The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
mix_ascend_bf16_infer_err (#6794)
zhoufan295635740259de27
diff --git a/comfy/model_management.py b/comfy/model_management.pyindex 29cd43b51..cbf4c4ea6 100644--- a/comfy/model_management.py+++ b/comfy/model_management.py@@ -1082,6 +1082,9 @@ def should_use_bf16(device=None, model_params=0, prioritize_performance=True, maif is_intel_xpu():return True++ if is_ascend_npu():+ return Trueprops = torch.cuda.get_device_properties(device)if props.major >= 8: