News for hackers. Stuff that matters.
Re:A hyper-fast local vector database for use with LLM Agents by superkuh on Sunday April 16, 2023 @10:42PM (#35596313)
A "DB performance vs VC funding raised" plot is an odd choice for a github repo dealing with code.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by theolivenbaum on Sunday April 16, 2023 @10:44PM (#35596327)
I think that was perfectly on point if you reviewed the code.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by brigadier132 on Sunday April 16, 2023 @10:43PM (#35596323)
Absolutely hilarious. You're ready to raise $18 million.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by bugglebeetle on Sunday April 16, 2023 @10:47PM (#35596340)
Perfect joke example of an AI MVP because you pay OpenAI to create the embeddings.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by egeozcan on Sunday April 16, 2023 @10:56PM (#35596380)
https://github.com/jdagdelen/hyperDB/blob/main/hyperdb/galax...

This was a great laugh. Also, after the benchmark, it says:

> Benchmark Credit: Benim Kıçım

"Benim Kıçım" means "my ass" in Turkish.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by kristianp on Monday April 17, 2023 @02:04AM (#35597286)
References "[Submitted on 31 Mar 2023] Programmed but Arbitrary Control Minimization of Amplitude and phase for speckle Nulling (PACMAN)" https://arxiv.org/abs/2303.18231

No idea if that's a joke paper.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by 0x008 on Monday April 17, 2023 @02:28AM (#35597404)
> HyperSVMRanking (Such Vector, Much Ranking) algorithm proposed by Andrej Karpathy (2023) https://arxiv.org/abs/2303.18231
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by shreyansh26 on Sunday April 16, 2023 @11:00PM (#35596414)
Haha, so all one has to do is to add some fancy terms like LLMs, vectors, hyper-fast and 300+ Github stars. What a a joke lol!
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by orasis on Sunday April 16, 2023 @11:02PM (#35596430)
It’s a prank.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by jazzido on Sunday April 16, 2023 @11:22PM (#35596529)
`def derridaean_similarity`

Instant classic.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by celestialcheese on Sunday April 16, 2023 @11:25PM (#35596553)
Damn this is good - well played.

For those who aren't neck deep in LLM hype - ChromaDB raised $18m at a $75m valuation a couple weeks ago with what's essentially just a thin wrapper around duckdb with a parquet import/export for persistence.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by elforce002 on Monday April 17, 2023 @07:14AM (#35599172)
Hahaha. Really?! Good for them.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by merpnderp on Sunday April 16, 2023 @11:28PM (#35596567)
I couldn't believe they were raising $35 million rounds with a brand new embedding database that was open source. Then I poked through the dozens of lines of code and realized they had implemented the same DB I had.

LOL, much funny.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by alfor on Sunday April 16, 2023 @11:32PM (#35596586)
It’s a joke, but is it a functional one?

I found it crazy to run auto-GPT with an external DB over the internet, how can this be ever fast and efficient? (unless the dataset is humongous)

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by gavi on Tuesday April 18, 2023 @02:27PM (#35617870)
Yes, it used OpenAI embeddings and numpy to find cosine similarity
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by yawnxyz on Sunday April 16, 2023 @11:55PM (#35596695)
aside from making fun of Chroma, this seems to 100% work though? What would be the pros/cons of actually using this instead?
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by jdaggers on Monday April 17, 2023 @11:33AM (#35602296)
Author here. Yes, it actually does work. If I’m ever indexing fewer than ~100k documents/paragraphs I would use something like this approach rather than reaching for a vector database.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by gavi on Tuesday April 18, 2023 @02:27PM (#35617883)
Great work!
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by ShamelessC on Monday April 17, 2023 @12:08AM (#35596754)
Damn - did this get flagged? Went from top-of-HN to missing from the front page _pretty_ quick.
[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by jdaggers on Monday April 17, 2023 @12:53AM (#35596970)
Yeah, this was a little strange. I'm the author of the repo linked here.

I got alerted to it hitting #1 on HN and was able to see it there for only a little bit before it immediately got tanked to # 60. I'm a little bummed, but c'est la vie.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by ShamelessC on Monday April 17, 2023 @01:18AM (#35597097)
Certainly suspicious - almost like VC's can't take a joke ;) I'm sure there's an innocent explanation though.

Great satire in any case! Strikes the perfect balance between highly plausible and glaringly obvious.

[ Parent | on HN ]
Re:A hyper-fast local vector database for use with LLM Agents by arthurcolle on Monday April 17, 2023 @12:09AM (#35596763)
https://raw.githubusercontent.com/jdagdelen/hyperDB/main/_st...

Y-axis on this chart is actually hilarious

[ Parent | on HN ]