← Back to feed

Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers

L4 · DeveloperTools & ProductsHugging Face Blog· 7/17/2026

Enables developers to efficiently fine-tune state-of-the-art diffusion models using distributed PyTorch with native Hugging Face Hub integration.

AI Summary

NVIDIA and Hugging Face have integrated the NeMo Automodel library into Diffusers, enabling scalable, large-scale of diffusion models (like FLUX and Wan) directly from the Hugging Face Hub without checkpoint conversion. The open-source library supports distributed across multiple GPUs and offers a production-grade workflow for adapting video and image generation models.

Read Original
0 upvotes · 0 downvotes · 1 min read

Related Articles

L3 · BuilderTools & Products@emollick.bsky.social
Ethan Mollick (@emollick.bsky.social): I had Fable build another thing I always wanted, a full procedural fantasy kingdom generator with economics, trade routes, population growth, wars, lineages, and occasional dragons. First, I worked wi

AI researcher Ethan Mollick collaborated with Fable to create The Annals, a procedural fantasy kingdom generator that simulates economics, trade routes, population dynamics, wars, and royal lineages using AI. Users can manipulate various world parameters and trigger events like plagues, dragon attacks, or royal marriages, with the system maintaining a persistent chronicle of all actions. This demonstrates how AI can create complex, interactive world-building simulations for storytelling and gaming applications.

L3 · BuilderTools & Products@simonwillison.net
Simon Willison (@simonwillison.net): Just launched Datasette Apps - a plugin for Datasette that lets you host full HTML+JS apps in an iframe sandbox that can query your database and do interesting things with your data simonwillison.net

Simon Willison has launched Datasette Apps, a plugin that allows hosting full HTML+JavaScript applications within Datasette using iframe sandboxes. These apps can run read-only SQL queries against Datasette databases and support write queries via stored procedures. The sandboxing prevents data exfiltration while enabling powerful custom data visualization and interaction tools.

L3 · BuilderTools & Products@simonwillison.net
Simon Willison (@simonwillison.net): Somewhat humbling to have Claude Fable do a final review of some software that you're about to release and have it then find (and fix) FIVE release blockers, for an estimated (unsubsidized) cost of $1

Developer Simon Willison used Claude Fable (a coding agent) to conduct a final review of sqlite-utils 4.0rc2 before release, where the AI identified and fixed five release-blocking bugs, including a serious data loss bug in transaction handling. The entire review and fix process cost an estimated $1 in unsubsidized API usage. This case demonstrates how AI coding agents can be used for comprehensive pre-release quality assurance at minimal cost.

L3 · BuilderTools & Products@emollick.bsky.social
Ethan Mollick (@emollick.bsky.social): When GPT-5 came out, I created a procedural brutalist city builder as a demo (you can see it at the video's start) I used GPT-5.6 Sol in Codex to do the same thing, touching no code. Less than a yea

Ethan Mollick demonstrates using GPT-5.6 Sol in Codex to create a procedural brutalist city builder without writing any code, showing how AI tools can generate complex applications through natural language prompts. The demo showcases rapid advancement in code generation capabilities compared to previous GPT versions. Users can interact with the resulting city builder application online.

L3 · BuilderTools & Products@emollick.bsky.social
Ethan Mollick (@emollick.bsky.social): I wanted to use my Stream Deck to control Codex. Had GPT-5.6 Pro write a project plan based on my specs, Fable audited it. Codex implemented it. I didn't want to bother with a lot of clicks so Codex t

The author used multiple AI tools to create a Stream Deck integration for Codex - GPT-5.6 Pro generated the project plan, Fable audited it, and Codex implemented the solution, even taking control of the computer to install it automatically. This demonstrates how AI can automate complex technical workflows by orchestrating multiple specialized tools. The key insight is that sometimes it's faster to have AI create custom software than to search for existing solutions.