# huggingface/finetrainers

Scalable and memory-optimized training of diffusion models

Repository: https://github.com/huggingface/finetrainers
Canonical: https://ross.abutalabs.com/products/finetrainers
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai, art, artificial-intelligence, diffusers, diffusion, diffusion-models, pytorch, transformers
Last push: 2026-05-26T10:33:13+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 84, release rhythm 40, longevity 50
- inputs: {"age_days": 707, "days_push": 99, "days_rel": 495, "gap_med": 30.0, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1358, forks 141 (observed 2026-08-28T04:04:29.693931+00:00)

## What it is
finetrainers is a Hugging Face library for scalable, memory-optimized training (fine-tuning) of diffusion models, including LoRA training of video generation models like CogVideoX, LTX-Video, and Wan. It builds on PyTorch and diffusers and provides reproducible example training scripts and dataset utilities.

## Use cases
- train a lora for a video diffusion model
- fine-tune cogvideox on my own dataset
- memory-efficient training of diffusion transformers
- replicate pika effects style video generation
- train ltx-video on custom video-caption pairs
- run sft training for wan text-to-video model

## When to choose
- you want to fine-tune or LoRA-train supported diffusion/video models with memory optimizations
- you want reproducible example training scripts built on diffusers
- you need scalable training on limited GPU memory

## When to avoid
- you need training for models outside its support matrix
- you want a stable, fully polished tool - the project describes itself as work-in-progress
- you need image-only diffusion training with broad model coverage like other trainers

## Facets
- artifact type: library
- maturity: active
- function: llm-training, machine-learning, deep-learning
- domain: machine-learning, deep-learning, artificial-intelligence, image-processing
- platform: python, cross-platform
- tags: diffusion-models, lora, fine-tuning, pytorch, diffusers, video-generation, text-to-video, memory-optimization, video, gpu, linux

## Member repositories
- huggingface/finetrainers (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:29.693931+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 2026-08-30T04:41:46.137757+00:00, confidence not recorded.
  - readme: https://github.com/huggingface/finetrainers (fetched 2026-08-28T04:04:29.693931+00:00, sha 5a4dbfa530fe)
- Data as of 2026-08-30T08:39:29.467469+00:00.