E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303 on Docker Debian 9 #34

dan-developer · 2020-05-20T14:39:14Z

I have this error using debian on the docker, I would like to disable cuda support, is there how?

Below is the output:

Using TensorFlow backend.
^[[AUsing VGG-Face model backend and cosine distance.
2020-05-20 14:35:39.426369: E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303)
Verification: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 1.46it/s]
{'pair_1': {'distance': 0.23476946353912354,
'max_threshold_to_verify': 0.4,
'model': 'VGG-Face',
'similarity_metric': 'cosine',
'verified': True}}

serengil · 2020-05-20T14:52:08Z

Try to put the following code block in the first line of your program.

import os
os.environ["CUDA_VISIBLE_DEVICES"]=""

serengil · 2020-05-21T05:31:27Z

Alternatively, you can call the following command. It tries to allocate required GPU memory. If it cannot allocate, then continue with CPU.

DeepFace.allocateMemory()

dan-developer · 2020-05-21T13:14:14Z

I tried both methods, but the error doesn't go away:

Using TensorFlow backend.
Analyzing your system...
DeepFace will run on CPU
Using VGG-Face model backend and cosine distance.
2020-05-21 13:13:01.045449: E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303)
Verification: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00,  1.31it/s]
{'pair_1': {'distance': 0.29093843698501587,
            'max_threshold_to_verify': 0.4,
            'model': 'VGG-Face',
            'similarity_metric': 'cosine',
            'verified': True}}

The python code:

from pprint import pprint
import os
from deepface import DeepFace
from deepface.basemodels import VGGFace, OpenFace, Facenet, FbDeepFace

os.environ["CUDA_VISIBLE_DEVICES"]=""
DeepFace.allocateMemory()


obj = DeepFace.verify([
    ["db/01_eu.jpg", "./db/S5/1.tif"],

])

pprint(obj)

serengil · 2020-05-21T13:32:20Z

Even though you got an error, you can see the content of obj variable. It is interesting.

serengil · 2020-05-21T13:33:22Z

Could you try to install tensorflow instead of tensorflow-gpu

pip install tensorflow==1.9.0

dan-developer · 2020-05-21T13:36:35Z

Using TensorFlow backend.
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])
Traceback (most recent call last):
  File "teste-deepface.py", line 3, in <module>
    from deepface import DeepFace
  File "/usr/local/lib/python3.6/site-packages/deepface/DeepFace.py", line 1, in <module>
    from keras.preprocessing import image
  File "/usr/local/lib/python3.6/site-packages/keras/__init__.py", line 3, in <module>
    from . import utils
  File "/usr/local/lib/python3.6/site-packages/keras/utils/__init__.py", line 6, in <module>
    from . import conv_utils
  File "/usr/local/lib/python3.6/site-packages/keras/utils/conv_utils.py", line 9, in <module>
    from .. import backend as K
  File "/usr/local/lib/python3.6/site-packages/keras/backend/__init__.py", line 1, in <module>
    from .load_backend import epsilon
  File "/usr/local/lib/python3.6/site-packages/keras/backend/load_backend.py", line 90, in <module>
    from .tensorflow_backend import *
  File "/usr/local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py", line 54, in <module>
    get_graph = tf_keras_backend.get_graph
AttributeError: module 'tensorflow.python.keras.backend' has no attribute 'get_graph'

serengil · 2020-05-21T13:38:23Z

Could you run the following code block?

import tensorflow as tf
print(tf.__version__)
import keras
print(keras.__version__)

dan-developer · 2020-05-21T14:07:43Z

Continues the same error:

/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])
1.9.0
Using TensorFlow backend.
Traceback (most recent call last):
  File "teste-deepface.py", line 3, in <module>
    import keras
  File "/usr/local/lib/python3.6/site-packages/keras/__init__.py", line 3, in <module>
    from . import utils
  File "/usr/local/lib/python3.6/site-packages/keras/utils/__init__.py", line 6, in <module>
    from . import conv_utils
  File "/usr/local/lib/python3.6/site-packages/keras/utils/conv_utils.py", line 9, in <module>
    from .. import backend as K
  File "/usr/local/lib/python3.6/site-packages/keras/backend/__init__.py", line 1, in <module>
    from .load_backend import epsilon
  File "/usr/local/lib/python3.6/site-packages/keras/backend/load_backend.py", line 90, in <module>
    from .tensorflow_backend import *
  File "/usr/local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py", line 54, in <module>
    get_graph = tf_keras_backend.get_graph
AttributeError: module 'tensorflow.python.keras.backend' has no attribute 'get_graph'

serengil · 2020-05-22T06:08:04Z

I cannot see your keras versions. If line 3 has keras import, then your keras installation is problematic. Try to reinstall the package.

