You need to train an XGBoost model on a small dataset. Your training code requires custom dependencies. You want to minimize the startup time of your training job. How should you set up your Vertex…

Question

Options

• AStore the data in a Cloud Storage bucket, and create a custom container with your training
• BUse the XGBoost prebuilt custom container. Create a Python source distribution that includes the
• CCreate a custom container that includes the data. In your training application, load the data into a
• DStore the data in a Cloud Storage bucket, and use the XGBoost prebuilt custom container to run

• A
  70% (26)
• B
  19% (7)
• C
  3% (1)
• D
  8% (3)

Explanation

The key constraint is 'custom dependencies' and 'minimize startup time.' When a training job starts, any dependencies not pre-installed must be fetched and installed at runtime-this adds significant latency. A custom container solves this by baking all dependencies into the Docker image at build time, so the container is ready to run immediately on startup with zero installation overhead. Option B uses a prebuilt XGBoost container with a Python source distribution, but custom dependencies would still need to be installed at job startup, increasing startup time. Option C embeds data inside the container, which bloats the image unnecessarily-data should live in Cloud Storage. Option D uses a prebuilt container, which also lacks the custom dependencies pre-installed. Therefore, Option A (custom container + data in Cloud Storage) minimizes startup time.

No community discussion yet for this question.