Hello there How can I use llama index

You'll want to use a local LLM and local embedding model. You'll need at least 15GB of VRAM though

Check out the GPU section of this notebook
https://colab.research.google.com/drive/16QMQePkONNlDpgiltOi7oRQgmB8dU5fl?usp=sharing

Thanks, I will read.