Databricks-GenAI-Assoc Application Development — All Questions With Answers

{
  "model_name": "my_model",
  "model_version": "1",
  "config": {
    "served_models": [
      {
        "model_name": "my_model",
        "model_version": "1",
        "workload_type": "CPU",
        "scale_to_zero_enabled": true
      }
    ]
  }
}

Refer to the exhibit.

# Configuration snippet for model serving
{
  "name": "my-llm-endpoint",
  "config": {
    "served_models": [{
      "model_name": "my-model",
      "model_version": "1",
      "workload_type": "CPU",
      "scale_to_zero_enabled": true
    }]
  }
}

Refer to the exhibit.

# Model Serving Policy Configuration
{
  "traffic_config": {
    "routes": [
      {
        "served_model_name": "model-v1",
        "traffic_percentage": 90
      },
      {
        "served_model_name": "model-v2",
        "traffic_percentage": 10
      }
    ]
  }
}

bundle:
  name: my_rag_app

include:
  - resources/*.yml

targets:
  dev:
    default: true
    workspace:
      host: https://adb-123456789.azuredatabricks.net
  prod:
    workspace:
      host: https://adb-987654321.azuredatabricks.net
      root_path: /Shared/bundle_prod

{
  "model_name": "customer_support_llm",
  "task": "llm/v1/chat",
  "endpoint_config": {
    "auto_capture_request_payload": false,
    "traffic_config": {
      "routes": [{"served_model_name": "v1", "traffic_percentage": 100}]
    }
  }
}

import mlflow
mlflow.set_tracking_uri("databricks")
# Missing configuration here
model_info = mlflow.pyfunc.log_model(artifact_path="llm_model", python_model=model)

{
  "model_signature": {
    "inputs": [{"name": "prompt", "type": "string"}],
    "outputs": [{"name": "response", "type": "string"}]
  }
}

{
  "endpoint_name": "rag-bot-v1",
  "config": {
    "served_models": [
      {
        "model_name": "llama-3-8b",
        "model_version": "5",
        "workload_size": "Small",
        "scale_to_zero_enabled": true
      }
    ]
  }
}

Error: 403 Forbidden
Message: The calling principal does not have 'Can query' permission on the endpoint.
Context: model_serving_endpoint_name: 'llama-3-rag-prod'