search

h2oai / deepwater

Deep Learning in H2O using Native GPU Backends
Apache License 2.0
282 stars 93 forks source link

[2017-07-10 11:37:44] failure_stack_traces: java.lang.RuntimeException: Unable to initialize the native Deep Learning backend: Could not initialize class deepwater.backends.mxnet.MXNetBackend$MXNetLoader #46

Open datalee opened 7 years ago

datalee commented 7 years ago

[2017-07-10 11:37:44] failure_details: Unable to initialize the native Deep Learning backend: null [2017-07-10 11:37:44] failure_stack_traces: java.lang.RuntimeException: Unable to initialize the native Deep Learning backend: null at hex.deepwater.DeepWaterModelInfo.setupNativeBackend(DeepWaterModelInfo.java:267) at hex.deepwater.DeepWaterModelInfo.(DeepWaterModelInfo.java:214) at hex.deepwater.DeepWaterModel.(DeepWaterModel.java:227) at hex.deepwater.DeepWater$DeepWaterDriver.buildModel(DeepWater.java:131) at hex.deepwater.DeepWater$DeepWaterDriver.computeImpl(DeepWater.java:118) at hex.ModelBuilder$Driver.compute2(ModelBuilder.java:173) at hex.deepwater.DeepWater$DeepWaterDriver.compute2(DeepWater.java:111) at water.H2O$H2OCountedCompleter.compute(H2O.java:1256) at jsr166y.CountedCompleter.exec(CountedCompleter.java:468) at jsr166y.ForkJoinTask.doExec(ForkJoinTask.java:263) at jsr166y.ForkJoinPool$WorkQueue.runTask(ForkJoinPool.java:974) at jsr166y.ForkJoinPool.runWorker(ForkJoinPool.java:1477) at jsr166y.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:104)

mdymczyk commented 7 years ago

@datalee Please provide:

1) your environment (OS, hardware)

2) which version are you using

3) example of failing code + data

datalee commented 7 years ago

1.environment ubuntu 16.04 LTS NVIDIA-SMI 375.66 CUDA Version 8.0.61 CUDNN 5.1

2.R VERSION R: wget http://s3.amazonaws.com/h2o-deepwater/public/nightly/ latest/h2o_3.13.0.tar.gz R CMD INSTALL h2o_3.13.0.tar.gz

3.demo https://github.com/h2oai/h2o-3/blob/master/examples/deeplearning/notebooks/deeplearning_grid_iris_R.ipynb

if i need to install mxnet/tensorflow alone

datalee commented 7 years ago

@mdymczyk how can i tell the h2o to find the mxnet/tensorflow?

datalee commented 7 years ago

2017-07-10 14:23:49.568855: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] Ignoring visible gpu device (device: 0, name: GeForce 820M, pci bus id: 0000:04:00.0) with Cuda compute capability 2.1. The minimum required Cuda capability is 3.0.

mdymczyk commented 7 years ago

@datalee as for MXNet I will have to try it myself, might be a bug - should work out of the box

As for TF - as the message says your GPU is too old, TF requires cards with CUDA capability >= 3.0