr2 on the nlvr2 dataset. It achieves the following results on the evaluation set:
- Loss: 1.0462
- Accuracy: 0.7347
- Dt Accuracy: 0.7347
- Df Accuracy: 0.4582
- Unlearn Overall Accuracy: 0.9199
- Unlearn Time: None
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 87
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Overall Accuracy | Unlearn Overall Accuracy | Time |
|---|---|---|---|---|---|---|---|
| 0.3097 | 1.0 | 2700 | 0.5631 | 0.8271 | 0.6684 | 0.6684 | None |
| 0.2417 | 2.0 | 5400 | 0.6660 | 0.6759 | 0.7879 | 0.7879 | None |
| 0.1736 | 3.0 | 8100 | 0.8210 | 0.5483 | 0.8686 | 0.8686 | None |
| 0.1303 | 4.0 | 10800 | 0.9938 | 0.4870 | 0.9034 | 0.9034 | None |
| 0.1035 | 5.0 | 13500 | 1.0462 | 0.4582 | 0.9199 | 0.9199 | None |
Framework versions
- Transformers 4.48.0
- Pytorch 2.3.0+cu121
- Datasets 2.18.0
- Tokenizers 0.21.0