FileNotFoundError: No secrets files found. Valid paths for a secrets.toml file are: /root/.streamlit/secrets.toml, /app/.streamlit/secrets.toml

[71walceli]

Hi! I just wanted to test it out, and I got this error:

This is run locally, right? Or do I require some sort of key?

[jurajdanis]

Same issue here. The error apparently blocks all the functionality.

[71walceli]

Is it possible and feasible to run this locally? Or this only works with the OpenAI LLM?

[71walceli]

Anyways, I would like to know the correct way to set up a key, I'll help write the instructions for this, though I haven't been able to figure out how...

[ricardobalk]

1) Secrets file

An example of the secrets file can be found at ./app/.streamlit/secrets.toml.example:

OPENAI_API_KEY = "<YOUR API KEY HERE>"

This file should be copied or renamed to secrets.toml and contain your own OpenAI API key.

Keep in mind that even though it looks like it's running locally, you're still using OpenAI's API, so requests are sent over the internet and there's actually no LLM running locally.

2) Running an LLM locally

If you want to run an LLM locally, check out this repo I created just a few days ago: https://github.com/ricardobalk/streamlit-ollama - not as fancy as this Streamlit app, but it provides a basic setup allowing to run llama2/llama3 locally.

[ricardobalk]

Also as an update here, With PR #7 I added multi-provider support, i.e. you can switch between OpenAI, Local LLM via Ollama, Claude etc. - no more need for the repo mentioned before.