search

onnx / onnx-tensorrt

ONNX-TensorRT: TensorRT backend for ONNX
Apache License 2.0
2.95k stars 544 forks source link

IShuffleLayer reshape dimensions can't be 2? Thanks for any help. #264

Closed lmxlmx closed 4 years ago

lmxlmx commented 5 years ago

[2019-09-23 10:29:41 WARNING] Setting layouts of network and plugin input/output tensors to linear, as 3D operators are found and 3D non-linear IO formats are not supported, yet. [2019-09-23 10:29:41 ERROR] (Unnamed Layer* 182) [Shuffle]: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [1,5]

ShanQiong commented 5 years ago

I also meet the same problem when the onnx model come from torch.onnx.export, however the pretrained onnx model from https://github.com/onnx/models/tree/master/vision/classification/resnet works well. (Unnamed Layer* 71) [Shuffle]: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [1,4] [08/27/2019-16:57:13] [E] [TRT] Builder failed while configuring INT8 mode.

kevinch-nv commented 4 years ago

Is this still reproducible with the latest TensorRT version, if so can you provide the network you are using?

kevinch-nv commented 4 years ago

Closing due to inactivity - if you are still having issues feel free to reopen.

selimceylan commented 3 years ago

I have same problem and using: TensorRt 8.2 CUDA 10.2 cuDNN 8.2.1 W10

IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [1,2].

azuryl commented 2 years ago

Is this still reproducible with the latest TensorRT version, if so can you provide the network you are using?

@kevinch-nv I meet the same problem in TRT 8.2.1 CUDA 10.2 cuDNN 8.2.1 ubuntu18.04 [shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_12: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])

AmazingRoad commented 2 years ago

I meet the same problem, anyone has solutions? TRT 8.2.1.8

num-doc commented 2 years ago

I meet the same problem, anyone has solutions? TRT 8.4.1.5 CUDA Version: 11.1 CUDNN Version: 8.0.4 Operating System: win10 Python Version (if applicable): 3.7 PyTorch Version (if applicable): 1.11 [shuffleNode.cpp::nvinfer1::builder::ShuffleNode::symbolicExecute::392] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])

frankvp11 commented 2 years ago

Similar problem here too

ghazalehtrb commented 2 years ago

Similar problem

RuntimeError: While parsing node number 14:
ModelImporter.cpp:180 In function parseGraph:
[6] Invalid Node - Pad_14
[shuffleNode.cpp::symbolicExecute::392] Error Code 4: Internal Error (Reshape_3: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
DayDayupupupup commented 2 years ago

Same problem in TRT 8.2.1.8 

IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [1,5]

I found the error occurred in tf.repeat, then I switched to tf.tile，it worked fine.