Merge branch 'tensorflow:master' into panoptic-deeplab-modeling

official/vision/projects/video_ssl/tasks/pretrain.py → official/projects/video_ssl/tasks/pretrain.py

@@ -20,13 +20,13 @@ import tensorflow as tf
 # pylint: disable=unused-import
 from official.core import input_reader
 from official.core import task_factory
-from official.vision.beta.modeling import factory_3d
-from official.vision.beta.tasks import video_classification
-from official.vision.projects.video_ssl.configs import video_ssl as exp_cfg
-from official.vision.projects.video_ssl.dataloaders import video_ssl_input
-from official.vision.projects.video_ssl.losses import losses
-from official.vision.projects.video_ssl.modeling import video_ssl_model
-# pylint: disable=unused-import
+from official.projects.video_ssl.configs import video_ssl as exp_cfg
+from official.projects.video_ssl.dataloaders import video_ssl_input
+from official.projects.video_ssl.losses import losses
+from official.projects.video_ssl.modeling import video_ssl_model
+from official.vision.modeling import factory_3d
+from official.vision.tasks import video_classification
+# pylint: enable=unused-import
 
 
 @task_factory.register_task_cls(exp_cfg.VideoSSLPretrainTask)

official/vision/projects/video_ssl/tasks/pretrain_test.py → official/projects/video_ssl/tasks/pretrain_test.py

@@ -22,12 +22,13 @@ import orbit
 import tensorflow as tf
 
 # pylint: disable=unused-import
+from official import vision
 from official.core import exp_factory
 from official.core import task_factory
 from official.modeling import optimization
-from official.vision import beta
-from official.vision.beta.dataloaders import tfexample_utils
-from official.vision.projects.video_ssl.tasks import pretrain
+from official.projects.video_ssl.tasks import pretrain
+from official.vision.dataloaders import tfexample_utils
+# pylint: enable=unused-import
 
 
 class VideoClassificationTaskTest(tf.test.TestCase):

official/vision/projects/video_ssl/train.py → official/projects/video_ssl/train.py

@@ -20,16 +20,16 @@ from absl import flags
 import gin
 
 # pylint: disable=unused-import
-from official.common import registry_imports
 from official.common import distribute_utils
 from official.common import flags as tfm_flags
 from official.core import task_factory
 from official.core import train_lib
 from official.core import train_utils
 from official.modeling import performance
-from official.vision.projects.video_ssl.modeling import video_ssl_model
-from official.vision.projects.video_ssl.tasks import linear_eval
-from official.vision.projects.video_ssl.tasks import pretrain
+from official.projects.video_ssl.modeling import video_ssl_model
+from official.projects.video_ssl.tasks import linear_eval
+from official.projects.video_ssl.tasks.google import pretrain
+from official.vision import registry_imports
 # pylint: disable=unused-import
 
 FLAGS = flags.FLAGS

official/projects/vit/modeling/nn_blocks.py

@@ -22,10 +22,15 @@ from official.vision.beta.modeling.layers.nn_layers import StochasticDepth
 class TransformerEncoderBlock(modeling.layers.TransformerEncoderBlock):
   """TransformerEncoderBlock layer with stochastic depth."""
 
-  def __init__(self, *args, stochastic_depth_drop_rate=0.0, **kwargs):
+  def __init__(self,
+               *args,
+               stochastic_depth_drop_rate=0.0,
+               return_attention=False,
+               **kwargs):
     """Initializes TransformerEncoderBlock."""
     super().__init__(*args, **kwargs)
     self._stochastic_depth_drop_rate = stochastic_depth_drop_rate
+    self._return_attention = return_attention
 
   def build(self, input_shape):
     if self._stochastic_depth_drop_rate:
@@ -73,8 +78,9 @@ class TransformerEncoderBlock(modeling.layers.TransformerEncoderBlock):
 
     if key_value is None:
       key_value = input_tensor
-    attention_output = self._attention_layer(
-        query=target_tensor, value=key_value, attention_mask=attention_mask)
+    attention_output, attention_scores = self._attention_layer(
+        query=target_tensor, value=key_value, attention_mask=attention_mask,
+        return_attention_scores=True)
     attention_output = self._attention_dropout(attention_output)
 
     if self._norm_first:
@@ -95,12 +101,19 @@ class TransformerEncoderBlock(modeling.layers.TransformerEncoderBlock):
     layer_output = self._output_dropout(layer_output)
 
     if self._norm_first:
+      if self._return_attention:
+        return source_attention_output + self._stochastic_depth(
+            layer_output, training=training), attention_scores
+      else:
         return source_attention_output + self._stochastic_depth(
             layer_output, training=training)
 
