训练与微调MITPython

AI Toolkit

Ostris’ training suite for FLUX, video and image model LoRAs.

数据来自 GitHub API · 更新于 2026-10-15

AI Toolkit 是什么？

AI Toolkit is a popular trainer for FLUX and newer image/video model LoRAs, driven by YAML configs and a web UI for managing jobs and datasets.

根据项目 README 整理。开源项目更新很快，请以 GitHub 页面上的最新说明为准。

Clone, create a venv and install (install PyTorch first per the README).

git clone https://github.com/ostris/ai-toolkit.git
cd ai-toolkit
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

软件	Star 数	许可证	显存
AI Toolkit Ostris’ training suite for FLUX, video and image model LoRAs.	12.3K	MIT	24 GB recommended for FLUX
Kohya’s GUI GUI for training LoRA, DreamBooth and fine-tunes for SD, SDXL and FLUX.	12.6K	Apache-2.0	8 GB+ for SD1.5 LoRA; 12–24 GB for SDXL/FLUX
OneTrainer One-stop trainer for diffusion models with a desktop UI.	3.2K	AGPL-3.0	8 GB+
Unsloth Fine-tune LLMs 2× faster with up to 70% less memory.	78K	Apache-2.0	3 GB+ depending on model size
Axolotl Config-driven LLM fine-tuning for many models and techniques.	12.6K	Apache-2.0	—
LLaMA-Factory Unified fine-tuning for 100+ LLMs and VLMs with a no-code web UI.	75.5K	Apache-2.0	—

AI Toolkit 免费吗？

是的——AI Toolkit 是开源软件，采用 MIT 许可证。商用前请确认软件及模型权重的许可条款。

AI Toolkit 支持 Windows、Mac 和 Linux 吗？

AI Toolkit 支持：Linux, Windows。

运行 AI Toolkit 需要什么配置？

NVIDIA. VRAM: 24 GB recommended for FLUX

如何安装 AI Toolkit？

1. Clone, create a venv and install (install PyTorch first per the README). 2. Run a training config.

AI Toolkit 有哪些好的替代品？

常见替代品包括 Kohya’s GUI, OneTrainer, Unsloth, Axolotl，详见本页对比表。