TensorFlow Extended - TFX Overview and Pre-Training Workflow

Explore TensorFlow Extended (TFX), a comprehensive machine learning platform, in this 32-minute conference talk from the TensorFlow Dev Summit 2019. Dive into TFX's pre-training workflow, focusing on crucial aspects such as Data Validation and Transformation. Learn about building components, model export paths, data distribution visualization, and model tracking. Discover how to handle multiple models, utilize shared configuration models, and implement orchestration. Gain insights into data ingestion, analysis, and understanding through statistics and anomaly reports. Understand the importance of schema validation and various transformation techniques, including transform utility, pipeline, label, and graph. Enhance your machine learning pipeline knowledge with practical examples and workshop context provided by Product Manager Clemens Mewald.

Intro
TensorFlow Libraries
Building Components
Component Overview
Model Export Path
Data Distribution Visualization
Model Tracking
Multiple Models
Visualization
Driver Publisher
Shared Configuration Model
Orchestration
Examples
Workshop
Context
Model
Data Validation
Ingest Data
Data Analysis Validation
Data Understanding
Statistics
Why are my predictions bad in the morning
Schema
Payment Types
Example Validator
Anomaly Reports
Transformations
Transform Utility
Transform Pipeline
Transform Label
Transform Graph