因此如果tensorrt模型将Batch Size作为Dynamic 的就能起到很好… ... with Windows ML app, you'll need to convert the model to ONNX format. onnx documentation .
確定! 回上一頁