astronautas

7 ms·

31.
by astronautas 2y ago
Not a catalog though, still need to input s3 path...
32.
by astronautas 2y ago
This
33.
by astronautas 2y ago
I am confused... It's a cool example, but really an antipattern. Nowadays everyone gets analysts want access to raw data, since they know which aggregations they need best, whereas data engineers stay away from pre-aggregating and focu
34.
Let's build a Bloom filter. Part 1(astronautas.github.io)
1 points by astronautas 3y ago|0 comments
35.
by astronautas 3y ago
I've been recently toying around with dynamically switching between DuckDB and BigQuery on query time. So you have a layer of standard SQL in front, which uses sqlglot to translate it to compute-specific SQL. So e.g. you can use DuckDB
36.
by astronautas 3y ago
True!
37.
by astronautas 3y ago
Cool, thanks for sharing!
38.
by astronautas 3y ago
Write a blog post then about this! I can tell you it is hardly a solved problem.
39.
by astronautas 3y ago
eh, awesome! Seems this one, right? https://github.com/galeone/tfgo . Quite many stars.
40.
by astronautas 3y ago
Wow, make it open source quickly!!! :hype:. It's a classic Python REST API for model serving. But we have very low latency constraints. As such, rewriting in more high performant backend languages e.g. Go or Rust would substantially re
41.
by astronautas 3y ago
Is this the same as Nvidia's Triton?
42.
by astronautas 3y ago
Make it for Go, and I am sold. Running ML models in Go services is still an unsolved problem.
43.
CoreWeave raises $2.3B in debt collateralized by Nvidia chips(reuters.com)
2 points by astronautas 3y ago|0 comments
44.
by astronautas 3y ago
curious to hear your thoughts
45.
A warehouse table might just be enough as a feature store(astronautas.github.io)
2 points by astronautas 3y ago|1 comments
46.
by astronautas 4y ago
https://www.bloomberg.com/news/articles/2022-05-28/lithuania... On Bloomberg.
47.
Lithuanians Fundraise Bayraktar Drone for Ukraine in 3 Days(reuters.com)
5 points by astronautas 4y ago|1 comments