Polong Lin

@ivanmkc the test is failing because the BQML model was previously deployed in a previous test, and needs to be undeployed first in the test project (by first undeploying the...

@ivanmkc I've finished resolving the comments from @sarahcdugan , but the execution tests are failing (because of the issue stated earlier). ``` BadRequest: 400 Model Registry failed: com.google.cloud.ai.platform.common.errors.AiPlatformException: code=FAILED_PRECONDITION, message=The...

@ivanmkc Finally this has passed the CI tests! Could you check and merge if this looks good with you? Sarah has already approved.

@ivanmkc Regarding the folder -- we should also consider how these notebooks are used/discovered. If the entry point is Model Registry, then moving to the Model Registry makes sense, but...

Hi rain-ml, It looks like you may not have run one of the prior cells that creates the LTV_PARAMS dictionary: ``` LTV_PARAMS = { 'WINDOW_LENGTH': 0, 'WINDOW_STEP': 30, 'WINDOW_STEP_INITIAL': 90,...

Looking good overall. 1. Could you add in a README to your directory, similar to [this](https://github.com/slopp/analytics-componentized-patterns/tree/lopp-xai-code-snippets/retail/time-series/bqml-demand-forecasting )? 2. Could you rename the directory to: /finance/xai-fraud/bqml-looker/ Thank you!

Hi @Aakif-cloud , this can happen if the embedding model was not (for some reason) successfully able to create an embedding for the input text, and so the `embeddings` variable...

Thanks for flagging this! We just fixed this across the board for all the notebooks based on your feedback: https://github.com/GoogleCloudPlatform/generative-ai/pull/34

Hi @dvmorris, Indeed, the runtime doesn't need to be restarted, because Colab now has the latest version of `google-cloud-aiplatform` pre-installed (1.25.0) already. However, for future-proofing, it's still safer to keep...

After testing the notebook in Vertex AI Workbench, I've made some minor changes to the notebook and renamed the file to be simpler. Could you PTAL at the changes and...