Ollama
[Ollama]((https://github.com/ollama/ollam) allows you to run large language models locally.
Instal Ollama via brew
on Mac OSX:
brew install --cask ollama
Run the Ollam application and follow the step-by-step setup guide to install the Ollama command line tool.
To run a model locally:
# The models I am currently running locally:
ollama run phi3
ollama run llama3
ollama run gemma
Google Vertex AI
To run Jupyter notebooks in VS Code with google-cloud-sdk
:
brew install --cask google-cloud-sdk
gcloud config set project YOUR_PROJECT_ID
# To install or remove components at your current SDK version [475.0.0], run:
# gcloud components install COMPONENT_ID
# gcloud components remove COMPONENT_ID
# To update your SDK installation to the latest version [475.0.0], run:
# $ gcloud components update
# Log into Google Cloud
gcloud auth application-default login
# Verify the configuration
gcloud config list
If you receive an error (ModuleNotFoundError: No module named 'imp'
) when running any of the above commands, try:
brew upgrade google-cloud-sdk
# As of 13/05/2024, gloud components update does not support python3.12+)
# We need to use python 3.11 to run the command
export CLOUDSDK_PYTHON=$(which python3.11)
gcloud components update
# gloud does work with python3.12+,
# we can safely unset the environment variable
unset CLOUDSDK_PYTHON
gcloud version
If you receive PermissionDenied: 403 Your application is authenticating by using local Application Default Credentials. The aiplatform.googleapis.com API requires a quota project, which is not set by default. To learn how to set your quota project, see [https://cloud.google.com/docs/authentication/adc-troubleshooting/user-creds](https://cloud.google.com/docs/authentication/adc-troubleshooting/user-creds) . [reason: "SERVICE_DISABLED"
, use the command below to set up the quota project.
gcloud auth application-default set-quota-project YOUR_PROJECT