hanxiao

hanxiao / umap-mlx

Public

UMAP in pure MLX for Apple Silicon. 30x faster than umap-learn.

apple-silicon dimensionality-reduction machine-learning metal mlx
34
3
100% credibility
Found Mar 01, 2026 at 30 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

umap-mlx is an optimized implementation of the UMAP algorithm for fast dimensionality reduction and data visualization on Apple Silicon hardware.

How It Works

1
🔍 Discover the speedy visualizer

You hear about a super-fast tool that turns piles of complex data into easy-to-see pictures, perfect for Apple computers.

2
💻 Set it up on your Mac

You download it and get everything ready with a simple setup so it's all prepared to use.

3
📊 Load your data

You pick your data file full of numbers and points, like images or measurements.

4
⚡ Create the simple view

You ask it to simplify your data into a 2D picture, and it works lightning-fast, crunching everything in seconds.

5
👀 Watch clusters form

You see points group into colorful clusters, revealing hidden patterns in your data.

🎉 Share your insights

You now have beautiful, clear visualizations to understand and share what your data really means.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 30 to 34 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 umap-mlx?

umap-mlx brings UMAP dimensionality reduction to Apple Silicon via pure Python and MLX, running the full pipeline on Metal GPU. It delivers embeddings for visualization or clustering just like umap-learn on GitHub, but with datasets like 70K Fashion-MNIST finishing in seconds via a drop-in API: load data as numpy arrays, call fit_transform, get 2D plots. Targets users seeking fast umap github python alternatives without heavy dependencies.

Why is it gaining traction?

It crushes umap-learn speeds by 30-46x on Apple hardware, embedding large datasets in under 3s where competitors take minutes. Pure MLX plus numpy means lightweight installs—no PyTorch or SciPy—and parameters like n_neighbors, min_dist mirror the original for seamless swaps. Benchmarks and animations highlight crisp umap plot github results, drawing devs exploring parametric umap github or understanding umap github tools.

Who should use this?

Data scientists on M-series Macs visualizing high-dim embeddings for exploratory analysis, like clustering Fashion-MNIST or gene expression data. ML engineers prototyping umap efi github flows or umap matlab github ports who hit CPU bottlenecks with umap-learn. Teams needing quick iterations on 10K-100K samples without cloud GPUs.

Verdict

Grab it if you're on Apple Silicon and need blazing UMAP—installs easily, API is familiar, results match expectations. With 28 stars and 1.0% credibility score, it's early-stage (solid README, no tests yet), so test on your data before production.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.