PaddlePaddle / PaddleSeg

Easy-to-use image segmentation library with awesome pre-trained model zoo, supporting wide-range of practical tasks in Semantic Segmentation, Interactive Segmentation, Panoptic Segmentation, Image Matting, 3D Segmentation, etc.
https://arxiv.org/abs/2101.06175
Apache License 2.0
8.57k stars 1.68k forks source link

[Bug]DDRNet模型部署问题 #2679

Closed OpenCDR closed 1 year ago

OpenCDR commented 1 year ago

PaddleSeg版本:PaddleSeg release/2.6 paddleinference版本:4596b9a22540fb0ea5d369c3c804544de61d03d0 操作系统:Win10 20H2 CUDA 版本 11.6 CUDNN版本 8.4.1 TenserRT版本: 8.4.0 使用动态shape导出ddrnet模型后在该电脑上推理出现以下错误: Error: ../paddle/phi/kernels/funcs/elementwise_functor.h:545 Assertion b != 0 failed. InvalidArgumentError: Integer division by zero encountered in (floor) divide. Please check the input value. Error: ../paddle/phi/kernels/funcs/elementwise_functor.h:545 Assertion b != 0 failed. InvalidArgumentError: Integer division by zero encountered in (floor) divide. Please check the input value.


C++ Traceback (most recent call last):

Not support stack backtrace yet.


Error Message Summary:

ExternalError: CUDA error(719), unspecified launch failure. [Hint: Please search for the error code(719) on website (https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html#group__CUDART__TYPES_1g3f51e3575c2178246db0a94a430e0038) to get Nvidia's official solution and advice about CUDA Error.] (at ..\paddle\phi\backends\gpu\gpu_context.cc:435)

请问这种情况应该如何处理?相同情况下PPlite模型推理没有问题

juncaipeng commented 1 year ago

你执行代码是什么? 我在linux上执行如下导出、预测代码,是正确的,没有报错。

python tools/export.py --config configs/ddrnet/ddrnet23_cityscapes_1024x1024_120k.yml --save_dir output/tmp/ddrnet
python deploy/python/infer.py --config output/tmp/ddrnet/deploy.yaml --image_path tmp/cityscapes_demo.png --save_dir output/tmp
OpenCDR commented 1 year ago

导出代码是python tools/export.py --config configs/ddrnet/ddrnet23_cityscapes_1024x1024_120k.yml --save_dir output/tmp/ddrnet 动态shape收集的代码是python ../python/collect_dynamic_shape.py --config ddrnet/deploy.yaml --image_path images --dynamic_shape_path ./dynamic_shape.pbtxt 在使用的是CPP的部署方式而非Python。 CPP的代码就是例程里提供的。

juncaipeng commented 1 year ago

这个估计是paddle inference预测库的问题,建议先换个其他的预测库试试。

OpenCDR commented 1 year ago

好的谢谢您,我们现在只能用CPP的库。

juncaipeng commented 1 year ago

@OpenCDR 是换个其他版本的paddle inference cpp预测库。 我们会跟进下该问题。