unet3d is not access?

[8899012]

this is a small code to catch the problem:
`import torch
import torch_directml
import torch.nn as nn
device = torch_directml.device()

创建3D卷积层和五维输入

conv3d = nn.Conv3d(1, 1, kernel_size=3, padding=1).to(device)
x = torch.randn(1, 1, 16, 256, 256).to(device) # (B, C, D, H, W)

try:
output = conv3d(x)
print("3D卷积在DirectML下运行成功，输出维度：", output.shape)
except Exception as e:
print("错误信息：", e)`

python == 3.10
requirement.txt
what can i do to run the code successfully?