ModernBERT_base-cefr-all-classifier
This model is a fine-tuned version of answerdotai/ModernBERT-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1620
- F1: 0.9528
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: 3.6e-05
- train_batch_size: 2
- eval_batch_size: 3
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Use 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: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|---|---|---|---|---|
| 2.1765 | 1.0 | 13623 | 0.1331 | 0.9438 |
| 1.604 | 2.0 | 27246 | 0.1180 | 0.9503 |
| 1.1179 | 2.9998 | 40866 | 0.1620 | 0.9528 |
Framework versions
- Transformers 4.51.0
- Pytorch 2.5.1+cu121
- Datasets 3.1.0
- Tokenizers 0.21.0