The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Fix lowvram edge case.
comfyanonymous915fdb57454c
diff --git a/comfy/ops.py b/comfy/ops.pyindex 5e7c668eb..3c5ba0124 100644--- a/comfy/ops.py+++ b/comfy/ops.py@@ -264,10 +264,14 @@ def fp8_linear(self, input):scale_input = self.scale_inputif scale_weight is None:scale_weight = torch.ones((), device=input.device, dtype=torch.float32)+ else:+ scale_weight = scale_weight.to(input.device)+if scale_input is None:scale_input = torch.ones((), device=input.device, dtype=torch.float32)inn = input.reshape(-1, input.shape[2]).to(dtype)else:+ scale_input = scale_input.to(input.device)inn = (input * (1.0 / scale_input).to(input.dtype)).reshape(-1, input.shape[2]).to(dtype)if bias is not None: