argosml 0.0.4

A package for tracking and visualizing ML experiments

Argos Tracker Getting Started

• Clone Repo
• Make sure Anaconda is installed
• Create conda env with conda env create --name argos-tracker --file=environment.yml
• To update previously created envs use conda env update --file environment.yml --prune

Jupyter Lab

• Activate conda env with conda activate argos-tracker
• Create project specific ipykernel if it has not been created yet:

python -m ipykernel install --user --name argos-tracker --display-name "argos-tracker"

• Start jupyter lab: jupyter lab
• Open client using link provided on start
• Open test_notebook_mimic.ipynb and select the argos-tracker ipykernel

MLflow

• Open a new terminal window and activate the argos-tracker env
• Start mlflow: mlflow server --host 127.0.0.1 --port 8787 --app-name mlflow_cors

Argos Dashboard

• In a third terminal window activate the argos-tracker env
• Go to the reports_dashboard directory cd reports_dashboard
• Run npm ci to install dependencies for Vue frontend
• The following error may occur if using ubuntu with a version older than 24.04:

node: /lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.31' not found (required by /home/user/anaconda3/envs/argos-tracker/bin/../lib/libnode.so.127)

• Start client with npm run dev