PyTorch native Metrics. ... Accuracy(task="multiclass", num_classes=5) # move the metric to device you want computations to take place device = "cuda" if ...
確定! 回上一頁