teamSolve

Bot task are completed with extra requirements in their respective folders.

Clone the github repo from following command

git clone https://github.com/Drwaish/teamSolve.git
cd teamSolve

Now install the requirements using this command

pip install -r requirements.txt

SQL query generation

In this task I try with single and double tables and it responds accordingly.

If you want to run this task

cd SQL_query_gen

After this execute the main.py file.

python main.py

In this file, example of 2 tables are implemented. You need to specify the columns names and table names in input and it works accordingly.

Uncomment line 11 in main.py if you want to run on single table.

This will also works fine on large databases.

If you want to run this task

cd RAG_QuestioningAnswering

After this run following command.

python main.py

I update the vector at runtime with my resume.First code will create vector db of inout doc

and then update the vector db with resume and respond according to my resume.

I also try different input question it respond accurately.

In this tasks I use zephyr 7b beta for inference.

Export the Hugging face token or put in create_embedding.py line 74 to hassle free execution.