mlflow-project-template

mlflow-project-template

MLflow-project-template

MLflow project template

STEPS -

STEP 01- Create a repository by using template repository

STEP 02- Clone the new repository

STEP 03- Create a conda environment after opening the repository in VSCODE

conda create --prefix ./env python=3.7 -y

To access conda from c drive to d, through bash

---

. C:/Users/DELL/Anaconda3/etc/profile.d/conda.sh

---

conda activate ./env

OR

source activate ./env

STEP 04- install the requirements

pip install -r requirements.txt

STEP 05 - Create conda.yaml file -

conda env export > conda.yaml

Copy a template of python files

'''bash cp src/stage_00_template.py src/stage_01_get_data.py '''

STEP 06- commit and push the changes to the remote repository

MLFlow commands

Command to run MLFlow Project files

'''bash mlflow run . --no-conda '''

runany specific entry point in MLFlow Project files

'''bash mlflow run . -e get_data --no-conda '''

runany specific entry point with certain configuration in MLFlow Project files

'''bash mlflow run . -e get_data -P configs/your_config.yaml --no-conda '''