Live Coding A Machine Learning Model from Scratch (Google I/O'19)

Do you want to build a machine learning model, but not sure where to start? In this session, learn how to start with an empty Colab notebook, code a model using TensorFlow and Keras, train the model live, deploy it to Cloud AI Platform for serving, and use the deployed model to generate predictions from a web app.

Watch more #io19 here:

Speaker(s): Sara Robinson

T81AAF event: Google I/O 2019; re_ty: Publish; product: Cloud - AI and Machine Learning - AI building blocks; fullname: Sara Robinson;

[Steven-yisp]

While it won't make everything clear (due to time constraints and the fight for human attention), this is a really really helpful start to understanding ML!
Thanks a lot Sara! This is probably the first really informative beginner's resource I have found on the topic

[alllll]

Very informative video. You explanation of how ML is doing matrix multiplication under the hood was amazing! As were your descriptions of one-hot and bag-of-words encoding. Never heard of SHAP before, looks like something I would find very useful. Thank you!

[AnimeshSharma]

from model building to deployment in less than an hour! and handling the live demo gotcha gracefully #awesome

[jagadeeshn]

Beautifully explained AI from how to create a model, till how to test a models accuracy using shap

[suhasnama]

Awesome talk, thank you for the introduction to SHAP !!

[shashankbarki]

Thanks Sara and GCP ! This was an excellent presentation and demo.

[MrDonald]

Thanks for the content ! But I guess Colab is not adapted for long trainings that last several days right ?

[kushxmen]

This is inspiring, I have been doing it the hard way.

[JonMasters]

Awesome video! This is my favorite from I/O 2019 by far!

[wojciechduranowski]

very interesting and well-structured presentation

[harshadpatil]

Thanks Sara!! It was a very informative session!!

[narendraprasath]

Very nice hands on coding. It would be better if you do it on regular basis on different kinds of problems.

[ultraviolet.catastrophe]

My model just outputs "Wakanda Forever" pls help

[anoubhav]

Why don't use binary cross entropy and not categorical crossentropy ?

[akashdhar]

Why didn’t you use CountVectorizer from sklearn for tokenizing?

[username]

where is the colab file she used in the presentation?