tass-xlm-roberta-large-bs16
This model is a fine-tuned version of xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9776
- Precision: 0.6775
- Recall: 0.6798
- F1 Macro: 0.6779
- Accuracy: 0.6785
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- 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_steps: 0.1
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 Macro | Accuracy |
|---|---|---|---|---|---|---|---|
| 2.1909 | 1.0 | 121 | 1.0714 | 0.5796 | 0.3592 | 0.2984 | 0.3725 |
| 1.8080 | 2.0 | 242 | 0.8160 | 0.6572 | 0.6289 | 0.5964 | 0.6452 |
| 1.4524 | 3.0 | 363 | 0.8226 | 0.6586 | 0.6450 | 0.6394 | 0.6608 |
| 1.2175 | 4.0 | 484 | 0.7491 | 0.6904 | 0.6586 | 0.6662 | 0.6691 |
| 1.0099 | 5.0 | 605 | 0.7660 | 0.6829 | 0.6869 | 0.6829 | 0.6847 |
| 0.9551 | 6.0 | 726 | 0.8456 | 0.6782 | 0.6785 | 0.6783 | 0.6816 |
| 0.6754 | 7.0 | 847 | 0.8812 | 0.6798 | 0.6572 | 0.6640 | 0.6660 |
| 0.6418 | 8.0 | 968 | 0.9776 | 0.6775 | 0.6798 | 0.6779 | 0.6785 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 4.8.5
- Tokenizers 0.22.2