R Gui crashes with Keras Tensorflow windows cuda gpu -
edit: solved. not using proper version of cudnn. after downloading , copying on cudnn v5.1 library windows 10, works fine far. should have followed tensorflow windows guidelines more closely.
i spent time set windows system run r-tensorflow keras , tensorflow backend.
i followed these blog instructions, , able hello world example work. however, when continue r keras examples , several crash , exit rgui during running training phase, e.g. mnist_cnn fails @ training model block.
# train , evaluate model %>% fit( x_train, y_train, batch_size = batch_size, epochs = epochs, verbose = 1, validation_data = list(x_test, y_test) )
a visual studio jit popup shows "an unhandled win32 exception occured in rgui.exe[10456]." , vs editor shows, "unhandled exception @ 0x00007fffea2734be (ucrtbase.dll) in rgui.exe: fatal program exit requested."
there little load on cpu , or gpu during runtime (less 30%). ideas on how further debug appreciated.
hardware/software/setup: windows 10 pro 64 bit os x64 cpu intel i7-7700 @3.60ghz 64g ram r 64 bit 3.4.1 anaconda 3/python 3.5.4 nvidia 1060 6g cuda 8.0 v8.0.60 visual studio 2015 community edition
using
library(tensorflow) use_condaenv("r-tensorflow") library(keras)
i ran in rstudio w/ rdesktop log diagnostics on file , got following log after crash (i see cannot find file, not file).
20 aug 2017 01:29:32 [rdesktop] error system error 2 (the system cannot find file specified); occurred at: virtual void rstudio::core::http::namedpipeasyncclient::connectandwriterequest() c:/users/administrator/rstudio/src/cpp/core/include/core/http/namedpipeasyncclient.hpp:84; logged from: void rstudio::desktop::networkreply::onerror(const rstudio::core::error&) c:\users\administrator\rstudio\src\cpp\desktop\desktopnetworkreply.cpp:288 20 aug 2017 01:29:42 [rdesktop] error system error 2 (the system cannot find file specified); occurred at: virtual void rstudio::core::http::namedpipeasyncclient::connectandwriterequest() c:/users/administrator/rstudio/src/cpp/core/include/core/http/namedpipeasyncclient.hpp:84; logged from: void rstudio::desktop::networkreply::onerror(const rstudio::core::error&) c:\users\administrator\rstudio\src\cpp\desktop\desktopnetworkreply.cpp:288
edit: stepping through several more exampes, several seem work fine -- specific particular code , setup. don't know enough right troubleshoot code ideosyncracies. crash seem @ model fitting stage. i'll update list on crash/no-crash test. pass = no-crash, fail = crash , exit
mnist_hierarchical_rnn # pass mnist_transfer_cnn # fail variational_autoencoder_deconv # fail stateful_lstm # pass variational_autoencoder # pass reuters_mlp # pass mnist_mlp # pass mnist_irnn # pass mnist_cnn # fail mnist_antirectifier # fail
Comments
Post a Comment