search

tjmlabs / ColiVara-Eval

Colivara Evaluation
MIT License
2 stars 0 forks source link

Write collection names and file names in the readme #4

Closed HalemoGPA closed 1 day ago

HalemoGPA commented 2 days ago

The collection names in evaluate.py are fixed. Also, in main.py (upsert.py in the future). So, We need to specify them in the readme.md to make it easy for anyone to use our Eval.

HalemoGPA commented 1 day ago

I have another idea, and want your opinion @Jonathan-Adly

Instead of setting the names of collections fixed. and only adding them to the readme. I can update the evaluate.py to list collections and make the user choose the one to evaluate instead of typing the fixed name we specified earlier.

On the long term, if we had other datasets for evaluation, we wouldn't need to update readme each time to include them as I intended in the previous comment.

Example of the command: we have two options

  1. using python collection_manager.py --list then copying the name of collection we want. then use python evaluate.py --api_key "api_key" --collection_name "collection_name_got from_previous_command".

  2. making the use of command python evaluate.py --api_key "api_key" to print the list of collections numbered from 1 to 10 (currently they are 10) and asks for the user to enter the number corresponding to the collection he wants then evaluation starts.

  3. add collection names to readme and user copies from there as intended earlier.

Jonathan-Adly commented 1 day ago

No - have the collection name standardized and indicate on the readme is what we will go with.

HalemoGPA commented 1 day ago

Ok.