NVIDIA and Hugging Face team up to make training giant image and video AI models far easier

Key points

• NVIDIA and Hugging Face jointly released NeMo Automodel, a free, open-source training tool, in June 2025.
• The tool supports fine-tuning of models including FLUX.1-dev (12 billion parameters) and HunyuanVideo (13 billion parameters), two of the most capable open image and video generators available.
• Developers can start training directly from models hosted on the Hugging Face Hub, the world's largest public repository of AI models, without converting files to a special format first.
• The software is free to use under the Apache 2.0 licence, meaning anyone can use, modify, and build on it commercially.

Teaching an AI to generate pictures in a specific style, or to produce videos that match a company's visual brand, is expensive and technically demanding work. Until now, doing it at scale typically meant rewriting software from scratch every time a new model came along.

NVIDIA and Hugging Face want to change that.

The two companies have published NeMo Automodel, an open-source library, meaning free software whose code anyone can inspect and change, that connects NVIDIA's industrial-scale training machinery to the Hugging Face Diffusers library. Diffusers is the most widely used toolkit for running diffusion models, the class of AI that generates images and video by gradually refining random noise into a finished picture.

In plain terms: a developer who wants to teach FLUX.1-dev, one of the best publicly available text-to-image generators, to draw in the style of vintage tarot cards can now do that without touching any specialised code. They point the tool at their dataset, pick a configuration file, and run it.

What does this actually change for people building AI tools?

Before this release, moving a model from the Hugging Face library into a training system often required converting the model's saved weights, the numerical values the AI has learned, into a different file format, training, then converting back. NeMo Automodel skips that entirely. A fine-tuned model, one that has been further trained on new examples to specialise its skills, loads straight back into standard tools the moment training finishes.

The tool also handles scale automatically. A small team with one powerful machine and a large research group with hundreds of GPUs, the specialised chips that do the heavy number-crunching AI needs, can both use the same configuration files. Switching between different ways of splitting a model across many chips is a settings change, not a coding project.

For developers with limited hardware, the tool supports LoRA, a technique that fine-tunes only a small fraction of a model's values rather than all of them, cutting memory use sharply while keeping quality high.

The blog post, first published by Hugging Face, walks through a worked example: fine-tuning FLUX.1-dev on 78 public-domain tarot card images. The whole training run takes 200 steps and produces a model that generates tarot-style images on demand.

Supported models at launch include FLUX.1-dev for images and both Wan 2.1 and HunyuanVideo for video generation.

For most ordinary users, none of this changes anything directly today. The people this affects are the developers and researchers who build the image tools, creative apps, and business software that ordinary people eventually use. Better, cheaper training pipelines tend to mean more specialised AI features arriving faster and at lower cost.

The software is available now on GitHub and is documented in the official Diffusers training guide.