Скачать или смотреть DNN WindingTempEst-1-Experiment Tracking Using MLFlow On Google Colab

This video works through a trick to use MLFlow to track experiments on Google Colab and View results on the Local PC
00:00 Overview
01:40 Concept diagram
05:41 Go through the Notebook
11:30 MLflow tracking folder structure
16:12 Parsing the url
16:41 Set the experiment name
17:45 mlflow tracking during training
24:35 mlfow.log_metric
25:00 mlflow.log_figure
26:40 Problem: spin up the mlfow ui in Google colab won't work
28:28 Solution: Google Drive + Local mflow ui
31:00 Problem: Just spin up mlflow ui locally won't work
32:57 reason for picture and model not showing up
34:35 Solution 1: change the artifact_url in meta.yaml
36:45 Registered model- the proxy
39:35 (Better)Solution 2: use proxy magical address