What's new in Machine Learning
summary
Announces concrete releases and results: TensorFlow Hub: 10,000+ models and TensorFlow 2.5 adds tf.numpy for NumPy-style GPU/TPU work, Model Maker automates transfer learning, and TensorFlow Cloud sends notebooks to distributed training. Claims TensorFlow Lite runs on 4 billion devices, adds on‑device training and a TFLite runtime via Play Services to avoid APK bloat, plus Micro for kilobyte-class MCUs. Production pieces land too: TFX 1.0 stable and Vertex AI with pipelines, Feature Store, and a Trevor Project case logging ~5,000 training hours and 2,800 trained counselors.
AI-written summary · about
chapters (6)
- 0:00Start
- 1:11TensorFlow Evolution
- 2:11Generative Adversarial Network
- 5:19Know Your Data
- 7:04TensorFlow Cloud Easily train with the power of GCP
- 9:53Model Optimization Toolkit Results
related talks
07mMachine learning tools from Google
26mMachine Learning with TensorFlow
13mTaking Machine Learning to the Next Level
42mWhat's New with TensorFlow?
13mWhat's new in Android machine learning
10mMachine Learning for mobile
20mAI and Machine Learning for developers
54mAdventures in Javascript Machine Learning