Slow Inference on arm64 machines

[aryanagarwal9]

Describe the bug: A clear and concise description of what the bug is.
marqo is taking a lot of time to index and search

To Reproduce: Steps to reproduce the behavior:
On an arm64 machine, use docker to run marqo, and then try to index and search

Expected behavior: A clear and concise description of what you expected to happen.
Operations should be fast

Desktop (please complete the following information):
OS: macOS Monterey v12.5
Machine: M1 MacBookPro, 2020

[aryanagarwal9]

This can be fixed by following these steps to run marqo locally:

• Run marqo-os

 docker run --name marqo-os --platform linux/amd64 -id -p 9200:9200 -p 9600:9600 -e "discovery.type=single-node" marqoai/marqo-os:0.0.2 

• Clone the marqo repo and cd into the directory
• Run the following commands

brew install cmake;
brew install protobuf;
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs/ | sh;
pip install -e .

• cd into src/marqo/tensor_search and run:

export OPENSEARCH_URL="https://localhost:9200" && 
    export PYTHONPATH="${PYTHONPATH}:<your path here>/marqo/src" &&
    uvicorn api:app --host 0.0.0.0 --port 8882 --reload

Note: Insert the absolute path to the marqo repo

[aryanagarwal9]

The latest marqo docker image supports arm64. Follow these instructions in the README.