raphaelty

8 ms·

1.
Knowledge: You can just build your own AI feed to keep up, without the noise(github.com)
1 points by raphaelty 4mo ago|1 comments
2.
by raphaelty 4mo ago
Introducing a new version of knowledge which is my search engine / bookmark system in which I designed my own AI feed without the noise of X
3.
A semantic Code Search CLI but with grep filters(github.com)
1 points by raphaelty 7mo ago|1 comments
4.
by raphaelty 7mo ago
While semantic search shine at various tasks including advanced search engine and recommender systems, it did not shine that much on coding task so far appart from benchmarks available in papers. We think standard semantic code search tools
5.
Show HN: Hybrid Semantic Grep for Claude Code(github.com)
5 points by raphaelty 7mo ago|0 comments
6.
Multi-vector Grep for code agents, save 15% tokens, local(huggingface.co)
3 points by raphaelty 7mo ago|1 comments
7.
by raphaelty 9mo ago
Very interesting, wondering if there are other heavily used algorithm which could benefit a lot from a "Flash" version but don't have one today
8.
by raphaelty 1y ago
Very cool to spot such missing feature from Vision Language Models
9.
by raphaelty 1y ago
Very interesting, willing to try burn
10.
Information Retrieval, spawn Colbert in ms(lightonai.github.io)
1 points by raphaelty 1y ago|1 comments
11.
by raphaelty 1y ago
Blog introducing late-interaction models with max-similarity scoring function visualization. pylate-rs is a rust crate providing python and wasm bindings.
12.
by raphaelty 1y ago
Very interesting work! Starred the project. Would love to see such features integrated into the compiler itself Anyway, fully agree with you on the complementarity of ML and Rust
13.
Show HN: Rust Based Per-Token Late Interaction Dense Search(github.com)
8 points by raphaelty 1y ago|1 comments
14.
by raphaelty 1y ago
I created my personnal search engine to index and retrieve content I like. I like the idea of keeping track of content you like without too much effort. https://github.com/raphaelsty/knowledge I did not find yet a solu
15.
Torch.randperm isn't random and it can bias your trillion token-training run(twitter.com)
1 points by raphaelty 2y ago|0 comments
16.
Show HN: Search in HuggingFace Dataset from the URL(github.com)
2 points by raphaelty 2y ago|0 comments
17.
Show HN: DuckDB BM25 Client(github.com)
3 points by raphaelty 2y ago|0 comments
18.
Show HN: Late Interaction Retrieval(github.com)
1 points by raphaelty 2y ago|0 comments
19.
Show HN: FlashText with Rust for Python(github.com)
6 points by raphaelty 2y ago|0 comments
20.
by raphaelty 2y ago
Thank you for this ! Knowledge is primarly inspired by http://www.semanlink.net/sl/home
21.
by raphaelty 2y ago
Author here, I like the idea to have a public and personal search-engine. I'd love to deep dive in some people bookmarks such as expert in NLP field in order to get clue / a point of view on problem I want to solve but those infor
22.
by raphaelty 2y ago
Amazed by 3d-force-graph js library
23.
by raphaelty 2y ago
It upload content some of my friends like (github stars) and they have very good taste, so when I'm looking for a document / information on a new topic in machine learning or databases I like to give it a chance
24.
by raphaelty 2y ago
I did develop this when I was PhD student in NLP. I think it's fine to learn to develop a personal search system if you want to learn.
25.
by raphaelty 2y ago
The ask button is provided by chatgpt api here
26.
Show HN: Personal Knowledge Base Visualization(github.com)
150 points by raphaelty 2y ago|24 comments
27.
Show HN: Accelerate ColBERT with tree-based index(github.com)
4 points by raphaelty 3y ago|0 comments
28.
by raphaelty 3y ago
The library is clean and well documented
29.
by raphaelty 3y ago
You could recommend content based on user query, tag content produced by the user, use colbert as part of a ChatBot to show evidences to the user questions
30.
by raphaelty 3y ago
Yes, ColBERT is fast because you can pre-compute most embeddings. It's important to compute documents embeddings only once. neural-cherche do not compute embeddings on the fly and the retrieve method ask for queries and documents embed

More ›