Yang Chengyue
Yang Chengyue
In gradio_app/app.py, line 12 ``` c_dtype = torch.bfloat16 if device.type == "cpu" else torch.float ``` Isn't the cuda device that uses bfloat16?
if not , is there any possible alternative that can be used on windows platform?At this moment I can't run distributed trainning on windows os
if not , is there any possible alternative that can be used on windows platform?At this moment I can't run distributed trainning on windows os
### Expected Behavior I dont install ipex in my computer, thus intel xpu should not be enabled. ### Actual Behavior In comfy\model_management.py, the initial value of xpu_available is: ```py xpu_available...