tf_upgrade_v2 for network/classification/resnet.py

[ytfksw]

What this patch does to fix the issue.

I converted only the blueoil/network/classification/resnet.py with the command line tool tf_upgrade_v2 to run on tensorflow2 because there are a lot of targets in the network folder.

Here's what I did:

install newest tenosrflow (tensorflow-gpu==1.5.2)
Run the upgrade script tf_upgrade_v2
Check the upgrade report for warnings and errors
Run test (make test)

Link to any relevant issues or pull requests.

• Consider upgrading to TensorFlow 2.0 #479
• https://www.tensorflow.org/guide/upgrade

[ytfksw]

This is the whole convert log:

TensorFlow 2.0 Upgrade Script
-----------------------------
Converted 1 files
Detected 0 issues that require attention
--------------------------------------------------------------------------------
================================================================================
Detailed log follows:

================================================================================
--------------------------------------------------------------------------------
Processing file 'networks/classification/resnet.py'
 outputting to 'networks/classification/resnet.py'
--------------------------------------------------------------------------------

92:25: INFO: Added keywords to args of function 'tf.nn.avg_pool'
92:25: INFO: Renamed keyword argument for tf.nn.avg_pool from value to input
92:25: INFO: Renamed 'tf.nn.avg_pool' to 'tf.nn.avg_pool2d'
98:25: INFO: Added keywords to args of function 'tf.pad'
173:21: INFO: Renamed 'tf.Print' to 'tf.compat.v1.Print'
173:39: INFO: Added keywords to args of function 'tf.shape'
173:57: INFO: Added keywords to args of function 'tf.argmax'
174:22: INFO: Renamed 'tf.Print' to 'tf.compat.v1.Print'
174:41: INFO: Added keywords to args of function 'tf.shape'
174:60: INFO: Added keywords to args of function 'tf.argmax'
176:25: INFO: Added keywords to args of function 'tf.reduce_sum'
181:29: INFO: Added keywords to args of function 'tf.reduce_mean'
--------------------------------------------------------------------------------

[ytfksw]

tf_upgrade_v2 sometimes changed argment to keyword argument.
I deleted the keyword argument and checking it one by one (f172aa5), because the arguments become meaninglessly long .

Because API has changed by changing from tf1 to tf2.
It seems to have been converted to a keyword argument to avoid misunderstanding.

avg_pool
tf1:https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/avg_pool

tf.nn.avg_pool(
    value,
    ksize,
    strides,
    padding,
    data_format='NHWC',
    name=None,
    input=None
)

tf2:https://www.tensorflow.org/api_docs/python/tf/nn/avg_pool2d

tf.nn.avg_pool2d(
    input, ksize, strides, padding, data_format='NHWC', name=None
)

tf.pad
tf1:https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/pad

tf.pad(
    tensor,
    paddings,
    mode='CONSTANT',
    name=None,
    constant_values=0
)

tf2:https://www.tensorflow.org/api_docs/python/tf/pad

tf.shape(
    tensor, paddings, mode='CONSTANT', constant_values=0, name=None
)

tf.shape
tf1:https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/shape

tf.shape(
    input,
    name=None,
    out_type=tf.dtypes.int32
)

tf2:https://www.tensorflow.org/api_docs/python/tf/shape

tf.shape(
    input, out_type=tf.dtypes.int32, name=None
)

tf.argmax

tf1:https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/math/argmax

tf.math.argmax(
    input,
    axis=None,
    name=None,
    dimension=None,
    output_type=tf.dtypes.int64
)

tf2:https://www.tensorflow.org/api_docs/python/tf/math/argmax

tf.math.argmax(
    input, axis=None, output_type=tf.dtypes.int64, name=None
)

tf.reduce_sum
tf1:https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/math/reduce_sum

tf.math.reduce_sum(
    input_tensor,
    axis=None,
    keepdims=None,
    name=None,
    reduction_indices=None,
    keep_dims=None
)

tf2:https://www.tensorflow.org/api_docs/python/tf/math/reduce_sum

tf.math.reduce_sum(
    input_tensor, axis=None, keepdims=False, name=None
)

tf.reduce_mean
tf1:https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/math/reduce_mean

tf.math.reduce_mean(
    input_tensor,
    axis=None,
    keepdims=None,
    name=None,
    reduction_indices=None,
    keep_dims=None
)

tf2:https://www.tensorflow.org/api_docs/python/tf/math/reduce_mean

tf.math.reduce_mean(
    input_tensor, axis=None, keepdims=False, name=None
)

[bo-code-review-bot]

This PR needs Approvals as follows.

• Ownership Approval for / from iizukak, tkng, ruimashita
• Readability Approval for Python from tkng, tsawada, tfujiwar

Please choose reviewers and requet reviews!

Click to see how to approve each reviews

You can approve this PR by triggered comments as follows.

• Approve all reviews requested to you (readability and ownership) and LGTM review
  Approval, LGTM

• Approve all ownership reviews
  Ownership Approval or OA

• Approve all readability reviews
  Readability Approval or RA

• Approve specified review targets

  • Example of Ownership Reviewer of /: Ownership Approval for / or OA for /
  • Example of Readability Reviewer of Python: Readability Approval for Python or RA for Python

• Approve LGTM review
  LGTM

See all trigger comments

Please replace [Target] to review target

• Ownership Approval
  • Ownership Approval for [Target]
  • OA for [Target]
  • Ownership Approval
  • OA
  • Approval
• Readability Approval
  • Readability Approval for [Target]
  • RA for [Target]
  • [Target] Readability Approval
  • [Target] RA
  • Readability Approval
  • RA
  • Approval
• LGTM
  • LGTM
  • lgtm

[iizukak]

OA

[tfujiwar]

What about this? It's more consistent with the other function calls in this file.

labels = tf.compat.v1.Print(
    labels,
    [tf.shape(labels), tf.argmax(labels, axis=1)],
    message="labels:",
    summarize=200
)

[ytfksw]

Thank you. OK. I fixed it for consistency!

[tfujiwar]

RA

[iizukak]

/ready

[bo-mergebot]

⏳Merge job is queued...