Spaces:
Sleeping
Sleeping
File size: 3,309 Bytes
b619545 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 |
---
title: Gemma Fine-tuning UI
emoji: 🤖
colorFrom: blue
colorTo: indigo
sdk: streamlit
sdk_version: 1.30.0
app_file: app.py
pinned: false
---
# Gemma Fine-tuning UI
A web-based user interface for fine-tuning Google's Gemma models using Hugging Face infrastructure.
## Features
- **Dataset Upload**: Upload and preprocess your custom training data in CSV, JSON, or JSONL format
- **Model Configuration**: Configure Gemma model version and hyperparameters
- **Training Management**: Start, monitor, and manage fine-tuning jobs
- **Evaluation**: Test your fine-tuned model with interactive generation
- **Export Options**: Use your model directly from Hugging Face or export in various formats
## Installation
1. Clone this repository:
```bash
git clone https://github.com/yourusername/gemma-finetuning-ui.git
cd gemma-finetuning-ui
```
2. Create a virtual environment and install dependencies:
```bash
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txt
```
3. Run the application:
```bash
streamlit run app.py
```
## Usage
1. **Authentication**: Provide your Hugging Face API token with write permissions
2. **Dataset Preparation**: Upload your dataset and configure column mappings
3. **Model Selection**: Choose between Gemma 2B or 7B and customize training parameters
4. **Training**: Start the fine-tuning process and monitor progress
5. **Evaluation**: Test your fine-tuned model with custom prompts
6. **Deployment**: Export or directly use your model from Hugging Face
## Hugging Face Spaces Deployment
This application is designed to be deployed easily on Hugging Face Spaces:
1. Create a new Space on [Hugging Face Spaces](https://huggingface.co/spaces)
2. Select Streamlit as the SDK
3. Connect your GitHub repository or upload the files directly
4. The Space will automatically detect and install the requirements
## Requirements
- Python 3.8+
- Streamlit 1.30.0+
- Hugging Face Account with API token
- For training: GPU access (recommended)
## Project Structure
```
.
├── app.py # Main Streamlit application
├── pages/ # Multi-page app components
│ ├── 01_Dataset_Upload.py
│ ├── 02_Model_Configuration.py
│ ├── 03_Training_Monitor.py
│ └── 04_Evaluation.py
├── utils/ # Utility functions
│ ├── auth.py # Authentication utilities
│ ├── huggingface.py # Hugging Face API integration
│ ├── training.py # Training utilities
│ └── ui.py # UI components and styling
├── data/ # Sample data and uploads
└── requirements.txt # Project dependencies
```
## License
This project is licensed under the MIT License - see the LICENSE file for details.
## Acknowledgements
- [Google Gemma Models](https://huggingface.co/google/gemma-7b)
- [Hugging Face Transformers](https://huggingface.co/docs/transformers/index)
- [Streamlit](https://streamlit.io/)
- [PEFT - Parameter-Efficient Fine-Tuning](https://github.com/huggingface/peft)
---
Developed as a simplified interface for fine-tuning Gemma models with Hugging Face. |