Many of us knows how to train & deploy ML models in cloud, but doing so have we become redundant. Running multiple experiments in single machine & waiting for tasks to complete cannot be time-efficient for big datasets. Hence, we need an automation which can take over repetitive manual tasks & spare us the time to do other important stuff. Aim is to show how to deploy ML architecture in 60 SECONDS
ML pipeline consists of many manual tasks such as Data collection, Data cleaning, training environment setup, training configuration, monitoring progress or model evaluation gig, all these components should be automated & what you should be left with is just a single CONFIGURATION document with information of different set of experiments.