Question

我正在使用tensorflow-gpu 1.15和keras 2.3.1测试Yolo-v3（https://github.com/experiencor/keras-yolo3）。培训过程开始于：

runfile("train.py",'-c config.json')

以下是打印出来的消息：

Using TensorFlow backend.
WARNING:tensorflow:From train.py:40: The name tf.keras.backend.set_session is deprecated. Please use tf.compat.v1.keras.backend.set_session instead.

valid_annot_folder not exists. Spliting the trainining set.
Seen labels:    {'kangaroo': 266}

Given labels:   ['kangaroo']

Training on:    ['kangaroo']

WARNING:tensorflow:From C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\ops\resource_variable_ops.py:1630: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.
.....


Loading pretrained weights.

C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\callbacks\callbacks.py:998: UserWarning: `epsilon` argument is deprecated and will be removed, use `min_delta` instead.
  warnings.warn('`epsilon` argument is deprecated and '
Traceback (most recent call last):

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 1365, in _do_call
    return fn(*args)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 1348, in _run_fn
    self._extend_graph()

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 1388, in _extend_graph
    tf_session.ExtendSession(self._session)

InvalidArgumentError: Cannot assign a device for operation replica_0/lambda_1/Shape: {{node replica_0/lambda_1/Shape}} was explicitly assigned to /device:GPU:0 but available devices are [ /job:localhost/replica:0/task:0/device:CPU:0 ]. Make sure the device specification refers to a valid device.
     [[replica_0/lambda_1/Shape]]


During handling of the above exception, another exception occurred:

Traceback (most recent call last):

  File "train.py", line 305, in <module>
    _main_(args)

  File "train.py", line 282, in _main_
    max_queue_size   = 8

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\legacy\interfaces.py", line 91, in wrapper
    return func(*args, **kwargs)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\engine\training.py", line 1732, in fit_generator
    initial_epoch=initial_epoch)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\engine\training_generator.py", line 42, in fit_generator
    model._make_train_function()

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\engine\training.py", line 333, in _make_train_function
    **self._function_kwargs)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\backend\tensorflow_backend.py", line 3006, in function
    v1_variable_initialization()

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\backend\tensorflow_backend.py", line 420, in v1_variable_initialization
    session = get_session()

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\keras\backend\tensorflow_backend.py", line 385, in get_session
    return tf_keras_backend.get_session()

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\keras\backend.py", line 486, in get_session
    _initialize_variables(session)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\keras\backend.py", line 903, in _initialize_variables
    [variables_module.is_variable_initialized(v) for v in candidate_vars])

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 956, in run
    run_metadata_ptr)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 1180, in _run
    feed_dict_tensor, options, run_metadata)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 1359, in _do_run
    run_metadata)

  File "C:\Users\Dy\Anaconda3\envs\tf1x\lib\site-packages\tensorflow_core\python\client\session.py", line 1384, in _do_call
    raise type(e)(node_def, op, message)

InvalidArgumentError：无法为操作copy_0_0 / lambda_1 / Shape分配设备：节点copy_0_0 / lambda_1 / Shape（在C：\ Users \ Dy \ Anaconda3 \ envs \ tf1x \ lib \ site-packages \ tensorflow_core \ python \ framework中定义\ ops.py：1748）已明确分配给/ device：GPU：0，但可用设备为[/ job：localhost /副本0 / task：0 / device：CPU：0]。确保设备规格引用的是有效设备。 [[replica_0 / lambda_1 /形状]]

我不明白是什么原因导致了InvalidArgumentError。我的tensoflow-gpu安装不正确吗？还是在部署gpu时发生冲突？

Answer 1

尝试将“gpus”值更改为“0”，如果它是其他任何东西。如果您在 GPU 中执行，它应该可以工作。

InvalidArgumentError：无法为操作copy_0_0 / lambda_1 / Shape分配设备

1 个答案: