This README walks through the setup of Watsonx resources that are needed to execute the notebooks from Adobe’s Cloud ML Ecosystem (CMLE) GitHub repo. The first two notebooks demonstrated how to query and create a dataset from AEP data sources to be used for machine learning; transfer that data using the AEP data flow service to DLZ; The following 3 notebooks will showcase how to train the ML model using a Watsonx; and finally create customer segments derived from model predictions that are delivered back to AEP using the data flow service.
Watsonx is an advanced AI and data platform, designed to accommodate the entire AI lifecycle, including building, managing, and deploying AI models. See Watsonx for more information.
The steps below assume you have an account on watsonx, if not please ensure you have one. See Watsonx for more information.
-
Navigate to watsonx
Project
tab and create new project(select
Create and empty Project
) Supply a name and description, save it. -
Next step we need to pull
user_name
anduser_key
to be able to interact with watson programmatically, navigate toProfile and settings
top right corner, here you have access touser_name
andkey
Additionally we need
instance_id
,url
andversion
these you could find by browsing, at the time of writing these notebooks the version was4.8
-
Next store these values in config.ini (which you used through notebooks 1 and 2) under
[Watsonx]
section. Do not share them with anyone. -
Next step we will upload our config.ini file (which you used through notebooks 1 and 2), go to
Assets
>New asset
>Upload asset to this project
confirm new
Data
asset type was created: -
We are ready to import notebooks, for that navigate to
Assets
>New Asset
>Jupyter Notebook editor
>URL
, supplyname
chooseRuntime 23.1 on Python 3.10
and providepath
to notebooks/watsonx/Week*.ipnynb notebooks from CMLE git repo.
We are all set.