• Hacker News
  • new|
  • comments|
  • show|
  • ask|
  • jobs|
  • spoaceman7777 15 hours

    Well. That is insane. O_O Fantastic job!

  • mskkm 5 hours

    There are already several openreview comments alleging academic misconduct around TurboQuant: https://openreview.net/forum?id=tO3ASKZlok

    Some write-ups argue that this was deliberate rather than a good-faith mistake: https://dev.to/gaoj0017/turboquant-and-rabitq-what-the-publi...

    And now this. Pretty bold AI slop.

  • esafak 16 hours

    lancedb and duckdb integrations would be great...

  • lmeyerov 7 hours

    Interestingly, while we don't fine-tune generative models for Louie.ai, we found fine-tuning embedding models to be a major $ saver. Instead of 1K-2K wide frontier embedding vector lens... Just 64. Huge savings on vector DB $$$.

    I'm curious how that works with something like turboquant. Not needed any more, still dominant, better together, ... .

  • sp1982 16 hours

    If anyone is looking to retrofit to an existing pipeline, I use similar ideas to compress vectors for job search, getting roughly 8x compression with about a 3.5% drop in quality. My experiment: https://corvi.careers/blog/vector-search-embedding-compressi...

  • OutOfHere 11 hours

    I am not convinced that Turbovec yields better retrieval than the same amount of bits of a Matryoshka embedding.

  • cat-whisperer 11 hours

    What's a good embedding model and search to run locally? something fast and lightweight.

  • tracespect 13 hours

    [flagged]

  • cute_boi 13 hours

    Another vibe coded slop where they can't even spend time on Readme or documentation around code...

  • refulgentis 14 hours

    Bloviating nonsense, 3rd time I’ve seen something like this in HN since TurboQuant came out. You don’t need float32, never did. Source: I’ve been writing on device embedding code for 4 years.

  • anthropic-dario 5 hours

    [dead]

  • myshapeprotocol 11 hours

    [dead]

  • bobmarleybiceps 13 hours

    people should read turboquant's open review comments: https://openreview.net/forum?id=tO3ASKZlok

    esafak 10 hours

    tl,dr: there is an allegedly better alternative, and it's already implemented everywhere: https://github.com/VectorDB-NTU/RaBitQ-Library#rabitq-in-ind...

  • ghm2199 16 hours

    Wow! 4GB for 10 million documents. This means one could build a reverse index much faster than before and devx processes like debugging, performance testing would become much smoother. Can't wait for the sqlite bindings to come out!

    ghm2199 16 hours

    Also the removal latency is on a log scale. Which is quite insane.

  • Eridrus 13 hours

    FAISS is no longer close to SoTA:

    https://ann-benchmarks.com/index.html https://vector-index-bench.github.io/ https://big-ann-benchmarks.com/neurips23.html

    ehsanu1 8 hours

    Surprised that usearch isn't in any of these, it's pretty fast.

    nl 11 hours

    I think their point is the size/performance tradeoff rather than outright performance. The point of TurboQuant is the size savings, while still giving high accuracy.

    It's been a while, but I do recall some high-performing vector matching indexes being very large.

  • burgerboii 16 hours

    Who is this co-author called t <t@t>?

    cute_boi 13 hours

    As it is heavily vibe coded, I think member of technical staff at antropic has no clue....

    Next Prompt: remove t@t and force commit.

  • zuzululu 16 hours

    what could i use this for as part of my agentic workflow? codebase indexing? docs ?

    kyxsc 16 hours

    notes/docs/wiki is a great use case

  • beernet 14 hours

    Why not just use Qdrant? They've been integrating TurboQuant for months, works well.

    kanungle 8 hours

    Integrated in 5 weeks and just expanded data types for turbo4 in last release. No longer need to store fp32 vectors if you don't need them

  • anishvarghese 16 hours

    This looks perfect for local, privacy first search, but since it's built in Rust, has anyone tried compiling it to WASM to run directly inside a browser extension?

    westurner 15 hours

    oxirs does embeddings and GraphRAG, and full text search with Tantivy; oxirs-vec, oxirs-graphrag

    There's an oxirs-wasm with RDF and SPARQL bindings with a query budget. Tantivy-wasm says that the release WASM bundle is 1.5 MB.

    cool-japan/oxirs: https://github.com/cool-japan/oxirs

    oxirs-wasm: https://crates.io/crates/oxirs-wasm

    tantivy-wasm: https://github.com/phiresky/tantivy-wasm

    Is there an advantage to adding an MCP local memory interface over agent instructions on how to use a rust CLI?

    And then write Markdown documents with Google OKF-like frontmatter YAML metadata for agents that work with tokens not linked data graphs; https://github.com/GoogleCloudPlatform/knowledge-catalog/blo...

    cpursley 15 hours

    Also interested.

    coredog64 14 hours

    Can WASM use AVX512-VNNI?

    LtdJorge 13 hours

    No, WASM only has 128b SIMD instructions, for now.

  • nharada 16 hours

    It would be nice to have the README be a little more human written for a project where you actually want people to adopt it

    badatnames 15 hours

    Anthropic employee. This is what your brain on kool aid looks like

    deeviant 15 hours

    Then again, if the only thing the human doing is bitching about AI use, it's not really that comparatively useful.

    righthand 11 hours

    Sure it is useful, the bitching is canary in the shit software mine. How do you know the software isnt shit if the Readme is shit?