phi-3-mini-4k-qlora-modular_new_prompt
This model is a fine-tuned version of microsoft/Phi-3-mini-4k-instruct on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8322
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: 0.0002
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Use OptimizerNames.ADAMW_TORCH 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: 100
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.1352 | 1.0 | 12 | 2.9673 |
| 2.5859 | 2.0 | 24 | 2.0797 |
| 1.4657 | 3.0 | 36 | 1.0276 |
| 0.876 | 4.0 | 48 | 0.7854 |
| 0.6362 | 5.0 | 60 | 0.6661 |
| 0.4369 | 6.0 | 72 | 0.6698 |
| 0.2396 | 7.0 | 84 | 0.6673 |
| 0.1336 | 8.0 | 96 | 0.8322 |
Framework versions
- PEFT 0.16.0
- Transformers 4.53.1
- Pytorch 2.6.0+cu124
- Datasets 2.14.4
- Tokenizers 0.21.2