C4W3 Assignment "Implementing Canary Releases ..."

Course Q&A Machine Learning Engineering for Production(MLOps) Deploying Machine Learning Models in Production
1

Hi,

In the section "Deploying ResNet50, creating a storage bucket in my project in run in the issue that bucket is not properly created although I believe I used my project ID as name prefix correctly. see attached screenshot - line 4 of the screenshot

39CFC5F3-B82B-4D37-8F54-C5F9269155AF_4_5005_c
39CFC5F3-B82B-4D37-8F54-C5F9269155AF_4_5005_c1146×228 108 KB

Pls. advise,
thanx in advance and best regards
Jörn

2

Hi Jorn! It seems you hardcoded the PROJECT_ID and that causes the MODEL_BUCKET string to be interpreted differently. Please use the command mentioned in the instructions:

export MODEL_BUCKET=${PROJECT_ID}-bucket
gsutil mb gs://${MODEL_BUCKET}

You already declared PROJECT_ID in the earlier sections of this exercise so you can simply reuse it as shown above. You can read more about string manipulation in shell scripting here. Hope this helps!

3

I have a similar issue. Created the bucket and copied resnet 101 and resnet 50 but the system does not recognize it.

image
image2516×326 148 KB

4

Hi! As mentioned in the previous comment, it seems you hardcoded the PROJECT_ID and that causes the MODEL_BUCKET string to be interpreted differently. Please use the command mentioned in the instructions:

export MODEL_BUCKET=${PROJECT_ID}-bucket
gsutil mb gs://${MODEL_BUCKET}

You already declared the variable PROJECT_ID in the earlier sections of this exercise so you can simply reuse it as shown above. You don’t have to type qwiklabs-gcp-02...... You can simply use PROJECT_ID. You can read more about string manipulation in shell scripting here. If it fails again, please take another screenshot of the same commands you showed above. Thanks and hope this helps!