fine tuning

data scientist task

Prompt Text:

SYSTEM: help me structure this idea:
i want to create a script that takes as input an excel file , a list of column names from that excel file [col1, col2,col3, col4] clo1 refers to the input , col2 refers the field name , col3 refers to the previous output , col4 the feedback .
what the script does is to use the provided colomn names from the excel and use the first col1 (which is a URL) to invoke a lambda called "v3e-dev-ml-l-reghub-rms-metadata" . that lambda takes as input the structure in of : input_structure.json , the only thing we need to change is "body": "{\n    \"DocumentSource\": \"https://insightchemicals.dev.3eiq.com/ArielFT/Doc/2/1612/en-US/636421977915908569.HTM\"
by replacing DocumentSource value with the value of col 1. 
then the lambda will return something that follows the structure of output_structure.json here we will will use the col2 wich contain the field name to get the corresponding value from the output body metadata json we will call it New_response.
now the script output will be an excel with the following columns [col1, col2,col3,col4,New_response ]