Create an account with AutoML Admin role and download the key as json file. After the credentials are downloaded set the environment variable GOOGLE_APPLICATION_CREDENTIALS as the absolute path to the credentials file.
$ GOOGLE_APPLICATION_CREDENTIALS="absolute/path/to/credentials"
Use this link to know how to set up one
Run the following code in the terminal or virtual environment to install the required dependencies.
$ pip install -r requirements.txt --upgrade
- Replace the project_id variable with your project_id on line 3 of index.py.
project_id = "your_project_id"
-
Replace gs bucket path to the dataset with your own path in the file index.py, line number 4.
path = "gs://path/to/your/dataset"
-
On line number 5 of index.py change the display_name variable to anything that you want it to be reflected on the dataset and model.
display_name="<any_display_name>"
NOTE Do not use any special character other than underscore ( _ )
-
Set the appropriate target column on line number 6
target_column_name = "Target Column"
-
And then run the script using the following command.
$ python index.py
OR
$ python3 index.py
Go to Tables datasets to view your dataset.
Go to Tables models to view your model.