# Unsloth

> Use this tool when you need to efficiently fine-tune large language models, reducing VRAM usage and training time. It solves problems like limited GPU resources and lengthy training periods, providing inputs like model selection and training data, and outputs like fine-tuned models and generated text. Ideal for use cases involving large-scale language model optimization, such as text generation and model export.

Canonical page: https://skillsregistry.net/skills/unsloth-llm-fine-tuning  
JSON: https://api.skillsregistry.net/v1/skills/unsloth-llm-fine-tuning

## Description

An MCP server for Unsloth, a library that dramatically improves large language model fine-tuning efficiency by reducing VRAM usage and training time. The server provides tools for checking Unsloth installation, listing supported models, loading models, fine-tuning, text generation, and model export, with optimizations like 4-bit quantization and extended context length support for models like Llama, Mistral, and Phi.

## Trust

- **Trust score (0–1):** 0.65
- **Verification tier:** scanned
- **Last scanned:** 2026-09-02

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** other
- **Updated:** 2026-09-02

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/unsloth-llm-fine-tuning)
- **Repository:** <https://github.com/ototao/unsloth-mcp-server>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "unsloth-llm-fine-tuning"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/unsloth-llm-fine-tuning` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/unsloth-llm-fine-tuning/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt