cs221-xlm-roberta-large-xnli-hin-finetuned-10-epochs
This model is a fine-tuned version of joeddav/xlm-roberta-large-xnli on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1252
- F1: 0.8655
- Roc Auc: 0.9205
- Accuracy: 0.8082
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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Roc Auc | Accuracy |
|---|---|---|---|---|---|---|
| 0.3935 | 1.0 | 64 | 0.3393 | 0.2137 | 0.5586 | 0.3092 |
| 0.2401 | 2.0 | 128 | 0.1938 | 0.6904 | 0.7911 | 0.6712 |
| 0.1466 | 3.0 | 192 | 0.1378 | 0.8240 | 0.8764 | 0.7886 |
| 0.1203 | 4.0 | 256 | 0.1263 | 0.8529 | 0.9051 | 0.8141 |
| 0.0811 | 5.0 | 320 | 0.1222 | 0.8516 | 0.9065 | 0.8063 |
| 0.0629 | 6.0 | 384 | 0.1281 | 0.8555 | 0.9099 | 0.7945 |
| 0.0523 | 7.0 | 448 | 0.1224 | 0.8581 | 0.9173 | 0.8023 |
| 0.038 | 8.0 | 512 | 0.1180 | 0.8649 | 0.9186 | 0.8082 |
| 0.0307 | 9.0 | 576 | 0.1252 | 0.8655 | 0.9205 | 0.8082 |
| 0.0275 | 10.0 | 640 | 0.1256 | 0.8639 | 0.9185 | 0.8082 |
Framework versions
- Transformers 4.48.0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0