teamchong

6 ms·

1.
by teamchong 3mo ago
built this to stop fable blowing up my usage limit https://github.com/teamchong/pxpipe
2.
Show HN: TweetDeck-style HN reader, built to experiment with Chrome's Prompt API(hndeck.teamchong.net)
2 points by teamchong 4mo ago|0 comments
3.
by teamchong 5mo ago
sorry it’s not working for you. I built this as a personal project for self-learning, but I plan to take a look at this issue next weekend. you can check out a video demo of it here: https://github.com/user-attachments/
4.
by teamchong 5mo ago
firefox has webgpu already, but the subgroups extension isn't in yet. every matmul / softmax kernel here leans on subgroupShuffleXor for reductions, that's the blocker. same reason mlc webllm and friends don't run on fir
5.
Show HN: Prompt-to-Excalidraw demo with Gemma 4 E2B in the browser (3.1GB)(teamchong.github.io)
163 points by teamchong 5mo ago|62 comments
6.
by teamchong 5mo ago
I made some adjustment, can you try again? Is it faster now? https://teamchong.github.io/turboquant-wasm/search.html
7.
by teamchong 6mo ago
you’re right that 32f is faster on raw query time, quantization adds extra step. main benefit on download size since gzip won’t help much, which matters most in browser contexts
8.
Show HN: TurboQuant-WASM – Google's vector quantization in the browser(github.com)
165 points by teamchong 6mo ago|7 comments
9.
by teamchong 7mo ago
I built this after hitting GC stalls parsing streaming tool calls in my AI agent. LLM outputs are getting large — code edits, file writes, 50-200KB JSON payloads. Every AI SDK I looked at (Vercel, Anthropic, TanStack, OpenClaw) does `buffer
10.
Show HN: VectorJSON – O(n) streaming parser to handle LLM JSON outputs(github.com)
1 points by teamchong 7mo ago|1 comments