     # During mixed precision training, layer norm output is always fp32 for now.
     # Casts fp32 for the subsequent add.
     layer_output = tf.cast(layer_output, tf.float32)
-    return self._output_layer_norm(
-        layer_output +
-        self._stochastic_depth(attention_output, training=training))
+    if self._return_attention:
+      return self._output_layer_norm(layer_output + self._stochastic_depth(
+          attention_output, training=training)), attention_scores
+    else:
+      return self._output_layer_norm(layer_output + self._stochastic_depth(
+          attention_output, training=training))

official/projects/volumetric_models/configs/semantic_segmentation_3d.py

@@ -23,7 +23,7 @@ from official.modeling import hyperparams
 from official.modeling import optimization
 from official.projects.volumetric_models.configs import backbones
 from official.projects.volumetric_models.configs import decoders
-from official.vision.beta.configs import common
+from official.vision.configs import common
 
 
 @dataclasses.dataclass

official/projects/volumetric_models/dataloaders/segmentation_input_3d.py

@@ -16,8 +16,8 @@
 
 from typing import Any, Dict, Sequence, Tuple
 import tensorflow as tf
-from official.vision.beta.dataloaders import decoder
-from official.vision.beta.dataloaders import parser
+from official.vision.dataloaders import decoder
+from official.vision.dataloaders import parser
 
 
 class Decoder(decoder.Decoder):

official/projects/volumetric_models/dataloaders/segmentation_input_3d_test.py

@@ -20,7 +20,7 @@ from absl.testing import parameterized
 import tensorflow as tf
 
 from official.projects.volumetric_models.dataloaders import segmentation_input_3d
-from official.vision.beta.dataloaders import tfexample_utils
+from official.vision.dataloaders import tfexample_utils
 
 
 class InputReaderTest(parameterized.TestCase, tf.test.TestCase):

official/projects/volumetric_models/modeling/backbones/unet_3d.py

@@ -25,7 +25,7 @@ from typing import Any, Mapping, Sequence
 import tensorflow as tf
 from official.modeling import hyperparams
 from official.projects.volumetric_models.modeling import nn_blocks_3d
-from official.vision.beta.modeling.backbones import factory
+from official.vision.modeling.backbones import factory
 
 layers = tf.keras.layers
 

official/projects/volumetric_models/modeling/factory.py

@@ -21,8 +21,8 @@ import tensorflow as tf
 from official.modeling import hyperparams
 from official.projects.volumetric_models.modeling.decoders import factory as decoder_factory
 from official.projects.volumetric_models.modeling.heads import segmentation_heads_3d
-from official.vision.beta.modeling import segmentation_model
-from official.vision.beta.modeling.backbones import factory as backbone_factory
+from official.vision.modeling import segmentation_model
+from official.vision.modeling.backbones import factory as backbone_factory
 
 
 def build_segmentation_model_3d(

official/projects/volumetric_models/modeling/nn_blocks_3d.py

@@ -20,7 +20,7 @@ from typing import Sequence, Union
 import tensorflow as tf
 
 from official.modeling import tf_utils
-from official.vision.beta.modeling.layers import nn_layers
+from official.vision.modeling.layers import nn_layers
 
