# Nerogar/OneTrainer

OneTrainer is a one-stop solution for all your Diffusion training needs.

Repository: https://github.com/Nerogar/OneTrainer
Canonical: https://ross.abutalabs.com/products/onetrainer
Language: Python
License: AGPL-3.0
License Family: copyleft
Topics: fine-tuning, lora, training, image-model-training
Last push: 2026-08-24T19:50:12+00:00

## Health v2 (maintenance only)
Score: 74/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 88
- inputs: {"age_days": 1239, "days_push": 9, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3184, forks 322 (observed 2026-08-28T04:07:47.997722+00:00)

## What it is
OneTrainer is a GUI and CLI application for fine-tuning diffusion image models, supporting full fine-tuning, LoRA, and embeddings across many model families like Stable Diffusion, SDXL, and FLUX. It includes dataset tooling, masked training, aspect ratio bucketing, automatic backups, and a built-in sampling UI.

## Use cases
- train a lora for stable diffusion
- fine-tune flux.1 on my own images
- create embeddings for sdxl
- caption an image dataset automatically with blip
- train a diffusion model with masked regions
- convert diffusion model checkpoints to diffusers format
- sample images during training without leaving the app

## When to choose
- you want a GUI-driven workflow for training diffusion models
- you need LoRA, embedding, or full fine-tuning support across many model architectures
- you want built-in dataset captioning, masking, and augmentation
- you want automatic backups and resumable training

## When to avoid
- you need to train LLMs or non-diffusion models
- you prefer writing your own training scripts with raw PyTorch
- you need distributed multi-node training at scale

## Facets
- artifact type: application
- maturity: active
- function: machine-learning, llm-training, image-processing, data-science, gui
- domain: machine-learning, deep-learning, image-processing, artificial-intelligence
- platform: windows, python
- tags: diffusion-models, lora-training, fine-tuning, stable-diffusion, flux, sdxl, text-to-image, model-training, tensorboard, linux, macos, gpu, desktop

## Member repositories
- Nerogar/OneTrainer (main) score 74

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:47.997722+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 2026-08-30T07:25:00.763494+00:00, confidence not recorded.
  - readme: https://github.com/Nerogar/OneTrainer (fetched 2026-08-28T04:07:47.997722+00:00, sha bdf593d8062b)
- Data as of 2026-08-30T08:39:29.467469+00:00.