shawgpt-ft-lr0.0002-wd0.01
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: 3.2788
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 |
| 25.7033 | 1.5714 | 2 | 4.1610 |
| 24.6801 | 2.5714 | 3 | 3.9766 |
| 23.5882 | 3.5714 | 4 | 3.8070 |
| 22.4572 | 4.5714 | 5 | 3.6633 |
| 21.6525 | 5.5714 | 6 | 3.5396 |
| 20.9261 | 6.5714 | 7 | 3.4377 |
| 20.4749 | 7.5714 | 8 | 3.3594 |
| 20.0629 | 8.5714 | 9 | 3.3061 |
| 12.9157 | 9.5714 | 10 | 3.2788 |
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
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for Jonasbukhave/shawgpt-ft-lr0.0002-wd0.01
Base model
mistralai/Mistral-7B-Instruct-v0.2 Quantized
TheBloke/Mistral-7B-Instruct-v0.2-GPTQ