update/optimization_and_modeling->[update] optimzation.py and modeling.py for tensorflow 2x. #1414
+48
−44
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
MOTIVATION PR to resolve `AttributeError: module 'tensorflow._api.v2.train' has no attribute 'Optimizer'` `optimization.py`: - change `tf.train.Optimizer` -> `tf.compat.v1.train.Optimizer` due to tensorflow 2x. src: [ https://www.tensorflow.org/api_docs/python/tf/compat/v1 ] `modeling.py`: (due to changes made in `optimization.py`) - add `@tf.function` decorator above `get_shape_list(...)` due to eager execution enabled by default in tensorflow 2x. src: [ https://www.tensorflow.org/guide/function ] - change `tf.get_variable` -> `tf.compat.v1.get_variable`. - change `tf.variable_score` -> `tf.compat.v1.variable_score`. - change `tf.truncated_normal_initializer` -> `tf.compat.v1.truncated_normal_initializer`. - change `tf.assert_less_equal` -> `tf.compat.v1.assert_less_equal`. - change `tf.get_variable_scope` -> `tf.compat.v1.get_variable_scope`. src: [ https://www.tensorflow.org/api_docs/python/tf/compat/v1 ] - change `tf.layers.dense` -> `tf.keras.layers.Dense` due to Keras 3. - change `tf.contrib.layers.layer_norm` -> `tf.keras.layers.LayerNormalization`. src: [ https://www.tensorflow.org/api_docs/python/tf/keras/layers/Dense ] src: [ https://stackoverflow.com/a/62357941/11492382 ] CONFIGURATION software : Windows 11 Home [21H2]. WSL version 2.2.4.0; openSUSE tumbleweed 20240629 hardware : 64-bit operating system, x64-based processor conda_env : cudnn 8.9.7.29 conda-forge python 3.12.2 conda-forge tensorflow 2.16.1 conda-forge Reported-by : Майкл Шодеке Signed-off-by: Майкл Шодеке
p0lyMth
force-pushed
the
update/optimization_and_modeling
branch
from
65a9d49
to
938a485
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
MOTIVATION
PR to resolve AttributeError: module 'tensorflow._api.v2.train' has no attribute 'Optimizer'
optimization.py:tf.train.Optimizer->tf.compat.v1.train.Optimizerdue to tensorflow 2x. src: [ https://www.tensorflow.org/api_docs/python/tf/compat/v1 ]modeling.py: (due to changes made inoptimization.py)add
@tf.functiondecorator aboveget_shape_list(...)due to eager execution enabled by default in tensorflow 2x. src: [ https://www.tensorflow.org/guide/function ]change
tf.get_variable->tf.compat.v1.get_variable.change
tf.variable_score->tf.compat.v1.variable_score.change
tf.truncated_normal_initializer->tf.compat.v1.truncated_normal_initializer.change
tf.assert_less_equal->tf.compat.v1.assert_less_equal.change
tf.get_variable_scope->tf.compat.v1.get_variable_scope. src: [ https://www.tensorflow.org/api_docs/python/tf/compat/v1 ]change
tf.layers.dense->tf.keras.layers.Densedue to Keras 3.change
tf.contrib.layers.layer_norm->tf.keras.layers.LayerNormalization. src: [ https://www.tensorflow.org/api_docs/python/tf/keras/layers/Dense ] src: [ https://stackoverflow.com/a/62357941/11492382 ]CONFIGURATION
Reported-by : Майкл Шодеке
Signed-off-by: Майкл Шодеке