yua_j
RRunPod
•Created by yua_j on 3/9/2025 in #⛅｜pods-clusters
Issue with Checkpoint Switching and Runtime Error
Hello, I’m experiencing an issue when switching checkpoints on RunPod. When I change the checkpoint, it looks like it switches, but in reality, it doesn’t. Then, when I try to generate an image, I get the following error:
"RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument index in method wrapper_CUDA__index_select)"
It seems like the model is not correctly loading onto the GPU. I’ve tried restarting the instance, but the problem persists. Does anyone know how to fix this?
Thanks in advance!
2 replies