search

statmike / vertex-ai-mlops

Google Cloud Platform Vertex AI end-to-end workflows for machine learning operations
Apache License 2.0
507 stars 233 forks source link

Applied Forecasting AutoML Forecasting Python client.ipynb #16

Closed stevewalker-de closed 1 year ago

stevewalker-de commented 2 years ago

Cell 6, 'Create Dataset (Link to BigQuery Table)' requires notebook 1 'prepped' table to be created. I had to figure that out and go find it, and execute it. Didn't know if that was your intent. -steve

stevewalker-de commented 2 years ago

And cell 10 errors out with TIMESTAMP

NameError Traceback (most recent call last) /tmp/ipykernel_22050/3421282980.py in 15 validation_options = "fail-pipeline", 16 budget_milli_node_hours = 100, ---> 17 model_displayname = f"{NOTEBOOK}{DATANAME}_{TIMESTAMP}", 18 model_labels = {'notebook':f'{NOTEBOOK}'}, 19 holiday_regions = ['GLOBAL', 'NA', 'US'],

NameError: name 'TIMESTAMP' is not defined

statmike commented 2 years ago

Hi @stevewalker-de , Sorry about this. I have been updating the repository layout and uniformity across notebooks. The last 2 sections on my list for updates are AutoML (02) and This folder (Applied Forecasting). I am working on Applied Forecasting this week so it will be fixed shortly. Sorry for the inconvenience. A quick fix for the TIMESTAMP is adding a cell with this above its use:

from datetime import datetime
TIMESTAMP = datetime.now().strftime("%Y%m%d%H%M%S")
stevewalker-de commented 2 years ago

No worries at all, worked through it fine, was just letting you know. Cheers!

On Tue, Oct 25, 2022 at 6:03 AM Mike Henderson @.***> wrote:

Hi @stevewalker-de https://github.com/stevewalker-de , Sorry about this. I have been updating the repository layout and uniformity across notebooks. The last 2 sections on my list for updates are AutoML (02) and This folder (Applied Forecasting). I am working on Applied Forecasting this week so it will be fixed shortly. Sorry for the inconvenience. A quick fix for the TIMESTAMP is adding a cell with this above its use:

from datetime import datetime TIMESTAMP = datetime.now().strftime("%Y%m%d%H%M%S")

— Reply to this email directly, view it on GitHub https://github.com/statmike/vertex-ai-mlops/issues/16#issuecomment-1290299898, or unsubscribe https://github.com/notifications/unsubscribe-auth/AQF3N77CLWAEA2CVAJGCPATWE6WAZANCNFSM6AAAAAARNH5POQ . You are receiving this because you were mentioned.Message ID: @.***>

--

Steve Walker

@.***

Developer Advocacy, Data Analytics, Google Cloud

statmike commented 1 year ago

Marking this one as complete as the Applied Forecasting series just got a big update pass that included fixing this, many other items, and adding many new features and methods.