itrummer

5 ms·

1.
Try AI Operators on PostgreSQL(samtsql.com)
2 points by itrummer 3mo ago|0 comments
2.
by itrummer 11mo ago
The approximation in ThalamusDB is relative to the best accuracy that can be achieved using the associated language models (LLMs). E.g., if ThalamusDB processes a subset of rows using LLMs, it can reason about possible results when applying
3.
by itrummer 11mo ago
:-) Actually, you can also use models of other providers (e.g., Google's Gemini models). You just have to set your access key by the corresponding provider and configure the models you'd like to use in this file: https:/&#x2
4.
by itrummer 11mo ago
I think the previous post refers to a different project. But yes: ThalamusDB can process all rows if necessary, including matching all images that have the same persons in them.
5.
by itrummer 11mo ago
We use mocking to replace actual LLM calls when testing for the correctness of the ThalamusDB code. In terms of performance benchmarking, we ran quite a few experiments measuring time, costs (fees for LLM calls), and result accuracy. The la
6.
by itrummer 11mo ago
Well, it definitely goes beyond a traditional DBMS, but yes :-) If processing the same amount of data via pure SQL versus SQL with LLM calls, it will be slower and more expensive when using LLMs. Note that 600s is just the default timeout,
7.
by itrummer 11mo ago
LlamaIndex relies heavily on RAG-style approaches, e.g., we're using items whose embedding vectors are close to the embedding vectors of the question (what you describe). RAG-style approaches work great if the answer depends only on a
8.
by itrummer 11mo ago
Thank you :-)
9.
Show HN: SQL with AI Operators on Text, Images, and Sound Files(github.com)
1 points by itrummer 11mo ago|0 comments
10.
ThalamusDB: Query text, tables, images, and audio(github.com)
53 points by itrummer 1y ago|18 comments
11.
From Data to Reports – Instantly with AI(speedylytics.com)
1 points by itrummer 1y ago|0 comments
12.
SQL processing over images, text, and audio(itrummer.github.io)
3 points by itrummer 1y ago|0 comments
13.
by itrummer 2y ago
Thanks for the feedback! No, it wasn't written by an AI writing assistant. :-) I think the problem is the following: until you buy the book, you'll only see every fourth page or so in the chapter preview. Not my choice, that'
14.
by itrummer 2y ago
I've been giving a few tutorials on large language models for data scientists over the past years and ultimately decided to turn it into a book, now available as an early-access version. It's a gentle introduction to analyzing var
15.
Writing a Book: "AI-Assisted Data Science"(manning.com)
2 points by itrummer 2y ago|3 comments