mDeBERTa-v3-base-mnli-xnli-finetuned-econ_frame-nli-classification
This model is a fine-tuned version of MoritzLaurer/mDeBERTa-v3-base-mnli-xnli on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3348
- Accuracy: 0.9324
- Accuracy Balanced: 0.9098
- F1 Macro: 0.9191
- F1 Weighted: 0.9316
- F1 Positive: 0.8862
- Precision Positive: 0.925
- Recall Positive: 0.8506
- Precision Weighted: 0.9321
- Recall Weighted: 0.9324
- Mcc: 0.8397
- Roc Auc: 0.9414
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: 8
- eval_batch_size: 8
- 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: 0.06
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Accuracy Balanced | F1 Macro | F1 Weighted | F1 Positive | Precision Positive | Recall Positive | Precision Weighted | Recall Weighted | Mcc | Roc Auc |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 164 | 0.2904 | 0.9253 | 0.9173 | 0.9134 | 0.9256 | 0.8814 | 0.8667 | 0.8966 | 0.9262 | 0.9253 | 0.8271 | 0.9373 |
| No log | 2.0 | 328 | 0.3162 | 0.9288 | 0.9041 | 0.9145 | 0.9278 | 0.8795 | 0.9241 | 0.8391 | 0.9286 | 0.9288 | 0.8311 | 0.9548 |
| No log | 3.0 | 492 | 0.3348 | 0.9324 | 0.9098 | 0.9191 | 0.9316 | 0.8862 | 0.925 | 0.8506 | 0.9321 | 0.9324 | 0.8397 | 0.9414 |
Framework versions
- Transformers 4.57.6
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2