Skip to content

corticalstack/genai-playground

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

21 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

πŸš€ GenAI Playground

A comprehensive toolkit and collection of examples for working with Large Language Models (LLMs), from environment setup to advanced fine-tuning, quantization, and evaluation.

πŸ“š Description

GenAI Playground provides a structured approach to experimenting with and deploying state-of-the-art language models. This repository contains practical examples and workflows for the entire LLM lifecycle, including environment setup, inference, dataset preparation, fine-tuning, quantization, model merging, and evaluation.

Each section includes Jupyter notebooks with detailed examples and conda environment configurations, making it easy to get started with different aspects of LLM development.

✨ Features

  • Environment Setup: Ready-to-use conda environment configurations for different LLM tasks
  • Batch Inference: Examples for running inference on LLMs with transformers
  • Inference Engines: Integration with popular inference engines like VLLM and Ooba
  • Dataset Preparation: Tools and examples for creating and formatting datasets for LLM training
  • Fine-tuning:
    • Supervised Fine-Tuning (SFT) examples with various models and datasets
    • Direct Preference Optimization (DPO) implementations
    • Support for efficient fine-tuning methods like QLoRA and Unsloth
  • Quantization: Examples for quantizing models to various formats (GGUF, GPTQ, AWQ, EXL2)
  • Model Merging: Techniques for merging models using methods like SLERP, TIES, and DARE
  • Evaluation: Integration with LLM Evaluation Harness for benchmarking model performance

πŸ”§ Prerequisites

  • CUDA-capable GPU
  • Python 3.11.7
  • Git
  • Conda package manager

πŸš€ Getting Started

1. Clone the repository

git clone https://github.com/corticalstack/genai-playground.git
cd genai-playground

2. Set up the environment

Choose the appropriate environment based on your task, but see the environment section README for more.

πŸ“š Resources

πŸ“„ License

This project is licensed under the MIT License - see the LICENSE file for details.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published