 
 @tf.keras.utils.register_keras_serializable(package='Vision')

official/projects/volumetric_models/modeling/segmentation_model_test.py

@@ -21,7 +21,7 @@ import tensorflow as tf
 from official.projects.volumetric_models.modeling import backbones
 from official.projects.volumetric_models.modeling import decoders
 from official.projects.volumetric_models.modeling.heads import segmentation_heads_3d
-from official.vision.beta.modeling import segmentation_model
+from official.vision.modeling import segmentation_model
 
 
 class SegmentationNetworkUNet3DTest(parameterized.TestCase, tf.test.TestCase):

official/projects/volumetric_models/serving/export_saved_model.py

@@ -43,7 +43,7 @@ from official.common import registry_imports  # pylint: disable=unused-import
 from official.core import exp_factory
 from official.modeling import hyperparams
 from official.projects.volumetric_models.serving import semantic_segmentation_3d
-from official.vision.beta.serving import export_saved_model_lib
+from official.vision.serving import export_saved_model_lib
 
 FLAGS = flags.FLAGS
 

official/projects/volumetric_models/serving/semantic_segmentation_3d.py

@@ -22,7 +22,7 @@ import tensorflow as tf
 from official.projects.volumetric_models.modeling import backbones
 from official.projects.volumetric_models.modeling import decoders
 from official.projects.volumetric_models.modeling import factory
-from official.vision.beta.serving import export_base
+from official.vision.serving import export_base
 
 
 class SegmentationModule(export_base.ExportModule):

official/projects/volumetric_models/tasks/semantic_segmentation_3d_test.py

@@ -30,7 +30,7 @@ from official.projects.volumetric_models.evaluation import segmentation_metrics
 from official.projects.volumetric_models.modeling import backbones
 from official.projects.volumetric_models.modeling import decoders
 from official.projects.volumetric_models.tasks import semantic_segmentation_3d as img_seg_task
-from official.vision.beta.dataloaders import tfexample_utils
+from official.vision.dataloaders import tfexample_utils
 
 
 class SemanticSegmentationTaskTest(tf.test.TestCase, parameterized.TestCase):

official/projects/volumetric_models/train.py

@@ -19,7 +19,7 @@ import gin  # pylint: disable=unused-import
 
 from official.common import flags as tfm_flags
 from official.projects.volumetric_models import registry_imports  # pylint: disable=unused-import
-from official.vision.beta import train
+from official.vision import train
 
 
 def main(_):

official/projects/volumetric_models/train_test.py

@@ -21,7 +21,7 @@ from absl import logging
 from absl.testing import flagsaver
 import tensorflow as tf
 from official.projects.volumetric_models import train as train_lib
-from official.vision.beta.dataloaders import tfexample_utils
+from official.vision.dataloaders import tfexample_utils
 
 FLAGS = flags.FLAGS
 

official/projects/yt8m/configs/yt8m.py

@@ -21,7 +21,7 @@ from official.core import config_definitions as cfg
 from official.core import exp_factory
 from official.modeling import hyperparams
 from official.modeling import optimization
-from official.vision.beta.configs import common
+from official.vision.configs import common
 
 FLAGS = flags.FLAGS
 

official/projects/yt8m/dataloaders/yt8m_input.py

@@ -27,9 +27,9 @@ from typing import Dict
 
 import tensorflow as tf
 from official.projects.yt8m.dataloaders import utils
-from official.vision.beta.configs import video_classification as exp_cfg
-from official.vision.beta.dataloaders import decoder
-from official.vision.beta.dataloaders import parser
+from official.vision.configs import video_classification as exp_cfg
+from official.vision.dataloaders import decoder
+from official.vision.dataloaders import parser
 
 
 def resize_axis(tensor, axis, new_size, fill_value=0):

official/projects/yt8m/train.py

@@ -21,7 +21,7 @@ from official.common import flags as tfm_flags
 from official.projects.yt8m.configs import yt8m
 from official.projects.yt8m.tasks import yt8m_task
 # pylint: enable=unused-import
-from official.vision.beta import train
+from official.vision import train
 
 
 if __name__ == '__main__':

official/projects/yt8m/train_test.py

@@ -22,7 +22,7 @@ from absl.testing import flagsaver
 import numpy as np
 import tensorflow as tf
 from official.projects.yt8m import train as train_lib
-from official.vision.beta.dataloaders import tfexample_utils
+from official.vision.dataloaders import tfexample_utils
 
 FLAGS = flags.FLAGS