Google Fly Cup Challenge: Pilot

Just change the NAME given in your lab 

export BUCKET_NAME=
export BUCKET_LOCATION=

### Task 1 

gsutil mb -l $BUCKET_LOCATION -b on gs://$BUCKET_NAME 

###  Task 2 

bq mk drl
bq load --source_format=CSV drl.pilot-event-times spls/gsp396/lab2.csv --auto-detect

###  Task 3 

gsutil cp gs://spls/gsp396/DRL2.ipynb .

### Task 4 

Run all of the cells in the DRL2.ipynb file.
 
When setting the PROJECT variable, replace its value with the unique project id of the qwiklabs project

###  Task 5

## Navigate to Vertex AI > Model Registry
 
Click Import
 
Name the model drl-pilot-ranking
 
Choose the dynamically selected region variable displayed in Task 4 for the region
 
# Click Model Settings
 
Choose 2.8 for the Model Framework Version
 
# Navigate and select the following path for the Model Artifact Location:
 
 
PROJECT-ID/model

###  Task 6 

Navigate to Vertex AI > Endpoints
 
Click Create Endpoint
 
Name the endpoint drl-pr-endpoint
 
Choose the dynamically generated region for the region
 
Ensure the traffic split is set to 100%
 
Click on advanced scaling settings
 
Choose n1-standard-2 for machine type
 
Click Create

###  Task 7 

## Return to your notebook tab
 
# Create a new cell and add the following code:
 

from google.cloud import aiplatform
 
PROJECT = '[project-id]'
LOCATION = '[region]'
ENDPOINT_ID = '[your-endpoint-id]'
 
endpoint = aiplatform.Endpoint(ENDPOINT_ID, project=PROJECT, location=LOCATION)
 
instances = [{'pilot_id': [11],
'ename': ['Practice'],
'events_name':['Project Manhattan'],
'minimum_time': [80.0]
}]
 
endpoint.predict(instances = instances)
 
# Replace [project-id] with the unique qwiklabs project id.
 
# Replace [region] with the dynamically selected region variable used for the endpoint
 
# Replace [your-endpoint-id] with the endpoint's unique ID
 
# Both the region and endpoint ID can be found in the Endpoints page in the console
 
# Run the cell in your Vertex Notebook