Easier data processing with Keras
summary
Shows how the Keras Preprocessing Layers API turns raw text files (positive/negative IMDb folders) into deployable models by: using text vectorization to lowercase, strip punctuation, whitespace-tokenize, keep top 10,000 words, convert via multi‑hot vectors, calling adapt on the training dataset to build the vocabulary. Demonstrates mapping the layer into a tf.data pipeline with parallel CPU threads to avoid GPU idling, and exporting an end‑to‑end model that takes raw strings for inference.
AI-written summary · about
chapters (5)
- 0:00Start
- 0:18Use Cases for Data Preprocessing
- 0:45Data Augmentation
- 2:00Keras Pre-Processing Layers Api
- 3:28Steps of Our Workflow
related talks
30mKick-Start your Understanding of Machine Learning with Python
54mNatural Language Processing with Deep Learning and TensorFlow
35mHow does text become data?
01h 38mNavigating the World of NLP and Text Analytics
44mTeaching Computers ABCs: A Quick Intro to Natural Language Processing
52mMachine Learning Data Pipelines
01h 30mIntroduction to Deep Learning & Natural Language Processing
34mGenerative AI Model Data Pre-Training on Kubernetes: A Use Case Study