hamudizz/amharic-all-dataset-fine-tuning

amharic-all-dataset-fine-tuning # Amharic LLM Fine-Tuning with Unsloth A comprehensive project for fine-tuning modern open-weight LLMs on Amharic datasets using Unsloth optimization. ## 📁 Project Structure ``` amharic-all-dataset-fine-tuning/ ├── src/ # Source code │ ├── __init__.py │ ├── data_loader.py # Dataset loading utilities │ ├── preprocessor.py # Data preprocessing │ ├── trainer.py # Training pipeline │ ├── evaluator.py # Model evaluation │ └── utils.py # Helper functions ├── data/ # Dataset storage │ ├── raw/ # Raw downloaded datasets │ ├── processed/ # Processed datasets │ └── unified/ # Unified training data ├── models/ # Model checkpoints │ ├── base/ # Base model downloads │ └── fine-tuned/ # Fine-tuned models ├── configs/ # Configuration files │ ├── dataset_config.yaml │ ├── model_config.yaml │ └── training_config.yaml ├── notebooks/ # Jupyter notebooks │ ├── 01_data_exploration.ipynb │ ├── 02_training.ipynb │ └── 03_evaluation.ipynb ├── scripts/ # Executable scripts │ ├── download_datasets.py │ ├── train.py │ └── inference.py ├── tests/ # Unit tests ├── outputs/ # Training outputs │ ├── logs/ │ └── checkpoints/ ├── requirements.txt # Dependencies ├── setup.py # Package setup └── README.md # Documentation ``` ## 🚀 Quick Start ### 1. Setup Environment ```bash # Create virtual environment python3 -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate # Install dependencies pip install -r requirements.txt ``` ### 2. Download Datasets ```bash python scripts/download_datasets.py ``` ### 3. Train Model ```bash python scripts/train.py --config configs/training_config.yaml ``` ### 4. Run Inference ```bash python scripts/inference.py --model models/fine-tuned/amharic-llama3 ``` …