# Unsloth Studio

Unsloth Studio is an open-source web interface for running and training machine learning models. It is open source and self-hosted, installable on Docker Hub and Unraid.

Canonical page: https://ourlocalhost.com/app/18/unsloth-studio

## Description

Unsloth Studio is an open-source web interface for running and training machine learning models. It lets users start model inference sessions, configure fine-tuning jobs, monitor progress, and review outputs from a browser. It is aimed at developers and homelab operators who want a guided local environment for model experimentation. It reduces setup friction compared with scripting inference and training workflows.

## Platform availability

Available on 3 of 9 platforms: Unraid, Mac, Linux.

## Desktop and other devices

Mac, Linux.

## Health score

89/100 (Excellent), based on 3 of 4 factors.
- Maintained: 100 (last commit 18 days ago)
- Popular: 98 (77,271 GitHub stars)
- Easy to install: 55 (3 install platforms, install notes, Docker image)

## Links

- Docker Hub: https://hub.docker.com/r/unsloth/unsloth
- desktop: https://github.com/unslothai/unsloth/releases/latest
- Official website: https://unsloth.ai
- GitHub: https://github.com/unslothai/unsloth
- Unraid: https://ca.unraid.net/apps/unsloth-studio-10lhn9b0s63lpv?q=Unsloth-Studio

## Common questions

### How can I open Unsloth Studio from another machine on my local network?

Start Studio with the host and port flags so it listens on the local network, such as port 8888. If you skip binding to all interfaces, it may only be reachable from the machine where it is installed.

### What installation method helps avoid dependency conflicts or the wrong GPU backend?

Install Unsloth inside a Python virtual environment and install it with the auto torch backend option. Skipping isolation can leave broken imports or the wrong GPU backend and make Studio fail to start.

### Why might agent requests fail even though the model is running?

Run the Hermes setup step first to configure inference providers and agent settings. If that step is skipped, provider addresses or credentials may be unset and agent requests can fail.

### Can external tools call the model through a standard API?

Yes, it can serve models through an OpenAI compatible API.

### Does it support multi-GPU and non-NVIDIA graphics hardware?

It supports multi-GPU setups, NVIDIA, AMD, and Intel GPUs, CPUs, and the Vulkan backend.

## Alternatives

- LLaMA Factory(Nvidia GPU): LLaMA Factory is an open-source, self-hosted framework for fine-tuning and serving large language models on NVIDIA GPU hardware.

---
Source: Our Local Host (https://ourlocalhost.com/app/18/unsloth-studio), the self-hosted and open-source app directory.