Mistral-7B-Instruct-v0.3-LoRA-Legal
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.8193
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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.1687 | 1.0 | 421 | 1.8837 |
| 1.1129 | 2.0 | 842 | 1.8581 |
| 1.0754 | 3.0 | 1263 | 1.8415 |
| 1.1104 | 4.0 | 1684 | 1.8316 |
| 1.0823 | 5.0 | 2105 | 1.8253 |
| 1.1225 | 6.0 | 2526 | 1.8225 |
| 1.0894 | 7.0 | 2947 | 1.8209 |
| 1.1285 | 8.0 | 3368 | 1.8200 |
| 1.1105 | 9.0 | 3789 | 1.8193 |
| 1.1307 | 10.0 | 4210 | 1.8193 |
Framework versions
- PEFT 0.17.1
- Transformers 4.56.0
- Pytorch 2.8.0+cu126
- Datasets 4.0.0
- Tokenizers 0.22.0