README.md

metadata

license: mit
title: Gemma FineTuner
sdk: gradio
emoji: 🔥
colorFrom: gray
colorTo: green
pinned: true

Gemma Fine-tuning UI

A user-friendly interface for fine-tuning Google's Gemma models using Unsloth optimizations.

Features

• Easy-to-use web interface for model fine-tuning
• Support for multiple data formats (CSV, JSONL, TEXT)
• Parameter-efficient fine-tuning with LoRA
• Real-time training progress visualization
• Model export in multiple formats
• Integrated text generation testing

Installation

git clone https://github.com/codewithdark-git/Gemma-Finetune.git
cd Gemma-Finetune
pip install -r requirements.txt

Usage

1. Run the application:

python main.py

2. Follow the UI steps:
   • Upload your dataset
   • Configure model parameters
   • Start training
   • Test and export your model

Requirements

See requirements.txt for detailed dependencies.

License

MIT License