Wordllama
AI
embedding models
SHOW HN2024
How HN took it
3 positive, 5 neutral or asking, 1 mixed, 1 critical, out of 10 comments about the project. Thread score +0.17.
Idea: 2 for, 0 againstPerformance: 0 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Tiny model size is an advantage over SBERT smallest models
From 1 comment
Nice. I like the tiny size a lot, that's already an advantage over SBERTs smallest models. But it seems quite dated technically - which I understand is a tradeoff for performance…
authorfly on HN
Useful for game making
From 1 comment
This is great for game making! Thank you!
ttpphd on HN
Objections
Benchmarks clearly worse than MiniLM models
From 2 comments
Its a 17 MB model that benchmarks obviously worse than MiniLM v2 (which is SBERT). I run V3 on ONNX on every platform you can think of with a 23 MB model. I don't intend for that…
refulgentis on HNLooks cool! Any advantages to the mini-lm model - it seems better on most mteb tasks but wondering if maybe inference or something is better.
dspoka on HN
Dated technically compared to LLM-based embeddings
From 1 comment
Nice. I like the tiny size a lot, that's already an advantage over SBERTs smallest models. But it seems quite dated technically - which I understand is a tradeoff for performance…
authorfly on HN
Requests
Support for multiple similarity types such as semantic, NLI and noun-abstract
From 1 comment
Nice. I like the tiny size a lot, that's already an advantage over SBERTs smallest models. But it seems quite dated technically - which I understand is a tradeoff for performance…
authorfly on HN
Multilingual or French language support
From 2 comments
Any plan for languages other than english? This would be a perfect tool for french language.
jcmeyrignac on HNVery interested in a multilingual version too! FYI huggingface hosts datasets too. And wikipedia has a nice portal for datasets :…
Ey7NFZ3P0nzAe on HN
Questions
Whether embeddings can encode opposite meanings like burning and freezing
From 2 comments
I assume that by "ChatGPT embeddings" you mean OpenAI embedding models. In that case, "burning" and "freezing" are not opposite at all, with a cosine similarity of 0.46 (running…
curl-up on HNThis might be a dumb question but... if I get the embeddings of words with a common theme like "burning", "warm", "cool", "freezing", would I be able to relatively well fit an…
magicalhippo on HN
Alternatives named
MiniLM / SBERT models
From 2 comments
Looks cool! Any advantages to the mini-lm model - it seems better on most mteb tasks but wondering if maybe inference or something is better.
dspoka on HNIts a 17 MB model that benchmarks obviously worse than MiniLM v2 (which is SBERT). I run V3 on ONNX on every platform you can think of with a 23 MB model. I don't intend for that…
refulgentis on HN
OpenAI embedding models
From 1 comment
I assume that by "ChatGPT embeddings" you mean OpenAI embedding models. In that case, "burning" and "freezing" are not opposite at all, with a cosine similarity of 0.46 (running…
curl-up on HN
TF-IDF with kmeans
From 1 comment
With a large corpus (10,000+ sentences--each sentence is a "document" in my usecase) I can get similar results by kmeans clustering TF-IDF spmatrix vectors but it looks like this…
xk3 on HN