Machine learning com TensorFlow na Vertex AI avaliações

Anshumaan Kumar P. · Revisado há about 5 years

CHHAY S. · Revisado há about 5 years

Christelle X. · Revisado há about 5 years

Federico U. · Revisado há about 5 years

Azmat S. · Revisado há about 5 years

Furkan K. · Revisado há about 5 years

The method proposed to edit the code is so painful that it gets rid of all the fun and learning. The lab text should include a final text version of model.py and task.py to allow comparison and copy-paste if needed. I suggest that first the code is edited on the gcloud instance with the graphical editor and then uploaded to the VM instance to avoid any python tabulation errors headaches with nano.

Antonio M. · Revisado há about 5 years

Usama A. · Revisado há about 5 years

I192103 A. · Revisado há about 5 years

Azmat S. · Revisado há about 5 years

Laziz T. · Revisado há about 5 years

HēT P. · Revisado há about 5 years

Nathália C. · Revisado há about 5 years

Tom D. · Revisado há about 5 years

Ramses A. · Revisado há about 5 years

karan p. · Revisado há about 5 years

Nick P. · Revisado há about 5 years

Onder Alp S. · Revisado há about 5 years

Seems outdated (tf 1.X instead of 2.X) and a lot of the steps are too manual (copy/pasting indented python code is painful). Also, glazes over some of the more interesting pieces like why hash buckets is not as great as using something like pandas.CategoricalDType (relegates this to the oreily book which not everyone has access to). Suggest using a ai platform notebook to automate a lot of the friction points.

Justin W. · Revisado há about 5 years

Marco R. · Revisado há about 5 years

José René L. · Revisado há about 5 years

Javier B. · Revisado há about 5 years

time is very tight. training on AI Platofrm took longer than expexted.

Jerome G. · Revisado há about 5 years

Shambhu S. · Revisado há about 5 years

not enough time to think about what I'm doing it while doing it

Johannes C. · Revisado há about 5 years