Hello Instructor We're reaching out to discuss strategies to promote your Udemy course. With your expertise and the course's value, we see impactful ways to boost visibility and expand your audience.
I am very glad to have found your channel. Your explanations are very clear and understandable, especially regarding the Faiss vector database. Perhaps it would be possible to have a tutorial on how to develop an LLM app with Flowise and the Faiss vector database?
This is very informative if we are planning to use only palm2. We can skip the boiler plate codes if we use langchain or llamaindex. That can simplify the development process with few lines of code.
Hi I have used your code for this and changed the images. To test the effectiveness of the program, I took an image from the database as a query image and ran the program. Ideally the top candidate image should be the same image as the query image, but for some images I get a totally different image with distance = 0. In fact for all my top candidates the image distance is 0. Why is this? BTW this is a very good and concise video.
Nevermind, I solved the bug. I gave the wrong order of filenames in the data array in comparison to my imgs folders. I changed the logic for the code where I use glob to directly take the images in the folder and embed them. So the orders of the filenames changed. 👍
Thanks for the information. Is there any way to store the category too in the FAISS DB and filter by category(tagging) before the similarity search for large data?