CUDA报错: Cannot create Cublas handle. Cublas won't be available. 以及:Check failed: status == CUBLAS

发布时间:2024-12-10 20:40

aita@aita-Alienware-Area-51-R5:~/AITA2/daisida/ssd-github/caffe$ make runtest -j8

.build_release/tools/caffe

caffe: command line brew

usage: caffe <command> <args>

commands:

  train           train or finetune a model

  test            score a model

  device_query    show GPU diagnostic information

  time            benchmark model execution time

  Flags from tools/caffe.cpp:

    -gpu (Optional; run in GPU mode on given device IDs separated by ','.Use

      '-gpu all' to run on all available GPUs. The effective training batch

      size is multiplied by the number of devices.) type: string default: ""

    -iterations (The number of iterations to run.) type: int32 default: 50

    -level (Optional; network level.) type: int32 default: 0

    -model (The model definition protocol buffer text file.) type: string

      default: ""

    -phase (Optional; network phase (TRAIN or TEST). Only used for 'time'.)

      type: string default: ""

    -sighup_effect (Optional; action to take when a SIGHUP signal is received:

      snapshot, stop or none.) type: string default: "snapshot"

    -sigint_effect (Optional; action to take when a SIGINT signal is received:

      snapshot, stop or none.) type: string default: "stop"

    -snapshot (Optional; the snapshot solver state to resume training.)

      type: string default: ""

    -solver (The solver definition protocol buffer text file.) type: string

      default: ""

    -stage (Optional; network stages (not to be confused with phase), separated

      by ','.) type: string default: ""

    -weights (Optional; the pretrained weights to initialize finetuning,

      separated by ','. Cannot be set simultaneously with snapshot.)

      type: string default: ""

.build_release/test/test_all.testbin 0 --gtest_shuffle

Cuda number of devices: 3

Setting to use device 0

Current device id: 0

Current device name: GeForce GTX 1080 Ti

Note: Randomizing tests' orders with a seed of 48866 .

[==========] Running 2361 tests from 309 test cases.

[----------] Global test environment set-up.

[----------] 7 tests from DetectionOutputLayerTest/2, where TypeParam = caffe::GPUDevice<float>

[ RUN      ] DetectionOutputLayerTest/2.TestForwardShareLocationTopK

E0103 00:37:53.042623 19470 common.cpp:113] Cannot create Cublas handle. Cublas won't be available.

[       OK ] DetectionOutputLayerTest/2.TestForwardShareLocationTopK (219 ms)

[ RUN      ] DetectionOutputLayerTest/2.TestForwardNoShareLocationNeg0TopK

[       OK ] DetectionOutputLayerTest/2.TestForwardNoShareLocationNeg0TopK (2 ms)

[ RUN      ] DetectionOutputLayerTest/2.TestSetup

[       OK ] DetectionOutputLayerTest/2.TestSetup (1 ms)

[ RUN      ] DetectionOutputLayerTest/2.TestForwardNoShareLocationNeg0

[       OK ] DetectionOutputLayerTest/2.TestForwardNoShareLocationNeg0 (2 ms)

[ RUN      ] DetectionOutputLayerTest/2.TestForwardNoShareLocation

[       OK ] DetectionOutputLayerTest/2.TestForwardNoShareLocation (2 ms)

[ RUN      ] DetectionOutputLayerTest/2.TestForwardShareLocation

[       OK ] DetectionOutputLayerTest/2.TestForwardShareLocation (1 ms)

[ RUN      ] DetectionOutputLayerTest/2.TestForwardNoShareLocationTopK

[       OK ] DetectionOutputLayerTest/2.TestForwardNoShareLocationTopK (2 ms)

[----------] 7 tests from DetectionOutputLayerTest/2 (229 ms total)

[----------] 2 tests from EuclideanLossLayerTest/2, where TypeParam = caffe::GPUDevice<float>

[ RUN      ] EuclideanLossLayerTest/2.TestGradient

F0103 00:37:53.068140 19470 math_functions.cu:110] Check failed: status == CUBLAS_STATUS_SUCCESS (1 vs. 0)  CUBLAS_STATUS_NOT_INITIALIZED

*** Check failure stack trace: ***

    @     0x7f9210daf5cd  google::LogMessage::Fail()

    @     0x7f9210db1433  google::LogMessage::SendToLog()

    @     0x7f9210daf15b  google::LogMessage::Flush()

    @     0x7f9210db1e1e  google::LogMessageFatal::~LogMessageFatal()

    @     0x7f920c7ad43a  caffe::caffe_gpu_dot<>()

    @     0x7f920c7ec7c3  caffe::EuclideanLossLayer<>::Forward_gpu()

    @           0x48ae96  caffe::Layer<>::Forward()

    @           0x48d445  caffe::GradientChecker<>::CheckGradientSingle()

    @           0x4aea53  caffe::GradientChecker<>::CheckGradientExhaustive()

    @           0x848f0c  caffe::EuclideanLossLayerTest_TestGradient_Test<>::TestBody()

    @           0xa17c23  testing::internal::HandleExceptionsInMethodIfSupported<>()

    @           0xa1123a  testing::Test::Run()

    @           0xa11388  testing::TestInfo::Run()

    @           0xa11465  testing::TestCase::Run()

    @           0xa1273f  testing::internal::UnitTestImpl::RunAllTests()

    @           0xa12a63  testing::UnitTest::Run()

    @           0x47a98d  main

    @     0x7f920ba46830  __libc_start_main

    @           0x483b49  _start

    @              (nil)  (unknown)

Makefile:526: recipe for target 'runtest' failed

make: *** [runtest] Aborted (core dumped)

网址:CUDA报错: Cannot create Cublas handle. Cublas won't be available. 以及:Check failed: status == CUBLAS http://c.mxgxt.com/news/view/132684

相关内容

[Err] 1064
This site can’t be reached
check一下 好帅
Apache Tomcat/7.0.11
Non 日历 2024「10月服装fashion check」 Non
体育明星代言的风险及防范措施
哈哈哈真·笑死! 当你唱错Die With A Smile和声be like
外贸大神私密分享:外贸完整流程中常用的邮件模板
2009020123 张敏 影视明星危机公关.doc
space.tv.cctv.com

随便看看