Scribd
Upload
2 views9 pages

RAG Overview

Uploaded by

Sai Bandari
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
2 views9 pages

RAG Overview

Uploaded by

Sai Bandari
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 9

Bot

1.Documentation,bing search,csv,sql,Chatgpt

2.GIS ( file,TASK------ Graphical Report)

a)Backend

1.main functionality are in utils.py

2.channael between frontend and backend was established using app.py

3.which it should use bot.py

b)Frontend

1.Home.py

User question ----- Bot.py------ llm,tools,agent,agent executor,awnser----user output

Question ---- decided the tool ---- docsearchagent--- GetdocsearchResult_tool---


customazureretriver

2) GIS Code

Frontend------- LLM_Geo_AP14.ipyb (Where all the Data Locations(Files),TASK then all the
functions and methods are called)

Backend------ All the Gis Functionality present

a)LLM_Geo_Kernel.py ------- Solution class

b) LLM_Geo_Constants.py ---- all the prompts for giving the LLM

c) helper.py ------- Extract the code or text from the LLM generated response and

return to the function

d) LLm_Unit_test.py----- Debug the LLM generated code

Prompts

a) Graph related
b) Operation related
c) Assembly Role
d) Debug Related
e) Operation Review related
f) Assembly review related
Code

a) Data Locations
b) Generated a graph
c) Extract operations from graph
d) For each operation LLm generates the Python Code

Operation

1)Files are Loaded

i)code is generated to load 1st file

i)code is generated to load 2st file

i)code is generated to load 3st file

2) Join the relevant File

Common Columns in Files

a) HW (geometry)

b) Boundary of the state (geometry,GEOID)

c) Population CSV Data (GEOID)


The code is for generating the graph ----- prompt related to graph is called --- get llm
response for graph ---- LLM --- generates the code --- is present in response variable--to
extract the code which was generated by the LLM is done using helper module
Import all modules

Given the file paths in DATA_LOCATIONS

Solution class is initialized

To generated the code related to graph, then get_LLM_reply_for_graph function is called then
inside it get_LLm_reply is called

Prompt---- to the get_LLm_reply--- response ---- from the response we need to extract
the code ---- helper.extract code

Operation which are present in the graph

Clear upto graphs ,

3 files

First thing we need to load this files(Read the file)

1)haz_waste_gdf 2)tract_boundary _gdf 3)tract_population_df csv

Join the files

1st and 2nd file based on the common column GEOID

You might also like