sketch

This is great~ There's been some really rapid progress on Text2SQL in the last 6 months, and I really thinking this will have a real impact on the modern data stack ecosystem! I had similar success with lambdaprompt for solving Text2SQL (https://github.com/approximatelabs/lambdaprompt/) where one of the first projects we built and tested was a Text-to-SQL very similar to this Similar learnings as well: - Data content matters and helps these models do Text2SQL a lot - Asking for multiple queries, and selecting from the best is really important - Asking for re-writes of failed queries (happens occasionally) also helps The main challenge I think with a lot of these "look it works" tools for data applications, is how do you get an interface that actually will be easy to adopt. The chat-bot style shown here (discord and slack integration) I ca

Another comment from this thread seems pretty cool to me. https://github.com/approximatelabs/sketch

In terms of applications with it, I have made things like sketch: https://github.com/approximatelabs/sketch Raw prompt-structure ideas i've worked with: - Iterate on a prompt with another "discriminator" prompt, that determines the result is good / safe - Write N-trials of an answer, then use another prompt to select the best answer - When doing code writing (SQL or Pandas) write the output, then use a parser (eg. `ast` in python) to validate code is valid, if not, feed back into a prompt for fixes - Logical negation checks (check if X, and if ~X, give opposite answers, then it's likely consistent, if it's both "affirmative" (as the models tend to bias towards), then it's definitely hallucinating) Other 'product' ideas i've tried: - A chat style interface (I made a chat-bot last year, similar to chatGPT) - A "google-this-for-me" style cha

Right now, it sends the first five rows of the dataframe: https://github.com/approximatelabs/sketch/blob/9d567ec161015...

From https://github.com/approximatelabs/sketch/blob/main/sketch/p... it appears that this library is calling a remote API, which obviates the utility of the demonstrated use case. Upon closer inspection, it looks like https://github.com/approximatelabs/sketch interfaces with the model via https://github.com/approximatelabs/lambdaprompt, which is made by the same organization. This suggests to me that the former may be a toy demonstration of the latter. Interesting how as of the time of writing this, most of the comments here (i.e. dozens) are praising this as a legitimate use case. Maybe I'm missing something obvious, but it seems clear to me that uploading data to a third party to verify whether that data contains PII is a non-starter for any serious application.

Well, I'm locked out of my github account right now and don't feel like going through all those hoops right now but I wanted to point something minor out. In this line, https://github.com/approximatelabs/sketch/blob/9d567ec161015... I think you can end up marking control characters as "UNKNOWN" characters by accident by assuming that in all contexts/environments that dictionary.items() always returns items in a consistent order. This isn't always true. edit: actually with the way the code is written if you have any overlapping ranges at all you'll end up double/triple/etc. counting a character into multiple categories.

This is sending summary statistics to a cloud machine by default (for ease of immediate use. https://github.com/approximatelabs/sketch#sketch-currently-u... You can run using your own OpenAI key by setting 2 environment variables: (1) SKETCH_USE_REMOTE_LAMBDAPROMPT=False (2) OPENAI_API_KEY=YOUR_API_KEY To run entirely locally (using your own GPU and a model like Bloom) one would have to add a new prompt type to `lambdaprompt` (the package that this depends on), have a machine with enough GPU resources, and then add a slight modification to sketch.

digging thru the code https://github.com/approximatelabs/sketch/blob/9d567ec161015... this seems to be using their gpt3 frameowrk: https://github.com/approximatelabs/lambdaprompt which uses text-davinci-003 by default https://github.com/approximatelabs/lambdaprompt/blob/main/la...

Hey HN! I’m excited to share sketch: a tool to help anyone who uses python and pandas quickly iterate and get to answers for their data questions. Sketch installs as a pandas extension that offers utility functions that operate on natural language prompts. Using the `ask` interface you can get answers in natural language. Using the `howto` interface you can get get python and pandas code directly. The primary benefit of this over copilot and chatGPT is that this adds data-content based context so that the generated answers are much more accurate and relevant to the data problem at hand. Check out the demo video[1] and try it out using the colab notebook (on github)! [1] https://user-images.githubusercontent.com/916073/212602281-4...