so umm i will start with the first project created by me which was just to understand how a rag works and how a rag basically opparates and how the syntax of lang chain looks and what is does and essentially to run this project i had to import os(oparating system than i had basically use that os to download all the librarys required for the code to run ) than the rag bassically does its thing the code starts with definig the web page the website opporates on and bassically didines diffrent input bars to take in inputs for ex the first one(google api key was ment to be private)so we use the type password so that it doesnt show the previously typed chrachters this code can only opporate on pdf type files and only takes them as inputs and only reads them so essentially this code only works for pdf files then the code goes on to define how basically the the text should be split into peices of text as that is required to get the most relvent peice of text from the pdf file then the fungtion to give the prompt is defind which just gives it an additional text with the question and the context of the convorsation then this fungtion basically defines what kind of api you use write fungtion basically displays the out put of the model then we basically go on to define the webpage in form of main then we run main! Screenshot 2025-02-05 010830 this is basically how this looked completly oporational then i starte with the second code in which i did similar things but i used open ais chatgpt 3.5 model but to further understand how gans acctually work i did not use langchain while creating this model i still cant understand somethings i did while creating this model but i understood a lot more about the code of a gan as when i dont use langchain we have to code out the fungtions individually and have to create them on our own in this code we finnaly intigrated 3 input system for the 3 types of files docx pdf or text files the api used was the same as the one posted before

now we get to the last bit of code i wrote this code was bassically adding the voice based features to the code without langchain i tried like 3 4 librarys but none of seemed to be properly install on stream lit properly but after like 15 attempts of doing it the webpage was finally executed and the voice input was visible and the code worked and then i finally started to make changes to the program bassicaly makeing changes till i break the code and yup it did break and i coudnt recover it from git hub but this is how the code looks opporational as i took a screen shot of how it looked! Screenshot 2025-02-06 172045