# meta-pytorch/torchtune

PyTorch native post-training library

Repository: https://github.com/meta-pytorch/torchtune
Canonical: https://ross.abutalabs.com/products/torchtune
Homepage: https://pytorch.org/torchtune/main/
Language: Python
License: BSD-3-Clause
License Family: permissive
Last push: 2026-08-26T11:43:39+00:00

## Health v2 (maintenance only)
Score: 69/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 28, longevity 74
- inputs: {"age_days": 1048, "days_push": 7, "days_rel": 513, "gap_med": 36, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5801, forks 747 (observed 2026-08-28T04:09:29.749587+00:00)

## What it is
Torchtune is a PyTorch-native library for authoring, post-training, and experimenting with large language models. It provides hackable training recipes (SFT, DPO, PPO, GRPO, distillation, QAT), simple implementations of popular LLMs, and YAML-based configuration for training, evaluation, quantization, and inference.

## Use cases
- fine-tune llama on my own dataset
- run LoRA or QLoRA finetuning of an open LLM
- train a model with DPO or PPO preference optimization
- distill a large language model into a smaller one
- experiment with LLM training recipes in pure PyTorch
- finetune a multimodal vision-language model
- quantization-aware training for LLMs

## When to choose
- you want hackable, PyTorch-native fine-tuning code rather than a black-box trainer
- you need memory-efficient single- or multi-node finetuning of Llama, Qwen, Gemma, Mistral, or Phi models
- you want YAML-configured recipes for SFT, DPO, PPO, GRPO, or distillation

## When to avoid
- you need a library under active development - torchtune wound down in 2025 and is no longer actively maintained
- you need point-and-click GUI-based fine-tuning
- you need inference-only serving rather than post-training

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-training, machine-learning, deep-learning, cli, configuration-management
- domain: large-language-models, machine-learning, deep-learning, artificial-intelligence
- platform: python, cli, cross-platform
- tags: fine-tuning, lora, qlora, sft, dpo, ppo, grpo, knowledge-distillation, pytorch, llama, qwen, gemma, mistral, multimodal, yaml-configs, wound-down, gpu

## Member repositories
- meta-pytorch/torchtune (main) score 69

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:29.749587+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:52:46.228549+00:00, confidence not recorded.
  - readme: https://github.com/meta-pytorch/torchtune (fetched 2026-08-28T04:09:29.749587+00:00, sha a4bd20437b81)
  - homepage: https://pytorch.org/torchtune/main/ (fetched 2026-08-29T08:47:54.804554+00:00, sha b406e0b2662e)
  - site_page: https://meta-pytorch.org/torchtune/main/install.html (fetched 2026-08-29T08:47:54.813557+00:00, sha c331fbe12fe9)
  - site_page: https://meta-pytorch.org/torchtune/main/deep_dives/configs.html (fetched 2026-08-29T08:47:54.816582+00:00, sha 400874749612)
- Data as of 2026-08-30T08:39:29.467469+00:00.