Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ainesh93
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
ainesh93
3y ago
Hard to claim success with "complex" questions if you don't account for business context and organizational nuances. For example, "active" listings on Redfin may be a combination of days on market, last open house,
2.
▲
by
ainesh93
3y ago
> You also need a bunch of information about the real business that the data is describing. While the article focuses on finetuning GPT-3.5-turbo, how you use the text-to-SQL engine within the architecture of your overall solution is for
3.
▲
by
ainesh93
3y ago
I think the focus here isn't necessarily on compute cost. When companies hire data scientists or analysts, they're niche-skilled and expensive. If those people spend 50-60% of their time courting ad-hoc questions from various peop
4.
▲
by
ainesh93
3y ago
Although Spider is better known in the text-to-SQL world, you're right that BiRD may provide a better testing ground. Comparing against the current leaderboard on that standard is on the docket!
5.
▲
by
ainesh93
3y ago
Hi, you can find updated documentation on connecting to BigQuery here: https://dataherald.readthedocs.io/en/latest/api.database.htm... . We have also updated the ReadMe.
6.
▲
by
ainesh93
3y ago
Yes, opening the connection to the DB read-only would also work. That's what we're planning on doing.
7.
▲
by
ainesh93
3y ago
Hit the nail on the head! Not only is the context length a limitation, but the speed of response gets impacted as well. With a human in the loop, even providing a "mostly" correct SQL that takes a swing at the correct joins betwee
8.
▲
Text-to-SQL Benchmarks and the Current State-of-the-Art
(medium.com)
2 points
by
ainesh93
3y ago
|
1 comments
9.
▲
by
ainesh93
3y ago
Medium article presenting background on the most popular text-to-SQL benchmark datasets and current performance of text-to-SQL algorithms
10.
▲
by
ainesh93
4y ago
What a quick way to visualize and get the most telling insights from cool public data sources! I've always known this data is out there, but now I have a medium to understand what it's meant to show. Super cool tool!