Ryan talked about converting https://github.com/MxDkl/pls into Python in order to help translate natural language into command line things.
Scott asked about using ChatGPT to parse Slack conversations and create tasks.
Things to research:
- https://platform.openai.com/tokenizer
- ChatGPT playground is free to test out prompts and inputs
- gpt-3.5-turbo is the cheapest model
- https://github.com/openai/tiktoken
- Alternative: https://python.langchain.com/docs/integrations/document_loaders/slack
Tim asked about creating some kind of test case "thing" to bolt on to his existing framework
- Tim doesn't like Robotframework
- too heavy
- doesn't want to write too much "non Python" stuff
- Want's something basic to describe/run test cases
Things to research:
- pytest
- hypothesis api testing
- hitch might be an alternative to robotframework
- mimesis
- taverntesting
Dillon's looking for a step-by-step guide for creating LLMs
Things to research:
Ryan 6:11 PM
https://github.com/MxDkl/pls
Ryan 6:26 PM
https://python.langchain.com/docs/integrations/document_loaders/slack
Ryan 6:28 PM
https://platform.openai.com/tokenizer
https://github.com/openai/tiktoken
Ryan 6:40 PM
https://docs.pytest.org/en/7.1.x/example/nonpython.html
https://hypothesis.readthedocs.io/en/latest/examples.html#fuzzing-an-http-api
Ryan 6:47 PM
https://hitchdev.com/hitchstory/
Ryan 6:48 PM
https://mimesis.name/en/master/
Ryan 6:49 PM
https://taverntesting.github.io/
Ryan 6.?? PM
https://www.youtube.com/watch?v=kCc8FmEb1nY