关于“Introduction to TensorFlow Data Validation”的评价

Dima E. · 评论over 4 years之前

ヨシモトナオト ヨ. · 评论over 4 years之前

John N. · 评论over 4 years之前

Great lab

Viktor O. · 评论over 4 years之前

ok

konutech k. · 评论over 4 years之前

MOHAMED A. · 评论over 4 years之前

SOURAV S. · 评论over 4 years之前

Mohamed A. · 评论over 4 years之前

Dennis L. · 评论over 4 years之前

Brandon G. · 评论over 4 years之前

Neil S. · 评论over 4 years之前

Fernando M. · 评论over 4 years之前

Shreyas B. · 评论over 4 years之前

Paul C. · 评论over 4 years之前

Bougda A. · 评论over 4 years之前

Seth R. · 评论over 4 years之前

Sumit Y. · 评论over 4 years之前

Vinayak G. · 评论over 4 years之前

Gulshat K. · 评论over 4 years之前

The lab was great intro to tfdv. There is an error in this notebook that I need to make a pull request on. The version of pyarrow which they specify in the list of !pip3 install .. statements is missing the package pyarrow.vendored, which breaks the version of Beam that tfdv uses. This causes tfdv.generate_statistics_from_dataframe(any_pd_df) to throw a ValueError that alludes to a missing package pyarrow.vendored. When I remove the pip install statement that includes some version number (like 2.0.0 that they specify) and replace it with '! pip3 install pyarrow', then version 5.0 installs. If I repeat the tfdv.generate_statistics_from_dataframe(), it no longer throws a value error.

David T. · 评论over 4 years之前

Sidia -. · 评论over 4 years之前

Thomas S. · 评论over 4 years之前

Andrei B. · 评论over 4 years之前

Anish K. · 评论over 4 years之前

Maxime P. · 评论over 4 years之前