search

memo / ofxMSATensorFlow

C++ openframeworks addon and examples integrating Google's TensorFlow numerical computation and machine learning library
Apache License 2.0
456 stars 87 forks source link

CUBLAS_STATUS_NOT_INITIALIZED #42

Closed ghost closed 4 years ago

ghost commented 5 years ago

Hi,

I have been using the pix2pix example with tf 1.13. However I updated my tensorflow to 1.14 and the example stopped working. I can compile it but when I run it I get the following error:


2019-10-21 20:57:03.692076: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.7
2019-10-21 20:57:04.360814: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcublas.so.10.0
2019-10-21 20:57:04.442850: E tensorflow/stream_executor/cuda/cuda_blas.cc:238] failed to create cublas handle: CUBLAS_STATUS_NOT_INITIALIZED
2019-10-21 20:57:04.442881: W tensorflow/stream_executor/stream.cc:1916] attempting to perform BLAS operation using StreamExecutor without BLAS support
[ error ] ofxMSATensorFlow: msa::tf::SimpleModel | Internal: 2 root error(s) found.
  (0) Internal: Blas SGEMM launch failed : m=1, n=512, k=8192
     [[{{node generator/encoder_8/conv/Conv2D}}]]
  (1) Internal: Blas SGEMM launch failed : m=1, n=512, k=8192
     [[{{node generator/encoder_8/conv/Conv2D}}]]
     [[generator/generator_outputs/_3]]
0 successful operations.
0 derived errors ignored.
terminate called after throwing an instance of 'std::runtime_error'
  what():  msa::tf::SimpleModel | Internal: 2 root error(s) found.
  (0) Internal: Blas SGEMM launch failed : m=1, n=512, k=8192
     [[{{node generator/encoder_8/conv/Conv2D}}]]
  (1) Internal: Blas SGEMM launch failed : m=1, n=512, k=8192
     [[{{node generator/encoder_8/conv/Conv2D}}]]
     [[generator/generator_outputs/_3]]
0 successful operations.
0 derived errors ignored.

any idea?