shawgpt-ft-LoRA-rank-5
This model is a fine-tuned version of TheBloke/Mistral-7B-Instruct-v0.2-GPTQ on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 3.8768
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT 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: 2
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 25.5434 | 0.5714 | 1 | 4.2401 |
| 25.8374 | 1.5714 | 2 | 4.2223 |
| 25.352 | 2.5714 | 3 | 4.1701 |
| 25.0898 | 3.5714 | 4 | 4.1139 |
| 24.6202 | 4.5714 | 5 | 4.0553 |
| 24.2887 | 5.5714 | 6 | 3.9994 |
| 23.9322 | 6.5714 | 7 | 3.9512 |
| 23.7971 | 7.5714 | 8 | 3.9143 |
| 23.5872 | 8.5714 | 9 | 3.8893 |
| 15.2722 | 9.5714 | 10 | 3.8768 |
Framework versions
- PEFT 0.14.0
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.3.1
- Tokenizers 0.21.0
- Downloads last month
- 1
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for Jonasbukhave/shawgpt-ft-LoRA-rank-5
Base model
mistralai/Mistral-7B-Instruct-v0.2
Quantized
TheBloke/Mistral-7B-Instruct-v0.2-GPTQ