← back to the directory
Library / SDK Fine-tuning & Training

torchtune

1A 2PyTorch-native 3library 4for 5finetuning 6LLMs.

the six Ws · specification

W1 Who

Maintained by the PyTorch team.

W2 What

Offers hackable recipes for fine-tuning and evaluating language models.

W3 Where

A Python library on local or cloud GPUs.

W4 When

When fine-tuning an open model from recipes.

W5 Why

Native PyTorch fine-tuning without heavy abstractions.

W6 With

PyTorch, and optional Hugging Face weights.

W7 Watch

Official PyTorch project, BSD-licensed, no credentials; runs on your GPUs.

PyTorchLoRATraining

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.