Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Viewing profile — anotherCodder

anotherCodder

HN member
Joined
Fri, Sep 06, 2024, 12:33 PM UTC
HN karma
18
Public activity
27 items

About anotherCodder

Solo engineer. I build memra, an LLM inference engine in Rust and CUDA for RTX Blackwell: https://github.com/avifenesh/memra

I also run a hosted inference API on it: https://inference.tiyuvta.ai

Recent public activity

  1. comment
    Comment #49403507

    [flagged]

  2. comment
    Comment #49403502

    most of the time when a local model feels dumb its not the quant, its the chat template. a lot of gguf mints just drop the template from the metadata and the runtime silently falls…

  3. story
  4. comment
  5. comment
    Comment #49328778

    price update: $0.40 in / $0.10 cache / $2.90 out

  6. comment
  7. story
  8. comment
  9. comment
    Comment #49315005

    vllm on 4x 5090 is getting ~20 tok/s with mtp on (their own thread on the hf card). i had qwen3.8-27b up the day after release, one rtx pro 6000, 140 tok/s spec, 0.156s first token…

  10. comment
  11. story
  12. story
  13. story
  14. comment
    Comment #48291256

    [flagged]

  15. story
  16. story
    Computer-Use-Linux

    Hi, Since apps like Codex and Claude don't provide support for computer use for Linux desktop users, iv'e built my own own one. https://github.com/agent-sh/computer-use-linux `npm …

  17. story
    Glide-mq high performance, features rich, message queue based on valkey glide

    https://avifenesh.github.io/glide-mq.dev/ https://github.com/avifenesh/glide-mq

  18. story
    Show HN: I maintain Valkey GLIDE – built a Node queue doing 48k jobs/s

    Hey HN, I'm a maintainer of Valkey GLIDE (the official Valkey client) and work on the AWS ElastiCache team. Day job is deep in database client internals — which gives me a clear vi…

  19. comment
    Comment #46983893

    Hey HN - I built agnix because I kept losing time to the same class of bug: AI tool configs that are almost right but silently wrong. The trigger: I had a Claude Code skill named `…

  20. story
  21. story
    Show HN: My way – 18-agent autonomous workflow for ClaudeCode – issues to deploy

    Built a plugin that runs my entire dev cycle autonomously after I approve the plan. /next-task pulls from GitHub issues, scans the codebase, ranks tasks, I pick one and approve the…

  22. comment
    Comment #42301389

    For users of Node.js, Java, Python and very soon Go - Valkey-Glide: https://github.com/valkey-io/valkey-glide It will stay out of the hand of Redis, can promise that, and if you us…

  23. story
  24. comment
    Comment #41478821

    Want to understand what dev's will appreciate and will lead decision makers to choose our client. What will make you choose a client library over the other, and what will make cons…

  25. story