[DEPRECATED] Training and Deploying a TensorFlow Model in Vertex AI Reviews

Navish Vardanaa S. · Reviewed over 2 years ago

Sudipta G. · Reviewed over 2 years ago

Jan K. · Reviewed over 2 years ago

Tarik S. · Reviewed over 2 years ago

Mehdi V. · Reviewed over 2 years ago

Alessandro G. · Reviewed over 2 years ago

Jordi V. · Reviewed over 2 years ago

Lot of Time taken at Build and submit your container image to your Artifact Repository !gcloud builds submit --timeout=20m --config {MODEL_NAME}/cloudbuild.yaml {MODEL_NAME} Hence Cudnt Complete the Lab to see the Model with Test data and hosting the model

Vijayanand D. · Reviewed over 2 years ago

Szymon B. · Reviewed over 2 years ago

command cd training-data-analyst/self-paced-labs/vertex-ai/vertex-ai-qwikstart pip install -U -r requirements.txt cannot be executed. No such directory

Yuping Y. · Reviewed over 2 years ago

Adrian R. · Reviewed over 2 years ago

Steven K. · Reviewed over 2 years ago

Great lab!

Yhary E. · Reviewed over 2 years ago

Shashi G. · Reviewed over 2 years ago

Varinder S. · Reviewed over 2 years ago

2023q2 데. · Reviewed over 2 years ago

Eduardo J. · Reviewed over 2 years ago

Mathias D. · Reviewed over 2 years ago

James C. · Reviewed over 2 years ago

Eduardo P. · Reviewed over 2 years ago

Mauro C. · Reviewed over 2 years ago

Mario Alberto M. · Reviewed over 2 years ago

Step Credentialed Accounts output is outdated. The output is: ACTIVE: * ACCOUNT: <myaccount>@qwiklabs.net --- In the notebook, I had issues with: tf.keras.utils.plot_model(model, show_shapes=False, rankdir="LR") telling me that I had to install pydot. To solve it: - In the notebook, I did: pip install pydot pip install pydotplus - And in the terminal: sudo apt install graphviz As a first timer in the use of notebook, I had no idea that a [*] meant that the step was still ongoing. Waiting for 20 minutes for build and submit to the Artifact Registry is way too long. Because of different issues, I didn't manage to finish the lab within the 1 hour and half. This is very upsetting. I will need to redo it again... ------------ Second time, the endpoint deployment took 20 minutes instead of the 5 minutes mentioned in the notebook. ------------ I am very dissatisfied because I have waited 3 * 20 minutes in the last try.

Marie-Helene T. · Reviewed over 2 years ago

The time for the lab ( 1:30) was not enough to complete the lab !

Ahmed Z. · Reviewed over 2 years ago

Joyce C. · Reviewed over 2 years ago