pip uninstall tensorflow
pip install tensorflow==1.9.9
pip uninstall keras
pip install keras==2.2.0

dan-developer · 2020-05-22T18:28:44Z

ERROR: Could not find a version that satisfies the requirement tensorflow==1.9.9

serengil · 2020-05-22T20:05:22Z

My fault. The correct version are below.

pip uninstall tensorflow
pip install tensorflow==1.9.0
pip uninstall keras
pip install keras==2.2.0

dan-developer · 2020-05-22T21:00:24Z

/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.

dan-developer · 2020-05-22T21:00:35Z

Using TensorFlow backend.
Directory /root /.deepface created
Directory /root /.deepface/weights created
Analyzing your system...
DeepFace will run on CPU
Using VGG-Face model backend and cosine distance.
vgg_face_weights.h5 will be downloaded...
Access denied with the following error:

    Too many users have viewed or downloaded this file recently. Please
    try accessing the file again later. If the file you are trying to
    access is particularly large or is shared with many people, it may
    take up to 24 hours to be able to view or download the file. If you
    still can't access a file after 24 hours, contact your domain
    administrator.

You may still be able to access the file from the browser:

     https://drive.google.com/uc?id=1CPSeum3HpopfomUEK1gybeuIVoeJT_Eo

serengil · 2020-05-23T07:42:09Z

Oke, download this file manually and copy it to the /root /.deepface/weights folder.

rockstarvibu · 2020-05-24T18:00:11Z

Continues the same error:

/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
/usr/local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])
1.9.0
Using TensorFlow backend.
Traceback (most recent call last):
  File "teste-deepface.py", line 3, in <module>
    import keras
  File "/usr/local/lib/python3.6/site-packages/keras/__init__.py", line 3, in <module>
    from . import utils
  File "/usr/local/lib/python3.6/site-packages/keras/utils/__init__.py", line 6, in <module>
    from . import conv_utils
  File "/usr/local/lib/python3.6/site-packages/keras/utils/conv_utils.py", line 9, in <module>
    from .. import backend as K
  File "/usr/local/lib/python3.6/site-packages/keras/backend/__init__.py", line 1, in <module>
    from .load_backend import epsilon
  File "/usr/local/lib/python3.6/site-packages/keras/backend/load_backend.py", line 90, in <module>
    from .tensorflow_backend import *
  File "/usr/local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py", line 54, in <module>
    get_graph = tf_keras_backend.get_graph
AttributeError: module 'tensorflow.python.keras.backend' has no attribute 'get_graph'

in api.py at line 83 change from graph = tf.get_default_graph() to graph = tf.compat.v1.get_default_graph()

it will run.. getting same error but now i solved it...

serengil · 2020-05-25T07:01:33Z

in the exception message, it says 'module 'tensorflow.python.keras.backend'. You are running tensorflow 2.0? Because this seems to come from 2.0.

In previous message, I mentioned that you should have the following versions.

pip install tensorflow==1.9.0
pip install keras==2.2.0

serengil · 2020-06-15T10:40:10Z

no response

serengil closed this as completed Jun 15, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303 on Docker Debian 9 #34

E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303 on Docker Debian 9 #34

dan-developer commented May 20, 2020

serengil commented May 20, 2020

serengil commented May 21, 2020

dan-developer commented May 21, 2020 •

edited

serengil commented May 21, 2020

serengil commented May 21, 2020

dan-developer commented May 21, 2020

serengil commented May 21, 2020

dan-developer commented May 21, 2020

serengil commented May 22, 2020

dan-developer commented May 22, 2020

serengil commented May 22, 2020

dan-developer commented May 22, 2020

dan-developer commented May 22, 2020

serengil commented May 23, 2020

rockstarvibu commented May 24, 2020

serengil commented May 25, 2020

serengil commented Jun 15, 2020

E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303 on Docker Debian 9 #34

E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: UNKNOWN ERROR (303 on Docker Debian 9 #34

Comments

dan-developer commented May 20, 2020

serengil commented May 20, 2020

serengil commented May 21, 2020

dan-developer commented May 21, 2020 • edited

serengil commented May 21, 2020

serengil commented May 21, 2020

dan-developer commented May 21, 2020

serengil commented May 21, 2020

dan-developer commented May 21, 2020

serengil commented May 22, 2020

dan-developer commented May 22, 2020

serengil commented May 22, 2020

dan-developer commented May 22, 2020

dan-developer commented May 22, 2020

serengil commented May 23, 2020

rockstarvibu commented May 24, 2020

serengil commented May 25, 2020

serengil commented Jun 15, 2020

dan-developer commented May 21, 2020 •

edited