TPU Speed Data Pipelines Reviews

Mikael T. · Reviewed בערך 2 שנים ago

Ehsan u. · Reviewed בערך 2 שנים ago

Nikolas N. · Reviewed בערך 2 שנים ago

Instructions again make absolutely no sense - "iterate the files" is not an instruction that is complete is any way. Even the question numbers are wrong just listing question 1 repeatedly.

Oliver H. · Reviewed בערך 2 שנים ago

aishwarya b. · Reviewed בערך 2 שנים ago

Joost B. · Reviewed בערך 2 שנים ago

Thorvald Sahl J. · Reviewed בערך 2 שנים ago

It's kind of messy. We're not using accelerators (GPUs) and we're not using parallelism at all. I guess it's nice to learn the data loading.

Loading A. · Reviewed בערך 2 שנים ago

Aditya S. · Reviewed מעל 2 שנים ago

Brody I. · Reviewed מעל 2 שנים ago

Decent but there's a lot missing. TFRecord is introduced and then there is no explanation of tf.io.*, so you'll end up searching through a bunch of docs until maybe you come across parse_single_example or until you look at the solution.

chris w. · Reviewed מעל 2 שנים ago

LUCAS S. · Reviewed מעל 2 שנים ago

Sudharshan N. · Reviewed מעל 2 שנים ago

Alexey T. · Reviewed מעל 2 שנים ago

Suchita B. · Reviewed מעל 2 שנים ago

DANIEL M. · Reviewed מעל 2 שנים ago

Jayesh Arun B. · Reviewed מעל 2 שנים ago

ROBERTO O. · Reviewed מעל 2 שנים ago

Rahul M. · Reviewed מעל 2 שנים ago

David M. · Reviewed מעל 2 שנים ago

Harold H. · Reviewed מעל 2 שנים ago

2. To use TFRecord format to load training data efficiently from GCS it is more efficient is it because that it doesn't show the pictures, but it shows 000100 010000 01010?

Wanda W. · Reviewed מעל 2 שנים ago

Ming H. · Reviewed מעל 2 שנים ago

LUIS F. · Reviewed מעל 2 שנים ago

Kidus B. · Reviewed מעל 2 שנים ago