amins01

amins01 / tiny-models

Public

Small generative models built and trained from scratch to run locally in the browser.

amins01.github.iotiny-models
36
7
100% credibility
Found Feb 23, 2026 at 20 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A browser-based demo app with small AI generators that turn colored sketches into landscape images or edit uploaded photos via text prompts, running entirely locally.

How It Works

1
🌐 Discover Tiny Models

You stumble upon a cool web page showcasing tiny AI tools that create images from your drawings or edit photos, all happening right in your web browser.

2
🎛️ Pick Your Creative Mode

Choose between sketching a landscape scene or uploading a photo to edit with simple words.

3
🖌️ Draw or Describe

Grab the drawing tools to paint a rough landscape with colors like sky and trees, or upload your photo and type what changes you want, like 'add a sunset'.

4
✨ Start the Magic

Tap generate and watch as the image comes alive step by step, building your vision from noise into a stunning picture.

5
⚙️ Fine-Tune Results

Slide to adjust detail level or reset for fresh starts, seeing changes happen instantly on screen.

🖼️ Your Artwork Shines

Enjoy beautiful landscapes from your doodles or perfectly edited photos, all created privately on your own device without sending anything online.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 20 to 36 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is tiny-models?

This GitHub small project delivers tiny generative models trained from scratch that generate images entirely in your browser—no servers, no APIs. Draw semantic sketches with a color palette to create 256x256 landscapes, or upload a photo and add a text prompt for 128x128 edits, all powered by TypeScript React app with ONNX Runtime Web and WASM for CPU inference. It's small generative AI that prioritizes local privacy and zero costs for narrow tasks like these demos.

Why is it gaining traction?

Unlike bloated cloud-dependent models, these small generative models run smoothly client-side, showing real-time denoising steps and latency stats in a polished dashboard—perfect for devs testing browser AI limits. The included Python training pipelines let you retrain from scratch on custom datasets, bridging education and experimentation in one small GitHub repo. For small business generative AI or hobbyists, it's instant offline demos without infra hassles.

Who should use this?

Frontend devs embedding local AI in web apps, indie hackers prototyping small generative AI tools, or educators demoing diffusion models in classrooms without GPUs. Suited for narrow use cases like sketch-to-image converters or quick photo editors in PWAs.

Verdict

Grab it for learning or proofs-of-concept—browser inference works, docs via inline blog are solid, but with 12 stars and 1.0% credibility, it's early-stage; expect tweaks for production. Solid small GitHub project for tiny-models fans.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.