frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Show HN: I visualized the entire history of Citi Bike in the browser

https://bikemap.nyc/
20•freemanjiang•3h ago•8 comments

Show HN: An LLM response cache that's aware of dynamic data

https://blog.butter.dev/on-automatic-template-induction-for-response-caching
7•raymondtana•1h ago•0 comments

Show HN: I built a "Do not disturb" Device for my home office

https://apoorv.page/blogs/over-engineered-dnd
66•quacky_batak•4d ago•33 comments

Show HN: A to Z – A word game I built from a childhood road trip memory

https://a26z.fun/
7•jackhulbert•3h ago•2 comments

Show HN: Free and local browser tool for designing gear models for 3D printing

https://gears.dmtrkovalenko.dev
35•neogoose•14h ago•9 comments

Show HN: Flatagents: State machine orchestration with stateless LLM agents

https://github.com/memgrafter/flatagents
2•beacon294•1h ago•1 comments

Show HN: KeelTest – AI-driven VS Code unit test generator with bug discovery

https://keelcode.dev/keeltest
25•bulba4aur•9h ago•10 comments

Show HN: Tool for Testing MCP Servers

https://www.mcp-workbench.ai/
2•opiniateddev•1h ago•0 comments

Show HN: SMTP Tunnel – A SOCKS5 proxy disguised as email traffic to bypass DPI

https://github.com/x011/smtp-tunnel-proxy
122•lobito25•22h ago•40 comments

Show HN: bikemap.nyc – visualization of the entire history of Citi Bike

https://github.com/freeman-jiang/bikemap.nyc
6•freemanjiang•2h ago•3 comments

Show HN: Grammar of Graphics CLI tool made in Rust

https://github.com/williamcotton/gramgraph
2•williamcotton•2h ago•0 comments

Show HN: Comet MCP – Give Claude Code a browser that can click

https://github.com/hanzili/comet-mcp
25•hanzili•3d ago•25 comments

Show HN: VaultSandbox – Test your real MailGun/SES/etc. integration

https://vaultsandbox.com/
52•vaultsandbox•1d ago•9 comments

Show HN: Mantic.sh – A structural code search engine for AI agents

https://github.com/marcoaapfortes/Mantic.sh
73•marcoaapfortes•1d ago•34 comments

Show HN: 48-digit prime numbers every git commit

https://textonly.github.io/git-prime/
65•keepamovin•6d ago•52 comments

Show HN: Tylax – A bidirectional LaTeX to Typst converter in Rust

https://github.com/scipenai/tylax
16•democat•7h ago•2 comments

Show HN: Seapie – a Python debugger where breakpoints drop into a REPL

https://github.com/hirsimaki-markus/seapie
7•markushirsimaki•4h ago•1 comments

Show HN: The kissing number theorem predicts particle masses from sphere packing

https://colab.research.google.com/drive/1_zDIOONfs4WvnpG7GDEH6hzSM25Fsu93?usp=sharing
2•AlekseN•1h ago•3 comments

Show HN: Make audio loops online

https://makeloops.online/
66•bilalba•2d ago•22 comments

Show HN: Prism.Tools – Free and privacy-focused developer utilities

https://blgardner.github.io/prism.tools/
364•BLGardner•1d ago•99 comments

Show HN: Tailsnitch – A security auditor for Tailscale

https://github.com/Adversis/tailsnitch
271•thesubtlety•2d ago•28 comments

Show HN: Stash – Sync Markdown Files with Apple Notes via CLI

https://github.com/shakedlokits/stash
69•shuka•1d ago•21 comments

Show HN: I built a 3D World Map and multiplayer geography game using Three.js

https://www.mixora.xyz
2•qwrwenm•6h ago•0 comments

Show HN: DoNotNotify – Log and intelligently block notifications on Android

https://donotnotify.com/
339•awaaz•2d ago•163 comments

Show HN: llmgame.ai – The Wikipedia Game but with LLMs

https://www.llmgame.ai
24•jmcallister•1d ago•22 comments

Show HN: Jax-JS, array library in JavaScript targeting WebGPU

https://ss.ekzhang.com/p/jax-js-an-ml-library-for-the-web
79•ekzhang•1d ago•21 comments

Show HN: KektorDB – Lightweight, Embeddable Vector+Graph Database Written in Go

https://github.com/sanonone/kektordb
2•san0n•7h ago•1 comments

Show HN: Foundertrace – chain of YC startups founded by its employees

https://foundertrace.com/
38•loondri•3d ago•13 comments

Show HN: GPU Cuckoo Filter – faster queries than Blocked Bloom, with deletion

https://github.com/tdortman/cuckoo-filter
31•tdortman•1d ago•4 comments

Show HN: Can you hit replacement? A fertility SIM with cited sources

https://www.tfrsim.com/
3•joshuafkon•8h ago•0 comments
Open in hackernews

Show HN: Jax-JS, array library in JavaScript targeting WebGPU

https://ss.ekzhang.com/p/jax-js-an-ml-library-for-the-web
79•ekzhang•1d ago

Comments

esafak•1d ago
What is the state of web ML? Anybody doing cool things already? How about https://www.w3.org/TR/webnn/ ?
sroussey•1d ago
onnx on the web has the most models available and can use webgpu which is available everywhere.

