https://github.com/neuralwork/instruct-finetune-mistral
Fine-tune Mistral 7B to generate fashion style suggestions
https://github.com/neuralwork/instruct-finetune-mistral
finetuning-llms huggingface llm llm-inference mistral peft quantization
Last synced: 13 days ago
JSON representation
Fine-tune Mistral 7B to generate fashion style suggestions
- Host: GitHub
- URL: https://github.com/neuralwork/instruct-finetune-mistral
- Owner: neuralwork
- License: mit
- Created: 2023-12-06T15:22:59.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-01-02T17:13:12.000Z (over 1 year ago)
- Last Synced: 2024-01-02T18:33:20.961Z (over 1 year ago)
- Topics: finetuning-llms, huggingface, llm, llm-inference, mistral, peft, quantization
- Language: Python
- Homepage:
- Size: 19.5 KB
- Stars: 7
- Watchers: 0
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
## Fine-tuning LLMs with PEFT
This project is a tutorial on parameter-efficient fine-tuning (PEFT) and quantization of the [Mistral 7B v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) model. We use LoRA for PEFT and 4-bit quantization to compress the model, and fine-tune the model on a semi-manually crafted fashion style recommendation instruct [dataset](https://huggingface.co/datasets/neuralwork/fashion-style-instruct). For more information and a step by step guide, see our [blog post](https://blog.neuralwork.ai/an-llm-fine-tuning-cookbook-with-mistral-7b/).
## Usage
Start by cloning the repository, setting up a conda environment and installing the dependencies. We tested our scripts with python 3.9 and CUDA 11.7.
```
git clone https://github.com/neuralwork/finetune-mistral.git
cd finetune-mistral
conda create -n llm python=3.9
conda activate llm
pip install -r requirements.txt
```
You can finetune the model on our fashion-style-instruct [dataset](https://huggingface.co/datasets/neuralwork/fashion-style-instruct) or another dataset. Note that you will need to have the same features as our dataset and pass in your HF Hub token as an argument if using a private dataset. Fine-tuning takes about 2 hours on a single A40, you can either use the default accelerate settings or configure it to use multiple GPUS. To fine-tune the model:
```
accelerate config default
python finetune_model.py --dataset= --base_model="mistralai/Mistral-7B-v0.1" --model_name= --auth_token= --push_to_hub
```
One model training is completed, only the fine-tuned (LoRA) parameters are saved, which are loaded to overwrite the corresponding parameters of the base model during testing.
To test the fine-tuned model with a random sample selected from the dataset, run `python test.py`. To launch the full Gradio demo and play around with your own examples, launch the demo with `python app.py`
## License
This project is licensed under the [MIT license](https://github.com/neuralwork/finetune-mistral/blob/main/LICENSE).
From [neuralwork](https://neuralwork.ai/) with :heart: