Summary (in construction)
Data: Google Speech Commands Model: Wav2Keyword (rewritten in pytorch-lightning) Data storage: DVC + GDrive Experiment Tracking: MLFlow Model Repo: MLFlow + DVC Hyperopt: Ray Tune
Serving: TorchServe Monitoring: Prometheus + Grafana
UI: Streamlit