Given the dynamic nature of the PyTorch execution graph, however, the export process must ... scope: AlexNet % 32: Long = onnx:: Squeeze [axes = [0]](% 31), ...
確定! 回上一頁