Show HN: Template for Chatwithpdf File on Colab (colab.research.google.com)

I built this using the Mistral 7B model; the vector embeddings take time. I would love suggestions on how to run it faster on CPU and general feedback appreciated.