BentoML

hi Scott, could you give more details about "run GPU while developing with the `stable-diffusion-bentoml`?

when I run the bentoml locally and hit the service, it runs stable diffusion on the CPU, rather than the GPU as it will when I deploy it...I'd like to develop while using the GPU

does `torch.cuda.is_available()` return `True` in your development environment?

Tho, I am noticing now when I specify the gpu's to use in my `configuration.yaml` file, it says:

```[bentoml-cli] `serve` failed: GPU device index in [2, 4] is greater than the system available: [0]```
If `cuda.is_available()` is true then perhaps I have the wrong cuda stuff installed maybe and its not seeing my GPU?

This message means that your current GPU is at index 0, whereas in your configuration you are setting it to use GPU:2 and GPU:4, which either doesn’t exist or not available

makes sense, thank you! I got it all working :slightly_smiling_face: