Building Machine Learning Pipelines

Errata for Building Machine Learning Pipelines

Submit your own errata for this product.


The errata list is a list of errors and their corrections that were found after the product was released.

The following errata were submitted by our customers and have not yet been approved or disproved by the author or editor. They solely represent the opinion of the customer.


Color Key: Serious Technical Mistake Minor Technical Mistake Language or formatting error Typo Question Note Update



Version Location Description Submitted By Date Submitted
Safari Books Online S5
Beam Pipeline code box

beam.io.ReadFromTFRecord requires a coder to be specified, eg data_coder = tft.coders.ExampleProtoCoder(raw_data_metadata.schema) pipeline | beam.io.ReadFromTFRecord(tfrecords_filename, coder=data_coder))

Michael Shearer  Sep 06, 2020 
Printed Page 12-14
Figures 1-4, 2-1, 2-2

If I see it correctly, these 3 figures (1-4, 2-1, and 2-2) on pages 12 to 14 are almost identical and they only seem to vary slightly in different shades of grey. This is at the very least quite confusing because I don't see the added value of showing it repeatedly on 3 subsequent pages, and it makes me think that I have to find some important difference between the 3 figures but I can't really see it. Thanks a lot for the great book though!

Christopher Schiefer  May 13, 2021 
Printed Page 20
Table with frameworks by company

The table lists 'AirBnb' but it's later spelled 'AirBnB', whereas it's official name seems to be either 'Airbnb' or 'airbnb'.

Christopher Schiefer  May 13, 2021 
Printed Page 73
Code at top

the line img = tf.image.convert_image_dtype(img, tf.float32) refers to an object 'img' which is not defined in the function Also, it is not explained how the process_image function should be called in the preprocessing_fn function. I get errors when trying to implement.

Ward Van Driessche  Feb 12, 2021 
Printed Page 74
2nd code block

Need to specify a data coder on ReadFromTFRecord, eg tft.coders.ExampleProtoCoder(raw_data_metadata.schema)

Michael Shearer  Sep 12, 2020 
Printed Page 87
Example 6.4

The serving function refers to a function '_get_serve_tf_examples_fn', wheras the function in example 6.4 does not have the prefix underscore.

Ward Van Driessche  Feb 19, 2021 
Printed Page 116
Définition of the predict method

The def predict(test_examples) should be def predict(examples) as the name of the parameter used in the body of the method in the list decoration is 'examples' and not 'test_examples'.

Jérôme Massot  Jun 08, 2021 
Printed Page 135
Just above "Model Signatures"

"which saw" should become "which we saw"

Ward Van Driessche  Mar 02, 2021 
Printed Page 210
code

Hi, I'm really enjoying the book so far! Great quality of writing and clear explanations and insight. Maybe this is because of a tfx update but I think image_to_example does not take input_dict as an input. and then we have: input_base_uri = exec_properties[tfx.components.example_gen.utils.INPUT_BASE_KEY] Thanks, Raphael

Raphael Rzb  Oct 28, 2020