Python中TypeError: __init__() missing 2 required positional arguments

In Python TypeError: __init __ () missing 2 required positional arguments [Problems Points: 50 minutes]

hhu_luqi

Bbs1

Knot tie 0%

I think tensorflow training on the code on GitHub Mnist data set run, run in an environment of Python 3.5.2, on the Spyder.
Code files found
https://github.com/tensorflow/models/blob/master/official/mnist/mnist.py
run prompt TypeError mnist.py appear: __init __ () missing 2 required positional arguments: 'inputs' and' outputs '
Help my God big problem lie? How to correct?

Red arrow line is the IPython console prompt error.

 0 2018-02-28 21:16:17

Replies  4  Look landlord  references  to report  the landlord

Chaos crocodile

Bbs7

Blank Blank Blank

First, that I have never used TF
only just seen a moment TF in github source.
Model is inherited from the Network.
https://github.com/tensorflow/tensorflow/blob/master/tensorflow/python/keras/_impl/keras/engine/network.py

Super (Model, Self) .__ the init __ () There is no parameter to
is called
self. _init_subclassed_network (** kwargs)
 

Python code?

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

class Network(base_layer.Layer):

  """A `Network` is a composition of layers.

  It is the topological form of a "model". A `Model`

  is simply a `Network` with added training routines.

  """

 

  def __init__(self*args, **kwargs):  # pylint: disable=super-init-not-called

    # Signature detection

    if (len(args) == 2 or

        len(args) == 1 and 'outputs' in kwargs or

        'inputs' in kwargs and 'outputs' in kwargs):

      # Graph network

      self._init_graph_network(*args, **kwargs)

    else:

      # Subclassed network

      self._init_subclassed_network(**kwargs)



Proposed landlord first thing to learn python language based on it, to reach at least Intermediate level to start playing this.

1 2018-03-01 10:28:00

Look at TA  quote  Report  # 1 score 0

Solve the error: test_create_ssd_models_from_config + missing 2 required positional arguments : 'inputs' and 'out' if you're doing TensorFlow of target detection API interface debugging, error: ERROR: test_create_ssd_models_from_config + missing 2 required positional arguments: 'inputs' and 'out' please detects the version of your tensorflow and protoc being given reasons: version 1.5 ------ protoc tensorflow version 3.3 ...

hhu_luqi

Bbs1

Thank you, friends advice, go back for this tutorial classes and inheritance;
call self._init_subclassed_network (** kwargs), why Ah error?

0 2018-03-01 22:12:33

Look at TA  reference  report  # 2 score of 0

__init __ () missing 1 required positional argument: 'on_delete' learning django model, created two database tables. A second reference to the first table id as a foreign key, a reported __init __ () missing 1 required positional argument: 'on_delete' errors. Puzzled, and then query the information found then need a on_delete parameters, plus a good advice then, in which enter cmd command python manage.py mak

sinat_39619444

Bbs1

It should be tensorflow version of the reason, tf 1.6 under the API does not allow subclassing. GitHub on someone answers https://github.com/tensorflow/models/issues/3755, hope useful to you.

1 2018-05-24 15:49:26

Look at TA  reference  report  # 3 score 0

being given pip install TypeError: __call __ () missing 1 required positional argument: 'name' to solve the most recently used in Ubuntu14.04 time pip3 installation package appeared the following error: TypeError: __call __ () missing 1 required positional argument: 'name' ,

Published 91 original articles · won praise 47 · views 90000 +

Guess you like

Origin blog.csdn.net/qq_30007885/article/details/102564035