728x90
반응형
아래와 같은 에러는
2019-05-21 08:35:00.310722: E tensorflow/contrib/tensorrt/log/trt_logger.cc:38] DefaultLogger engine.cpp (99) - Cuda Error in initializeCommonContext: 4 (Could not initialize cudnn, please check cudnn installation.)
terminate called after throwing an instance of 'nvinfer1::CudaError'
what(): std::exception
Aborted (core dumped)
Session 돌릴 때 gpu memory 할당을 너무 높게 줘서 생긴 문제 일 수도 있다.
with tf.Session(config=tf.ConfigProto(gpu_options=tf.GPUOptions(per_process_gpu_memory_fraction=0.5))) as sess:
아래와 같이 0.4 로 설정하니까 성공
with tf.Session(config=tf.ConfigProto(gpu_options=tf.GPUOptions(per_process_gpu_memory_fraction=0.4))) as sess:728x90
반응형
'삽질 기록' 카테고리의 다른 글
| docker 에서 cannot connect to X server unix:0 (0) | 2019.06.10 |
|---|---|
| [Git] fatal: Authentication failed (4) | 2019.05.31 |
| No such file or directory # include "opencv2/xfeatures2d/cuda.hpp" (0) | 2019.04.10 |
| Could not find module FindOpenCV.cmake or a configuration file for package OpenCV (2) | 2019.04.10 |
| [TensorRT] SampleMNIST using RTX 2080 & TensorRT 5.0.2.6 (0) | 2019.04.09 |