Chat With Research Documents Privately

I assume many of you watching this video would agree that AI tools offer tremendous productivity benefits. But every time I input a query into any GPT interface or ask some interface a question, I get a sinking feeling in my stomach. That feeling stems from the privacy terms in the age of cloud computing, where AI service providers potentially store all the queries we make. Using large language models can help researchers speed up my reading and writing process or even help me boost productivity. However, each query I send to the cloud computing platform could fall into the hands of service providers. As researchers, some of our queries are not just data; some queries and questions are the seeds of future breakthroughs. So why not keep your chats private by using large language models from your own Computer. In this video, I show you how to set up PDF chat using Ollama, Gemma2, and nomic-embed-text. Links for all models and the Python Github repo featured in this video are below: Python app.py: Author (Sanjjushri Varshini R) https://github.com/Sanjjushri/rag-pdf-qa-llama3 Sanjjushri Varshini R (Youtube Channel)- https://youtube.com/@sanjjushri?si=tirCsBLACh9sfcQV Ollama : https://ollama.com/ Gemma2: https://ollama.com/library/gemma2 nomic-embed-text: https://ollama.com/library/nomic-embed-text To download models from Command Line Interface, type: ollama run "name of the model"