Merge branch 'master' of https://github.com/tensorflow/models into AXg

official/nlp/train_ctl_continuous_finetune.py

@@ -14,27 +14,16 @@
 # limitations under the License.
 # ==============================================================================
 """TFM continuous finetuning+eval training driver."""
-import gc
-import os
-import time
-from typing import Any, Mapping, Optional
-
 from absl import app
 from absl import flags
-from absl import logging
 import gin
-import tensorflow as tf
 
 # pylint: disable=unused-import
 from official.common import registry_imports
 # pylint: enable=unused-import
-from official.common import distribute_utils
 from official.common import flags as tfm_flags
-from official.core import config_definitions
-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.nlp import continuous_finetune_lib
 
 FLAGS = flags.FLAGS
 
@@ -44,140 +33,15 @@ flags.DEFINE_integer(
     help='The number of total training steps for the pretraining job.')
 
 
-def run_continuous_finetune(
-    mode: str,
-    params: config_definitions.ExperimentConfig,
-    model_dir: str,
-    run_post_eval: bool = False,
-    pretrain_steps: Optional[int] = None,
-) -> Mapping[str, Any]:
-  """Run modes with continuous training.
-
-  Currently only supports continuous_train_and_eval.
-
-  Args:
-    mode: A 'str', specifying the mode. continuous_train_and_eval - monitors a
-      checkpoint directory. Once a new checkpoint is discovered, loads the
-      checkpoint, finetune the model by training it (probably on another dataset
-      or with another task), then evaluate the finetuned model.
-    params: ExperimentConfig instance.
-    model_dir: A 'str', a path to store model checkpoints and summaries.
-    run_post_eval: Whether to run post eval once after training, metrics logs
-      are returned.
-    pretrain_steps: Optional, the number of total training steps for the
-      pretraining job.
-
-  Returns:
-    eval logs: returns eval metrics logs when run_post_eval is set to True,
-      othewise, returns {}.
-  """
-
-  assert mode == 'continuous_train_and_eval', (
-      'Only continuous_train_and_eval is supported by continuous_finetune. '
-      'Got mode: {}'.format(mode))
-
-  # Sets mixed_precision policy. Using 'mixed_float16' or 'mixed_bfloat16'
-  # can have significant impact on model speeds by utilizing float16 in case of
-  # GPUs, and bfloat16 in the case of TPUs. loss_scale takes effect only when
-  # dtype is float16
-  if params.runtime.mixed_precision_dtype:
-    performance.set_mixed_precision_policy(params.runtime.mixed_precision_dtype,
-                                           params.runtime.loss_scale)
-  distribution_strategy = distribute_utils.get_distribution_strategy(
-      distribution_strategy=params.runtime.distribution_strategy,
-      all_reduce_alg=params.runtime.all_reduce_alg,
-      num_gpus=params.runtime.num_gpus,
-      tpu_address=params.runtime.tpu)
-
-  retry_times = 0
-  while not tf.io.gfile.isdir(params.task.init_checkpoint):
-    # Wait for the init_checkpoint directory to be created.
-    if retry_times >= 60:
-      raise ValueError(
-          'ExperimentConfig.task.init_checkpoint must be a directory for '
-          'continuous_train_and_eval mode.')
-    retry_times += 1
-    time.sleep(60)
-
-  summary_writer = tf.summary.create_file_writer(
-      os.path.join(model_dir, 'eval'))
-
-  global_step = 0
-
-  def timeout_fn():
-    if pretrain_steps and global_step < pretrain_steps:
-      # Keeps waiting for another timeout period.
-      logging.info(
-          'Continue waiting for new checkpoint as current pretrain '
-          'global_step=%d and target is %d.', global_step, pretrain_steps)
-      return False
-    # Quits the loop.
-    return True
-
-  for pretrain_ckpt in tf.train.checkpoints_iterator(
-      checkpoint_dir=params.task.init_checkpoint,
-      min_interval_secs=10,
-      timeout=params.trainer.continuous_eval_timeout,
-      timeout_fn=timeout_fn):
-    with distribution_strategy.scope():
-      global_step = train_utils.read_global_step_from_checkpoint(pretrain_ckpt)
-
-    if params.trainer.best_checkpoint_export_subdir:
-      best_ckpt_subdir = '{}_{}'.format(
-          params.trainer.best_checkpoint_export_subdir, global_step)
-      params_replaced = params.replace(
-          task={'init_checkpoint': pretrain_ckpt},
-          trainer={'best_checkpoint_export_subdir': best_ckpt_subdir})
-    else:
-      params_replaced = params.replace(task={'init_checkpoint': pretrain_ckpt})
-    params_replaced.lock()
-    logging.info('Running finetuning with params: %s', params_replaced)
-
-    with distribution_strategy.scope():
-      task = task_factory.get_task(params_replaced.task, logging_dir=model_dir)
-
-    _, eval_metrics = train_lib.run_experiment(
-        distribution_strategy=distribution_strategy,
-        task=task,
-        mode='train_and_eval',
-        # replace params.task.init_checkpoint to make sure that we load
-        # exactly this pretrain checkpoint.
-        params=params_replaced,
-        model_dir=model_dir,
-        run_post_eval=True,
-        save_summary=False)
-    logging.info('Evaluation finished. Pretrain global_step: %d', global_step)
-    train_utils.write_json_summary(model_dir, global_step, eval_metrics)
-
-    if not os.path.basename(model_dir):  # if model_dir.endswith('/')
-      summary_grp = os.path.dirname(model_dir) + '_' + task.name
-    else:
-      summary_grp = os.path.basename(model_dir) + '_' + task.name
-    summaries = {}
-    for name, value in eval_metrics.items():
-      summaries[summary_grp + '/' + name] = value
-    train_utils.write_summary(summary_writer, global_step, summaries)
-
-    train_utils.remove_ckpts(model_dir)
-    # In TF2, the resource life cycle is bound with the python object life
-    # cycle. Force trigger python garbage collection here so those resources
-    # can be deallocated in time, so it doesn't cause OOM when allocating new
-    # objects.
-    # TODO(b/169178664): Fix cycle reference in Keras model and revisit to see
-    # if we need gc here.
-    gc.collect()
-
-  if run_post_eval:
-    return eval_metrics
-  return {}
-
-
 def main(_):
+  # TODO(b/177863554): consolidate to nlp/train.py
   gin.parse_config_files_and_bindings(FLAGS.gin_file, FLAGS.gin_params)
   params = train_utils.parse_configuration(FLAGS)
   model_dir = FLAGS.model_dir
   train_utils.serialize_config(params, model_dir)
-  run_continuous_finetune(FLAGS.mode, params, model_dir, FLAGS.pretrain_steps)
+  continuous_finetune_lib.run_continuous_finetune(FLAGS.mode, params, model_dir,
+                                                  FLAGS.pretrain_steps)
+  train_utils.save_gin_config(FLAGS.mode, model_dir)
 
