https://github.com/jxtngx/tunelab

A web app for post-training
https://github.com/jxtngx/tunelab

artificial-intelligence fastapi generative-ai nextjs python pytorch shadcn-ui typescript

Last synced: 3 months ago
JSON representation

A web app for post-training

• Host: GitHub
• URL: https://github.com/jxtngx/tunelab
• Owner: jxtngx
• License: apache-2.0
• Created: 2025-01-09T20:56:28.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2026-03-24T12:20:00.000Z (3 months ago)
• Last Synced: 2026-03-25T00:15:42.236Z (3 months ago)
• Topics: artificial-intelligence, fastapi, generative-ai, nextjs, python, pytorch, shadcn-ui, typescript
• Language: TypeScript
• Homepage:
• Size: 1.15 MB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Funding: .github/FUNDING.yml
  • License: License
  • Codeowners: .github/CODEOWNERS

Awesome Lists containing this project

README

          
# Tune Lab

Tune Lab is a [FastAPI](https://fastapi.tiangolo.com) + [NextJS](https://nextjs.org) application that is purpose built for post-training with [NVIDIA NeMo](https://github.com/NVIDIA/NeMo) and [HuggingFace TRL](https://github.com/huggingface/trl) on MacOS + MPS, NVIDIA DGX devices + CUDA, and CUDA compatible AWS EC2 DLAMI instances.

> [!IMPORTANT]

> Tune Lab is in active development and is not yet ready for use

> [!IMPORTANT]

> Tune Lab is built alongside [Cursor](https://cursor.com/) and [Anthropic's](https://www.anthropic.com/) Claude and is closed for contributions

## Intended Use

Tune Lab is intended to be cloned and used locally on MacOS or NVIDIA DGX devices, or hosted on AWS. 

> [!IMPORTANT]

> Cloning the repository does not entitle a user to support from the maintainer

## High-Level Features

- a RESTful API for managing and tuning models with NeMo and TRL

- a NextJS web interface for managing and tuning models

- a Cursor directory for advanced workflows across finetuning, backend infra, and frontend development

## Downstream Usage

Models tuned with Tune Lab are intended to be used with LangChain or AWS Strands or NVIDIA NeMo Agent Toolkit.