That is great! Question: if we rely on a multimodal model to extract text from an image, and the multimodal model can return slightly different text everytime, is there a way we could prevent from calling the multimodal model by referencing the hash of the file instead of the hash of the extracted text?
The world needs more llamaindex videos
I was wondering how this was done and now that you have explained it, I have a better understanding.
Love what you are all doing. Makes everything so much more manageable.
Keep making videos for all of us
Awesome!! Will def use this.
That is great! Question: if we rely on a multimodal model to extract text from an image, and the multimodal model can return slightly different text everytime, is there a way we could prevent from calling the multimodal model by referencing the hash of the file instead of the hash of the extracted text?
So how should this work if I'm working with a Vector Store like Elasticsearch, instead of persisting to the disk?
Thank you
Excellent! Could you share the notebook please
no links were added to the description
where is the code.,