mlsql
AI
natural language to sql
SHOW HN2019
How HN took it
3 positive, 17 neutral or asking, 0 mixed, 9 critical, out of 29 comments about the project. Thread score -0.20.
Idea: 2 for, 5 againstExecution: 0 for, 2 againstPerformance: 0 for, 2 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
Interesting open source example of natural language to SQL work.
From 3 comments
It's great to see a nice open source example of work in this area. It would be helpful if there were also some examples of where the model fails either by not being able to…
michaelmior on HNVery interesting. I did a similar project for my machine learning class in grad school. I'm curious what kind of architecture/model framework you used for this.
o10449366 on HNdoes it only work with a question that returns a single answer or it can be used to build something like…
c8g on HN
Objections
Requires too many resources (4 GB Docker image).
From 2 comments
Nice. Very interesting. But 4 gigs just to get I running. That's probably more than the DB I'm trying to query.
amolo on HNYes, I don’t know anything about how/what this is running, but I immediately can’t use it based on requirements. Too bad.
jtbayly on HN
Natural language is too imprecise compared to explicit SQL.
From 2 comments
I might be too much of a control freak to give up the explicit precision of SQL.
rezeroed on HNI think anyone who uses sql would find this tool (and a lot of similar abstractions) too imprecise, verbose, and error prone. What's the longest bridge in mariland - or was it…
sheeshkebab on HN
Complex real-world queries are hard to express in English.
From 1 comment
I feel that SQL queries that people will need help with are not the SQL 101 examples shown here. Looking at the queries I write as a product manager, I wouldn't have the…
ohadron on HN
Requests
Show examples of where the model fails.
From 1 comment
It's great to see a nice open source example of work in this area. It would be helpful if there were also some examples of where the model fails either by not being able to…
michaelmior on HN
Provide a demo environment.
From 1 comment
From a product managers pov this looks very interesting - is there a demo environment available somewhere to play with this?
monkeydust on HN
Questions
What architecture or model framework was used?
From 1 comment
Very interesting. I did a similar project for my machine learning class in grad school. I'm curious what kind of architecture/model framework you used for this.
o10449366 on HN
How does it handle case-sensitive string matching?
From 2 comments
The example is probably wrong or not working since the English questions uses "throgs neck" in lower case instead of the captialized "Throgs Neck" version. Those are two…
LunaSea on HNAren't the queries case insensitive by default?
laurent123456 on HN
Alternatives named
Microsoft SQL Server English Query
From 1 comment
Ah, this brings back fun memories. Over a decade ago, Microsoft SQL Server had English Query to do the same thing. Developers had to define all the words & relationships that…
BrentOzar on HN
SQLova (naver/sqlova)
From 1 comment
This is a wrapper for a pretrained model. Probably a better github link would be https://github.com/naver/sqlova/
mynegation on HN
ThoughtSpot
From 1 comment
does it only work with a question that returns a single answer or it can be used to build something like…
c8g on HN