 
 if __name__ == '__main__':

official/recommendation/create_ncf_data.py

@@ -14,17 +14,13 @@
 # ==============================================================================
 """Binary to generate training/evaluation dataset for NCF model."""
 
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-
 import json
 
 # pylint: disable=g-bad-import-order
 # Import libraries
 from absl import app
 from absl import flags
-import tensorflow.compat.v2 as tf
+import tensorflow as tf
 # pylint: enable=g-bad-import-order
 
 from official.recommendation import movielens

official/recommendation/ncf_input_pipeline.py

@@ -14,15 +14,10 @@
 # ==============================================================================
 """NCF model input pipeline."""
 
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-
 import functools
 
 # pylint: disable=g-bad-import-order
-
-import tensorflow.compat.v2 as tf
+import tensorflow as tf
 # pylint: enable=g-bad-import-order
 
 from official.recommendation import constants as rconst

official/recommendation/ncf_keras_main.py

@@ -18,10 +18,6 @@ The NeuMF model assembles both MF and MLP models under the NCF framework. Check
 `neumf_model.py` for more details about the models.
 """
 
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-
 import json
 import os
 
@@ -30,7 +26,7 @@ import os
 from absl import app
 from absl import flags
 from absl import logging
-import tensorflow.compat.v2 as tf
+import tensorflow as tf
 # pylint: enable=g-bad-import-order
 
 from official.common import distribute_utils

official/staging/training/grad_utils.py

@@ -14,14 +14,9 @@
 # ==============================================================================
 """Some gradient util functions to help users writing custom training loop."""
 
-from __future__ import absolute_import
-from __future__ import division
-# from __future__ import google_type_annotations
-from __future__ import print_function
-
 from absl import logging
 
-import tensorflow.compat.v2 as tf
+import tensorflow as tf
 
 
 def _filter_grads(grads_and_vars):

official/utils/flags/README.md

 # Adding Abseil (absl) flags quickstart
+
+**WARNING** This module is deprecated. We no long use it in new models and
+your projects should not depend on it. We will remove this module when
+all models using it are deprecated which may take time.
+
 ## Defining a flag
 absl flag definitions are similar to argparse, although they are defined on a global namespace.
 

