whereismyai
OneTrainer
OneTrainer is a software application for training diffusion models, including full fine-tuning, LoRA, and embeddings.
OneTrainer provides a unified environment for training various diffusion models, including Stable Diffusion, FLUX.1, PixArt-Alpha, and Hunyuan Video. Users can perform full fine-tuning, LoRA training, and embedding training through a graphical user interface or a command-line interface. It includes tools for automatic dataset captioning using BLIP or WD-1.4, masking with ClipSeg or Rembg, and model format conversion. The software supports aspect ratio bucketing, multi-resolution training, and internal sampling during the training process. It is designed for users who require comprehensive control over the model training workflow, offering features like automatic backups and TensorBoard integration for progress tracking.
OneTrainer at a glance
- Best for: Designer, Software engineer
- Runs on: Windows, Linux, macOS, CLI
- Pricing: Free
- Released: April 2023
- Category: AI image tool
- Website: github.com/Nerogar/OneTrainer
Key features
- Support for multiple diffusion model architectures
- Full fine-tuning, LoRA, and embedding training methods
- Automatic dataset captioning and mask creation
- Aspect ratio bucketing and multi-resolution training
- Integrated sampling UI for model testing during training
- Model format conversion between diffusers and ckpt
- TensorBoard integration for training progress tracking
- Automatic progress backups
Use it for
- Fine-tuning Stable Diffusion or FLUX models
- Training LoRA adapters for custom image styles
- Automating dataset captioning and preprocessing
- Creating masks for focused image training
Alternatives to OneTrainer
- SimpleTuner — It supports dozens of model families with advanced memory and enterprise features.
- LoRA Scripts — Provides a complete web interface and scripts for kohya-ss trainer.
- Diffusers — A modular toolbox for inference and training of diffusion models.
- LyCORIS — It provides diverse algorithms like LoHa, LoKr, and DyLoRA for Stable Diffusion fine-tuning.
- DiffusionBee — Runs generative artificial intelligence locally without cloud uploads.
- Easy Diffusion — Hassle-free 1-click installation with no technical knowledge required.
- stable-diffusion.cpp — Runs diffusion models efficiently in pure C and C++ with no external dependencies.
- Dsh Sd Bridge Guard — Connects DeepSeek directly to local Stable Diffusion without manual setup.
- Flux — Runs via API or as open weights on your own infrastructure.
- Krea — Every major image and video model plus a 21K enhancer in one tool
- KohakuMossTrainer — Fits adapter training onto cards with 6 GB of VRAM.
- SwarmUI — It offers modular powertools and high performance for managing multiple AI models.
All OneTrainer alternatives