Errata

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. If the error was corrected in a later version or reprint the date of the correction will be displayed in the column titled "Date Corrected".

The following errata were submitted by our customers and approved as valid errors by the author or editor.

Color key: Serious technical mistake Minor technical mistake Language or formatting error Typo Question Note Update

Version Location Description Submitted By Date submitted Date corrected
F6-2
Figure 6-2

Figure 6-2 shows feature narrative_xf, transform module.py creates consumer_complaint_narrative_xf (not consumer_disputed_xf as reported on 6th Sept)

Note from the Author or Editor:
In figure 6-2, "narrative_xf" should be replaced by "consumer_complaint_narrative_xf"

 Michael Shearer  Sep 13, 2020 
Chapter 4: Data Ingestion - Converting Parquet-Serialized Data to tf.Example

Hello!

I've been previewing this title and it is excellent so far. I wanted to submit a change you'll want to make based on the current API for Tensorflow Extended.

In chapter 4: Data Ingestion under the "Converting Parquet-Serialized Data to TF.Example", the API has changed so that section of code should be updated to the following:



from tfx.components.example_gen.component import FileBasedExampleGen 1
from tfx.components.example_gen.custom_executors import parquet_executor 2
from tfx.utils.dsl_utils import external_input
from tfx.components.base import executor_spec # new import

examples = external_input(parquet_dir_path)
example_gen = FileBasedExampleGen(
input=examples,
custom_executor_spec=executor_spec.ExecutorClassSpec(parquet_executor.Executor)) 3 # new api


I'm really enjoying the title so far! Thanks!

-Jason

Note from the Author or Editor:
We have made the change as suggested.

 Jason Wolosonovich  Apr 08, 2020 
below fig 2-4
lesson 2: section: What is ML Metadata?

lesson 2: section: What is ML Metadata?

be e.g. a raw datasets

should be a raw dataset

Note from the Author or Editor:
Thank you for your note. We will fix the typo in the final version.

 Yanjin Long  Jun 22, 2020 
below fig 2-5
lesson 2: section: Interactive Pipelines

original text: "...and the components artifacts..."
suggested: "...and the components' artifacts..."

 Yanjin Long  Jun 22, 2020 
Other Digital Version
1
lesson 1 section "Overview of the Steps in a Machine Learning Pipeline"

Lesson 1:

"Overview of the Steps in a Machine Learning Pipeline" section


"Feedback Loops" subsection first line "The last step of the machine learning piprline is often forgotten,"


piprline is a type, should be pipeline



Note from the Author or Editor:
Thank you for your message. We will correct the typo in the final version.
- Hannes

 Yanjin Long  Jun 22, 2020 
1
DEFAULT SPLITS in chapter 3

If we don’t specify any output configuration, the ExampleGen component splits the data set into a training and evaluation split with a ration of 2:1 by default.


Ration should be ratio

 Yanjin Long  Jul 09, 2020 
Printed
Page 90
Last para

prediction_fn() should be predict_fn()

 Michael Shearer  Sep 12, 2020 
Printed
Page 102
Definition of precision

Precision is defined as true positives/(true negatives + false positives), which is not correct. It should be: true positives/(true positives + false positives)

Note from the Author or Editor:
Precision is defined as "true positives/(true negatives + false positives)", but it should be "true positives/(true positives + false positives)"

 Ward Van Driessche  Feb 24, 2021 
Printed
Page 239
1st paragraph in the "OpFunc Functions" note box

"digital subscriber line(DSL) objects in Kubeflow Pipelines". It is "domain-specific language (DSL)" as in the kubeflow document (https://www.kubeflow.org/docs/pipelines/sdk/sdk-overview/)

Note from the Author or Editor:
This is correct - it should be "domain specific language" not "digital subscriber line".

 Anonymous  Dec 11, 2020