Keras: experimental_list_devices in tensorflow_backend.py (line 506)

[vgrankin1]

As we know experimental_list_devices is removed from tensorflow, so i got an error:
AttributeError: module 'tensorflow_core._api.v2.config' has no attribute 'experimental_list_devices'

when creating 'Sequential' model.
It is happened as i understood beacause tf.config.experimental_list_devices()

tf.__version__ is: 2.1.0
tf.keras.__version__ is: 2.2.4-tf

Kaggle notebook with error

[AlphaWu]

experimental_list_devices is deprecated in tf 2.1,use tf.config.list_logical_devices replaced.

`
def _get_available_gpus():
"""Get a list of available gpu devices (formatted as strings).

# Returns
    A list of available GPU devices.
"""
global _LOCAL_DEVICES
if _LOCAL_DEVICES is None:
    if _is_tf_1():
        devices = get_session().list_devices()
        _LOCAL_DEVICES = [x.name for x in devices]
    else:
        devices = tf.config.list_logical_devices()
        _LOCAL_DEVICES = [x.name for x in devices]
return [x for x in _LOCAL_DEVICES if 'device:gpu' in x.lower()]

`

[gagolucasm]

Same problem here, AlphaWu solution works. I have created a pull request

[DavidGervasio]

I solve this by installing an older version

[amirsayonara]

I solve this by installing an older version

which version of Keras is downgraded? and which version of Tensorflow?

[vgrankin1]

I solve this by installing an older version

which version of Keras is downgraded? and which version of Tensorflow?

Hello, I think better solution is an injection.

import tensorflow as tf
import keras.backend.tensorflow_backend as tfback
...ba bla
print("tf.__version__ is", tf.__version__)
print("tf.keras.__version__ is:", tf.keras.__version__)

def _get_available_gpus():
    """Get a list of available gpu devices (formatted as strings).

    # Returns
        A list of available GPU devices.
    """
    #global _LOCAL_DEVICES
    if tfback._LOCAL_DEVICES is None:
        devices = tf.config.list_logical_devices()
        tfback._LOCAL_DEVICES = [x.name for x in devices]
    return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]

tfback._get_available_gpus = _get_available_gpus

it is working fine for me, even in jupyter notebooks

[qNOVAq]

Had the same problem, fixed it with a simple argument
model.add(MaxPooling2D(2,2, dim_ordering='tf'))

dim_ordering ='th'
gave me the previous error, replacing it with 'tf' fixed it.

[muhammadhasnain12]

This code is working on my colab notebook

import tensorflow as tf
import keras.backend.tensorflow_backend as tfback
print("tf.__version__ is", tf.__version__)
print("tf.keras.__version__ is:", tf.keras.__version__)

def _get_available_gpus():
"""Get a list of available gpu devices (formatted as strings).

# Returns
    A list of available GPU devices.
"""
#global _LOCAL_DEVICES
if tfback._LOCAL_DEVICES is None:
    devices = tf.config.list_logical_devices()
    tfback._LOCAL_DEVICES = [x.name for x in devices]
return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]

tfback._get_available_gpus = _get_available_gpus

[JingCi-ZHu]

the problem has not solved when i train my net in the colab
504 if tfback._LOCAL_DEVICES is None:
505 devices = tf.config.list_logical_devices()
--> 506 tfback._LOCAL_DEVICES = [x.name for x in devices]
507 return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]
508

AttributeError: module 'tensorflow._api.v2.config' has no attribute 'experimental_list_devices'

[wissam-bouattou]

@JingCi-ZHu same for me , plz if you've found a solution can you share it with me ?

[nmodem2k]

what's the error in these?

fork11 = Convolution2D(nb_filters_1, nb_conv_init, nb_conv_init, activation="relu")(init)
fork12 = Convolution2D(nb_filters_1, nb_conv_init, nb_conv_init, activation="relu")(init)

showing error::AttributeError: module 'tensorflow_core._api.v2.config' has no attribute 'experimental_list_device

[wissam-bouattou]

@nmodem2k i tried the solution which is in this link and it works for me https://www.kaggle.com/vitost/kernel-keras-injection-sol might it helps u

[nmodem2k]

I am getting this error.plz help

[wissam-bouattou]

@nmodem2k add those 2 lignes before the function

 import tensorflow as tf
import keras.backend.tensorflow_backend as tfback

[JingCi-ZHu]

do you solve it now?

[wissam-bouattou]

@JingCi-ZHu for me it works , i followed the instructions given in this website https://www.kaggle.com/vitost/kernel-keras-injection-sol

[nmodem2k]

@JingCi-ZHu @wissam-bouattou yeah it worked for me too

[dagger112]

I solve this by installing an older version

i am having build problem with older version any suggestions?

[lwoog]

Use tf.keras instead of keras with tf 2.1.0

On May 18, 2020, at 4:14 PM, dagger112 [e-mail] wrote:

I solve this by installing an older version

i am having build problem with older version any suggestions?

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub https://github.com/keras-team/keras/issues/13684#issuecomment-630210229, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALL5BYDNORSDIOD2WVZ4BNTRSE7E5ANCNFSM4KFE7NMQ.

[qzylalala]

list_logical_devices()

It occurs another ERROR：'LogicalDevice' object has no attribute 'lower'

[biswajit310]

experimental_list_devices is deprecated in tf 2.1,use tf.config.list_logical_devices replaced.

`
def _get_available_gpus():
"""Get a list of available gpu devices (formatted as strings).

# Returns
    A list of available GPU devices.
"""
global _LOCAL_DEVICES
if _LOCAL_DEVICES is None:
    if _is_tf_1():
        devices = get_session().list_devices()
        _LOCAL_DEVICES = [x.name for x in devices]
    else:
        devices = tf.config.list_logical_devices()
        _LOCAL_DEVICES = [x.name for x in devices]
return [x for x in _LOCAL_DEVICES if 'device:gpu' in x.lower()]

`

I am still getting the same error after using "tf.config.list_logical_device". I am running it on google colab. Plz help me.

[KYHyeon]

I solved this problem using pip install -U sensorflow-gpu.

[vidhigandhi94]

list_logical_devices()

It occurs another ERROR：'LogicalDevice' object has no attribute 'lower'

I am also facing the same issue!! Did anyone else find a solution for this?