Databricks-GenAI-Assoc Assembling and Deploying Apps — All Questions With Answers

{"model_name": "genai-chat-bot", "endpoint_config": {"min_instances": 1, "max_instances": 5, "instance_type": "GPU_MEDIUM"}, "tags": {"env": "prod"}}

resources:
  jobs:
    training_job:
      name: "model_training"
      tasks:
        - task_key: "train"
          new_cluster:
            spark_version: "13.3.x-scala2.12"
            node_type_id: "i3.xlarge"
            num_workers: 2

log_model(model=model, artifact_path="model", 
          signature=signature, 
          pip_requirements=["scikit-learn==1.2.2", "pandas==1.5.3"])

# Deployment Config:
compute_config:
  gpu_enabled: true
  instance_type: "g4dn.xlarge"

log_entry: 2023-10-27 10:00:00 [ERROR] Model inference failed: Connection to Vector Store timed out. Retrying in 5s... 2023-10-27 10:00:05 [ERROR] Max retries exceeded.

{"model": "llama-3", "config": {"gpu": "nvidia_a10", "min_instances": 0}, "autoscaling": "enabled"}

Error: Model serving endpoint update failed. Reason: 'INSUFFICIENT_PERMISSIONS' - The service principal does not have access to the model in the Unity Catalog.

{
  "model_name": "fraud_detection",
  "version": 5,
  "endpoint_name": "fraud-prod",
  "traffic_config": {
    "routes": [
      {
        "served_model_name": "fraud_detection-5",
        "traffic_percentage": 100
      }
    ]
  }
}

{
  "model_name": "customer_churn",
  "model_version": "2",
  "endpoint_name": "churn_inference",
  "traffic_config": {
    "routes": [
      {
        "served_model_name": "churn_v1",
        "traffic_percentage": 90
      },
      {
        "served_model_name": "churn_v2",
        "traffic_percentage": 10
      }
    ]
  }
}