This repository is for the GDG Cloud Kolkata workshop on 27th August, 2022. Learn about minimal MLOps required for any ML practitioner. It uses Weights and Biases.
MIT License
This repository is for the GDG Cloud Kolkata workshop on 27th August, 2022. Learn about minimal MLOps required for any ML practitioner. Find the curated list of resources here.
⭐ Slide Deck: http://wandb.me/minimal-mlops-deck
If you don't have a free W&B account follow the steps:
To authenticate any machine to start logging experiments to your W&B account,
1. git clone https://github.com/ayulockin/minimal-mlops
2. cd minimal-mlops
3. pip install -e .
1. pip install -r requirements.txt
Everything is stitched together using configs file. You can find the config for this repo in configs/
dir.
python train.py --config configs/baseline.py
python train.py --config configs/baseline.py --wandb
python train.py --config configs/baseline.py --log_model
python train.py --config configs/baseline.py --log_model --wandb
python train.py --config configs/baseline.py --wandb --log_eval
python log_data.py --config configs/baseline.py
I hope you find it useful. If you encounter any issue please raise an issue. :)