thecalebf

2 tahun yang laludiscuss

the graphics in it make it so much better

2 tahun yang laludiscuss

The Chrome Extension is open source and is powered by DuckDB WASM which loads the parquet conversion of the datasets as views so you can query them with SQL.

Here's the repo: https://github.com/cfahlgren1/hf-data-explorer

thecalebf

2 tahun yang laludiscuss

Yes please do! Looks awesome, would love to help any way I can as well.

thecalebf

2 tahun yang laludiscuss

Sorry, it seems complicated. Since it is a finetune of Deepseek-Coder, I had to include their license.

Deepseek is pretty open, just says not to use it for: - military purposes - exploiting vulnerabilities etc.

Just trying to include as much information as possible from the initial base model.

thecalebf

2 tahun yang laludiscuss

Yes it is designed for users without SQL knowledge, however, it can still perform fairly well with questions on the difficult side (for non technical users) with queries having multiple joins, aggregations, ratios, and subqueries.

Next step is fine tuning and leveraging larger models that can handle very complex questions, reasoning, and data schemas since this is only a 7B.

thecalebf

2 tahun yang laludiscuss

Currently Postgres yes, already working on a dataset with more DDLs like MySQL, DuckDB, MSSQL, etc for a second iteration.

thecalebf

2 tahun yang laludiscuss

Neat, would you ever use a local model for that if it could work with ORMs?

thecalebf

2 tahun yang laludiscuss

That is really interesting. I took note of this. That would be really cool!

thecalebf

2 tahun yang laludiscuss

Great call out. Will definitely focus on that in the next iteration!

thecalebf

2 tahun yang laludiscuss

No, those are benchmark, evaluation questions. The fine tune dataset was a custom, synthetically generated dataset of ~20k PostgreSQL Text to SQL pairs covering different SQL categories and question types.

I mention a little more about it here https://x.com/calebfahlgren/status/1754247740291207198?s=20

thecalebf

2 tahun yang laludiscuss

Updated the title since it may have been confusing, appreciate the feedback!

thecalebf

2 tahun yang laludiscuss

Not yet, planning to build an open source, local, CLI tool that utilizes natural-sql though!

thecalebf

2 tahun yang laludiscuss

Sure, there is a blog post with some other examples of the first model iteration with more complex, multipart questions. https://www.chatdb.ai/post/naturalsql-vs-sqlcoder-for-text-t...

I will update that to be a more truly difficult question. Appreciate the feedback!

thecalebf

2 tahun yang laludiscuss

Thanks! Yes, that was an earlier iteration. The v1 is here https://huggingface.co/chatdb/natural-sql-7b. Plan to push to Ollama soon and build an open source free tool around it.

Would love to hear about what you are building!

thecalebf

2 tahun yang laludiscuss

Very neat! I am building something very similar, called ChatDB.ai

Submissions

Show HN: WebLLM Playground – Run LLMs in the Browser on WebGPU

Valve New Employee Handbook (2012) [pdf]

Show HN: SmolCopilot – 360M LLM writing assistant in the browser

Show HN: I Made a Chrome Extension to Query Hugging Face Datasets with SQL

[untitled]

[untitled]

Show HN: Natural-SQL-7B, a strong text-to-SQL model

[untitled]

[untitled]

Show HN: Query CSV files in the browser with SQL

comments