Pipelines de dados com velocidade de TPU avaliações

Mikael T. · Revisado há about 2 years

Ehsan u. · Revisado há about 2 years

Nikolas N. · Revisado há about 2 years

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. · Revisado há about 2 years

aishwarya b. · Revisado há about 2 years

Joost B. · Revisado há about 2 years

Thorvald Sahl J. · Revisado há about 2 years

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. · Revisado há about 2 years

Aditya S. · Revisado há over 2 years

Brody I. · Revisado há over 2 years

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. · Revisado há over 2 years

LUCAS S. · Revisado há over 2 years

Sudharshan N. · Revisado há over 2 years

Alexey T. · Revisado há over 2 years

Suchita B. · Revisado há over 2 years

DANIEL M. · Revisado há over 2 years

Jayesh Arun B. · Revisado há over 2 years

ROBERTO O. · Revisado há over 2 years

Rahul M. · Revisado há over 2 years

David M. · Revisado há over 2 years

Harold H. · Revisado há over 2 years

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. · Revisado há over 2 years

Ming H. · Revisado há over 2 years

LUIS F. · Revisado há over 2 years

Kidus B. · Revisado há over 2 years