I have been thinking to deploy vectara, I had 2 queries:
-
In the documentation, you guys have mentioned following code: python3 prepare_config.py /path/to/<YOUR-CONFIG.yaml> , what is profile here, what to put, which profile or id?
-
If I use the method mentioned on github documentation to deploy, will the changes that I did in the header.js file and logo changes after cloning the initial repo,also reflect in the deployed version, because we are not using those files.
Hey @Vishesh_Tripathi
- I assume you mean “python3 prepare_config.py /path/to/<YOUR-CONFIG.yaml> ”? This code assumes you have a file called secrets.toml (GitHub - vectara/vectara-answer: LLM-powered Conversational AI experience using Vectara) with the API Key. Specifically TOML files can host multiple secret configuration, each under a different “profile”. For example if you have
[profile1]
api_key=“XXX”
[my_profile2]
api_key=“YYY”
This file has two profiles called “profile1” and “my_profile2”
So the variable called <PROFILE. above should be the profile you want to use, and it will pull the appropriate api_key from that profile.
- I understand you are making changes to the initial cloned repo. Are you asking whether these changes will be applied to the deployed version? If yes, then the answer it should (in some cases you may need to re-run “npm build” - depending on what the change is that you made)