tensorflow出现这种错误是怎么回事？

Question

2021-05-24 15:39:43.417723: E tensorflow/stream_executor/cuda/cuda_blas.cc:226] failed to create cublas handle: CUBLAS_STATUS_NOT_INITIALIZED
Traceback (most recent call last):
  File "C:/Repositories/My_Deep_Learning_TensorFlow/cifar_100.py", line 122, in <module>
    main()
  File "C:/Repositories/My_Deep_Learning_TensorFlow/cifar_100.py", line 85, in main
    out = conv_net(x)  # [b, 32, 32, 3] => [b, 1, 1, 512]
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\base_layer.py", line 1012, in __call__
    outputs = call_fn(inputs, *args, **kwargs)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\sequential.py", line 375, in call
    return super(Sequential, self).call(inputs, training=training, mask=mask)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\functional.py", line 425, in call
    inputs, training=training, mask=mask)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\functional.py", line 560, in _run_internal_graph
    outputs = node.layer(*args, **kwargs)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\keras\engine\base_layer.py", line 1012, in __call__
    outputs = call_fn(inputs, *args, **kwargs)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\keras\layers\convolutional.py", line 248, in call
    outputs = self._convolution_op(inputs, self.kernel)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\util\dispatch.py", line 201, in wrapper
    return target(*args, **kwargs)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 1020, in convolution_v2
    name=name)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 1150, in convolution_internal
    name=name)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 2604, in _conv2d_expanded_batch
    name=name)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\ops\gen_nn_ops.py", line 931, in conv2d
    _ops.raise_from_not_ok_status(e, name)
  File "C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py", line 6862, in raise_from_not_ok_status
    six.raise_from(core._status_to_exception(e.code, message), None)
  File "<string>", line 3, in raise_from
tensorflow.python.framework.errors_impl.NotFoundError: No algorithm worked! [Op:Conv2D]

 

win10，Tensorflow2.4.0，Python3.7，cuda11.0

Answer 1

因为显存不够，降低batchsize即可。

参考(1条消息) tensorflow训练3dcnn报错：NotFoundError: No algorithm worked!_今天又是不求上进的一天的博客-CSDN博客

Answer 2

不行啊，各种方法都尝试了，降低batchsize没用

Answer 3

没事，我再找找方法

Answer 4

好的，谢谢