Tuning and scaling your ML models

Опубликовано 1 сентября 2022, 16:00

Hyperparameter tuning codelab → goo.gle/3AQi5Jo
Distributed training codelab → goo.gle/3pKBF3E
Distributed training with TensorFlow video → goo.gle/3QSJtfr

A huge part of the machine learning process is experimentation, luckily there are a few Vertex AI features that can help you with tuning and scaling your ML models. In this episode of Prototype to Production, Developer Advocate Nikita Namjoshi takes a look at hyperparameter tuning, distributed training, and experiment tracking. Watch this episode to learn how you can get models out of experimentation and into production with Vertex AI.

Chapters:
0:00 - Intro
0:49 - Hyperparameter tuning
1:28 - Hyperparameter tuning on Vertex AI
3:25 - Distributed training
5:16 - Configuring worker pools
6:00 - Experimentation with TensorBoard
7:03 - Vertex AI experiment tracking service
7:30 - Wrap up

Hyperparameter tuning on Vertex AI docs → goo.gle/3RiRxpT
Distributed training on Vertex AI docs → goo.gle/3QNgdXz
Vertex AI TensorBoard docs → goo.gle/3CBiMb4
Vertex AI Experiments docs → goo.gle/3pP0U4O
Vertex AI Experiments notebooks → goo.gle/3Kw7cQj

Prototype To Production playlist → goo.gle/PrototypeToProduction
Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech
​​
#Prototype2Production