The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
workaround also works on cudnn 91200 (#10375)
comfyanonymousb1293d50eff5
diff --git a/comfy/ops.py b/comfy/ops.pyindex 893ceda98..56b07b44c 100644--- a/comfy/ops.py+++ b/comfy/ops.py@@ -55,7 +55,7 @@ except (ModuleNotFoundError, TypeError):NVIDIA_MEMORY_CONV_BUG_WORKAROUND = Falsetry:if comfy.model_management.is_nvidia():- if torch.backends.cudnn.version() >= 91300 and comfy.model_management.torch_version_numeric >= (2, 9) and comfy.model_management.torch_version_numeric <= (2, 10):+ if torch.backends.cudnn.version() >= 91200 and comfy.model_management.torch_version_numeric >= (2, 9) and comfy.model_management.torch_version_numeric <= (2, 10):#TODO: change upper bound version once it's fixed'NVIDIA_MEMORY_CONV_BUG_WORKAROUND = Truelogging.info("working around nvidia conv3d memory bug.")