Errata: January 9, 2021

Thank you for purchasing Deep Learning with PyTorch . Please post errata not listed below in this book's LiveBook Errata thread. All formats are corrected during the second printing. Thank you!

Page xxviii

The "about this cover" section incorrectly identifies the individual on the cover as Kardinian, but he is Kabardian.

Page 5

Figure 1.1 was incorrectly flipped left-to-right.

Page 195

The third code snippet incorrectly referenced "i22" instead of "i13" in the first line. The full listing should be:

o11 = i11 * w00 + i12 * w01 + i13 * w02 +
      i21 * w10 + i22 * w11 + i23 * w12 +
      i31 * w20 + i32 * w21 + i33 * w22

Page 205

The reference to a "MaxPool3d" should be "MaxPool2d".

Page 252

The final paragraph of section 9.4.4 should read:

Once you have the candidates file and at least one subset downloaded, uncompressed, and put in the correct location, you should be able to start running the examples in the next chapter. If you want to jump ahead, you can use the p2ch10_explore_data.ipynb Jupyter Notebook to get started. Otherwise, we'll return to the notebook in more depth in the next chapter. Hopefully your downloads will finish before you start needing to use the data!

Page 258

The first code annotation should read:

Counts the number of lines that end with 1, which indicates it is a nodule.

The end of the first paragraph should read:

(it's a Boolean value: 0 for a candidate that is not a nodule, e.g. a blood vessel, and 1 for a candidate that is a nodule, which can be malignant or benign). We have 1,351 of the nodule candidates flagged as nodules.

Page 272

Delete the two text snippets "1((CO10-1))" and "1((CO10-2))" from Listing 10.12.


The code annotation "This is our training sample." should cover those two lines as well.

Page 277

The first code annotation should read:

# In[7]:
from p2ch10.vis import findPositiveSamples, showCandidate
positiveSample_list = findPositiveSamples()