Huggingface’s transformers.js uses it. And I use that for https://workglow.dev (also tensorflow mediapipe though that is using wasm).

I don’t think webnn has gone anywhere and is too restrictive.

ekzhang•1d ago
Since ONNX is just a model data format, you can actually parse and run ONNX files in jax-js as well. Here’s an example of running DETR ResNet-50 from Xenova’s transformers.js checkpoint in jax-js

https://jax-js.com/detr-resnet-50

I don’t think I intend to support everything in ONNX right now, especially quant/dequant, but eventually it would be interesting to see if we can help accelerate transformers.js with a jax-js backend + goodies like kernel fusion

jax-js is more trying to explore being an ML research library, rather than ONNX which is a runtime for exported models

mlajtos•1d ago
I have a project using tfjs and jax-js is very exciting alternative. However during porting I struggle a lot with `.ref` and `.dispose()` API. Coming from tfjs where you garbage collect with `tf.tidy(() => { ... })`, API in jax-js seems very low-level and error-prone. Is that something that can be improved or is it inherent to how jax-js works?

Would `using`[0] help here?

[0]: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Refe...

ekzhang•1d ago
I don’t think tf.tidy() is a sound API under jvp/grad transformations, also it prevents you from using async which makes it incompatible with GPU backends (or blocks the page), a pretty big issue. https://github.com/tensorflow/tfjs/issues/5468

Thanks for the feedback though, just explaining how we arrived at this API. I hope you’d at least try it out — hopefully you will see when developing that the refs are more flexible than alternatives.

mlajtos•16h ago
I'll grind jax-js more and see if refs become invisible then. Thanks for a great project!
yuppiemephisto•1d ago
This project is an inspiration, I've been working on porting tinygrad to [Lean](github.com/alok/tinygrad)
sestep•1d ago
Hey Eric, great to see you've now published this! I know we chatted about this briefly last year, but it would be awesome to see how the performance of jax-js compares against that of other autodiff tools on a broader and more standard set of benchmarks: https://github.com/gradbench/gradbench
ekzhang•1d ago
For sure! It looks like this is benchmarking the autodiff cpu time, not the actual kernels though, which (correct me if I’m wrong) isn’t really relevant for an ML library — it’s more for if you have a really complex scientific expression
sestep•1d ago
Nope, both are measured! In fact, the time to do the autodiff transformation isn't even reflected in the charts shown on the README and the website; those charts only show the time to actually run the computations.
ekzhang•1d ago
Hm okay, seems like an interesting set of benchmarks — let me know if there’s anything I can do to help make jax-js more compatible with your docker setup
sestep•1d ago
It should be fairly straightforward; feel free to open a PR following the instructions in CONTRIBUTING.md :)
ekzhang•23h ago
I don’t think this is straightforward but it may be a skill issue on my part. It would require dockerizing headless Chrome with WebGPU support and dynamically injecting custom bundled JavaScript into the page, then extracting the results with Chrome IPC
sestep•21h ago
Ahh no you're right, I forgot about the difficulties for GPU specifically; apologies for my overly curt earlier message. More accurately: I think this is definitely possible (Troels and I have talked a bit about this previously) and I'd be happy to work together if this is something you're interested in. I probably won't work on this if you're not interested on your end, though.
bobajeff•1d ago
This is really great. I don't do ML stuff. But I some mathy things that would benefit from running in the GPU so it's great to see the Web getting this.

I hope this will help grow the js science community.

maelito•1d ago
Could not run the demos on Firefox. On Chromium, the Great Expectations loads but then nothing happens.
ekzhang•1d ago
Firefox doesn’t support WebGPU yet, you can run programs in the REPL through other backends like Wasm/WebGL: https://jax-js.com/repl

See: https://caniuse.com/webgpu

forgotpwd16•15h ago
According to page WebGPU supported (`dom.webgpu.enabled` flag) but is only enabled by default on Windows & macOS (i.e. not Linux).
fouronnes3•1d ago
Congrats on the launch! This is a very exciting project because the only decent autodiff implementation in typescript was tensorflowjs, which has been completely abandonned by Google. Everyone uses onnx runtime web for inference but actually computing gradients in typescript was surprisingly absent from the ecosystem since tfjs died.

I will be following this project closely! Best of luck Eric! Do you have plans to keep working on it for sometime? Is it a side project or will you abe ble to commit to jax-js longer term?

ekzhang•1d ago
Yes, we are actively working on it! The goal is to be a full ML research library, not just a model inference runtime. You can join the Discord to follow along
forgotpwd16•15h ago
Very nice work. Like how it supports webgpu but also cpu/wasm/webgl. Would love to read more on the internals & design choices made like e.g. ref counting in README.

P.S. And thanks for taking your time working on this and releasing something polished rather a Claude slop made within few days as seems to be the norm now.