PaddleOCR icon indicating copy to clipboard operation
PaddleOCR copied to clipboard

PaddleOCR-2.6 GPU版本MFC运行报错:无法定位程序输入点dnnl_prelu_backward_desc_init于动态链接库paddle_inference.dll上

Open stq054188 opened this issue 2 years ago • 1 comments

  • Win10 64位 企业版 VS2017社区版

-PaddleOCR v2.6 GPU版本 , CUDA10.1, cudnn-10.1-windows10-x64-v7.6.5.32,TensorRT-6.0.1.5.Windows10.x86_64.cuda-10.1.cudnn7.6

  • 运行指令/Command Code:
  • 控制台运行指令:ppocr.exe --det_model_dir=./model_1/en_PP-OCRv3_det_infer/ --rec_model_dir=./model_1/en_PP-OCRv3_rec_infer/ --image_dir=./imgs/
  • 控制台运行结果正常:
  • I0919 19:55:04.800885 7316 naive_executor.cc:102] --- skip [feed], feed -> x I0919 19:55:04.803768 7316 naive_executor.cc:102] --- skip [softmax_2.tmp_0], fetch -> fetch predict img: ./imgs\A.png W0919 19:55:04.862386 7316 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 6.1, Driver API Version: 10.1, Runtime API Version: 10.1 W0919 19:55:04.862386 7316 gpu_resources.cc:91] device: 0, cuDNN Version: 7.6. ./imgs\A.png 0 det boxes: [[12,53],[636,48],[636,102],[12,107]] rec text: Improve Brand Awareness rec score: 0.960799 1 det boxes: [[95,137],[574,143],[573,197],[94,191]] rec text: The Best to Rely on rec score: 0.941441 2 det boxes: [[473,248],[630,247],[630,265],[473,266]] rec text: OpenCVA rec score: 0.911527 The detection visualized image saved in ./output//A.png

MFC环境下运行报错

  • 完整报错/Complete Error Message: 无法定位程序输入点dnnl_prelu_backward_desc_init于动态链接库paddle_inference.dll上 image

同样的Demo,使用PaddleOCR 2.6 CPU版本,MFC可正常运行: image

stq054188 avatar Sep 19 '22 11:09 stq054188

之前没遇到过,查了下,可以试着把dll加到系统环境变量里试下,或者可以先看下仅使用cpp代码(不依赖mfc),能否跑通gpu推理

littletomatodonkey avatar Sep 20 '22 07:09 littletomatodonkey