Yup. Basically, dstack allows you to run ML workflows in the cloud as if you did it locally. For example, you can specify how many GPUs you need or how much RAM and dstack will automatically create a cloud instance that satisfies the requirements to run the workflow.