Created by: pele228

你好， win10平台， 在同一处理线程下使用NVIDA Video_Codec_SDK 视频解码，解码的图像调用Yolov3进行目标检测， GPU模式下报错如下（CPU检测 正常，GPU不解码直接载入图像detection正常）：

C++ Call Stacks (More useful to developers):

Windows not support stack backtrace yet.

Python Call Stacks (More useful to users):

File "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/framework.py", line 2426, in append_op attrs=kwargs.get("attrs", None)) File "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layer_helper.py", line 43, in append_op return self.main_program.current_block().append_op(*args, **kwargs) File "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/nn.py", line 2801, in conv2d "data_format": data_format, File "/home/aistudio/PaddleDetection/ppdet/modeling/backbones/resnet.py", line 152, in _conv_norm name=_name + '.conv2d.output.1') File "/home/aistudio/PaddleDetection/ppdet/modeling/backbones/resnet.py", line 407, in c1_stage name=_name) File "/home/aistudio/PaddleDetection/ppdet/modeling/backbones/resnet.py", line 428, in call res = self.c1_stage(res) File "/home/aistudio/PaddleDetection/ppdet/modeling/architectures/yolov3.py", line 56, in build body_feats = self.backbone(im) File "/home/aistudio/PaddleDetection/ppdet/modeling/architectures/yolov3.py", line 86, in test return self.build(feed_vars, mode='test') File "tools/export_model.py", line 101, in main test_fetches = model.test(feed_vars) File "tools/export_model.py", line 118, in main()

Error Message Summary:

PaddleCheckError: CUDNN_STATUS_EXECUTION_FAILED at [D:/1.6.1/paddle/paddle/fluid/operators/conv_cudnn_op.cu:288] [operator < conv2d > error]

是PaddleDetection需要独占GPU吗？还是独占线程，需要额外加Detection处理线程呢？