Set “fallback” GPUs

GPU availabilities on Modal can fluctuate, especially for tightly-constrained requests, like for eight co-located GPUs in a specific region.

If your code can run on multiple different GPUs, you can specify your GPU request as a list, in order of preference, and whenever your Function scales up, we will try to schedule it on each requested GPU type in order.

The code below demonstrates the usage of the gpu parameter with a list of GPUs.