Fine tune LLMs using GPUs with Cloud Run jobs

Preview — GPU support for Cloud Run jobs

This feature is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms. Pre-GA features are available "as is" and might have limited support. For more information, see the launch stage descriptions.

You can fine tune a Gemma 3 model on a Cloud Run job, then serve the fine tuned model on a Cloud Run service using vLLM.

See a step-by-step instructional codelab at How to fine tune a model using Cloud Run jobs.