ONNX转换

转换到一半，这个是什么原因
%6595 : Float(1, 360, 80) = aten::empty(%6591, %6592, %6593, %6594), scope: DataParallel
  %6596 : int = prim::Constant[value=0](), scope: DataParallel
  %6597 : int = prim::Constant[value=0](), scope: DataParallel
  %6598 : int = prim::Constant[value=1](), scope: DataParallel
  %6599 : Float(1, 360, 80) = aten::narrow(%6595, %6596, %6597, %6598), scope: DataParallel
  %6600 : Float(1, 360, 80) = aten::expand_as(%6567, %6599), scope: DataParallel
  return (%6584, %6601);
}

:
/home/lix/.local/lib/python3.6/site-packages/torch/tensor.py(337): masked_fill
/home/lix/Fastspeech_Squeezewave/transformer/Modules.py(21): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(477): _slow_forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(487): __call__
/home/lix/Fastspeech_Squeezewave/transformer/SubLayers.py(60): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(477): _slow_forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(487): __call__
/home/lix/Fastspeech_Squeezewave/transformer/Layers.py(125): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(477): _slow_forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(487): __call__
/home/lix/Fastspeech_Squeezewave/transformer/Models.py(94): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(477): _slow_forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(487): __call__
/home/lix/Fastspeech_Squeezewave/fastspeech.py(26): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(477): _slow_forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(487): __call__
/home/lix/.local/lib/python3.6/site-packages/torch/nn/parallel/parallel_apply.py(59): _worker
/home/lix/.local/lib/python3.6/site-packages/torch/nn/parallel/parallel_apply.py(77): parallel_apply
/home/lix/.local/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py(153): parallel_apply
/home/lix/.local/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py(143): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(477): _slow_forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(487): __call__
/home/lix/.local/lib/python3.6/site-packages/torch/jit/__init__.py(252): forward
/home/lix/.local/lib/python3.6/site-packages/torch/nn/modules/module.py(489): __call__
/home/lix/.local/lib/python3.6/site-packages/torch/jit/__init__.py(197): get_trace_graph
/home/lix/.local/lib/python3.6/site-packages/torch/onnx/utils.py(192): _trace_and_get_graph_from_model
/home/lix/.local/lib/python3.6/site-packages/torch/onnx/utils.py(224): _model_to_graph
/home/lix/.local/lib/python3.6/site-packages/torch/onnx/utils.py(281): _export
/home/lix/.local/lib/python3.6/site-packages/torch/onnx/utils.py(104): export
/home/lix/.local/lib/python3.6/site-packages/torch/onnx/__init__.py(27): export
synthesis.py(89): <module>

欢迎加入本站置顶帖qq群交流