official/utils/misc/callstack_sampler.py

-"""A simple Python callstack sampler."""
-
-import contextlib
-import datetime
-import signal
-import traceback
-
-
-class CallstackSampler(object):
-  """A simple signal-based Python callstack sampler.
-  """
-
-  def __init__(self, interval=None):
-    self.stacks = []
-    self.interval = 0.001 if interval is None else interval
-
-  def _sample(self, signum, frame):
-    """Samples the current stack."""
-    del signum
-    stack = traceback.extract_stack(frame)
-    formatted_stack = []
-    formatted_stack.append(datetime.datetime.utcnow())
-    for filename, lineno, function_name, text in stack:
-      formatted_frame = '{}:{}({})({})'.format(filename, lineno, function_name,
-                                               text)
-      formatted_stack.append(formatted_frame)
-    self.stacks.append(formatted_stack)
-    signal.setitimer(signal.ITIMER_VIRTUAL, self.interval, 0)
-
-  @contextlib.contextmanager
-  def profile(self):
-    signal.signal(signal.SIGVTALRM, self._sample)
-    signal.setitimer(signal.ITIMER_VIRTUAL, self.interval, 0)
-    try:
-      yield
-    finally:
-      signal.setitimer(signal.ITIMER_VIRTUAL, 0)
-
-  def save(self, fname):
-    with open(fname, 'w') as f:
-      for s in self.stacks:
-        for l in s:
-          f.write('%s\n' % l)
-        f.write('\n')
-
-
-@contextlib.contextmanager
-def callstack_sampling(filename, interval=None):
-  """Periodically samples the Python callstack.
-
-  Args:
-    filename: the filename
-    interval: the sampling interval, in seconds. Defaults to 0.001.
-
-  Yields:
-   nothing
-  """
-  sampler = CallstackSampler(interval=interval)
-  with sampler.profile():
-    yield
-  sampler.save(filename)
-

official/utils/testing/scripts/ci_sanity.sh

@@ -34,10 +34,10 @@ do_pylint() {
     #   --incremental  Performs check on only the python files changed in the
     #                  last non-merge git commit.
 
-    # Use this list to whitelist pylint errors
-    ERROR_WHITELIST=""
+    # Use this list to ALLOWLIST pylint errors
+    ERROR_ALLOWLIST=""
 
-    echo "ERROR_WHITELIST=\"${ERROR_WHITELIST}\""
+    echo "ERROR_ALLOWLIST=\"${ERROR_ALLOWLIST}\""
 
     PYLINT_BIN="python3 -m pylint"
 
@@ -92,16 +92,16 @@ do_pylint() {
 
     N_ERRORS=0
     while read -r LINE; do
-        IS_WHITELISTED=0
-        for WL_REGEX in ${ERROR_WHITELIST}; do
+        IS_ALLOWLISTED=0
+        for WL_REGEX in ${ERROR_ALLOWLIST}; do
             if echo ${LINE} | grep -q "${WL_REGEX}"; then
-                echo "Found a whitelisted error:"
+                echo "Found a ALLOWLISTed error:"
                 echo "  ${LINE}"
-                IS_WHITELISTED=1
+                IS_ALLOWLISTED=1
             fi
         done
 
-        if [[ ${IS_WHITELISTED} == "0" ]]; then
+        if [[ ${IS_ALLOWLISTED} == "0" ]]; then
             echo "${LINE}" >> ${NONWL_ERRORS_FILE}
             echo "" >> ${NONWL_ERRORS_FILE}
             ((N_ERRORS++))
@@ -116,7 +116,7 @@ do_pylint() {
         cat "${NONWL_ERRORS_FILE}"
         return 1
     else
-        echo "PASS: No non-whitelisted pylint errors were found."
+        echo "PASS: No non-ALLOWLISTed pylint errors were found."
         return 0
     fi
 }

official/vision/beta/configs/experiments/maskrcnn/coco_spinenet143_tpu.yaml

+# Expect to reach: box mAP: 49.3%, mask mAP: 43.4% on COCO
+runtime:
+  distribution_strategy: 'tpu'
+  mixed_precision_dtype: 'bfloat16'
+task:
+  init_checkpoint: null
+  train_data:
+    global_batch_size: 256
+    parser:
+      aug_rand_hflip: true
+      aug_scale_min: 0.1
+      aug_scale_max: 2.0
+  losses:
+    l2_weight_decay: 0.00004
+  model:
+    anchor:
+      anchor_size: 4.0
+      num_scales: 3
+    min_level: 3
+    max_level: 7
+    input_size: [1280, 1280, 3]
+    backbone:
+      spinenet:
+        stochastic_depth_drop_rate: 0.2
+        model_id: '143'
+      type: 'spinenet'
+    decoder:
+      type: 'identity'
+    norm_activation:
+      norm_epsilon: 0.001
+      norm_momentum: 0.99
+      use_sync_bn: true
+    detection_generator:
+      pre_nms_top_k: 1000
+trainer:
+  train_steps: 231000
+  optimizer_config:
+    learning_rate:
+      type: 'stepwise'
+      stepwise:
+        boundaries: [219450, 226380]
+        values: [0.32, 0.032, 0.0032]
+    warmup:
+      type: 'linear'
+      linear:
+        warmup_steps: 2000
+        warmup_learning_rate: 0.0067

official/vision/beta/configs/experiments/maskrcnn/coco_spinenet49_tpu.yaml

+# Expect to reach: box mAP: 43.2%, mask mAP: 38.3% on COCO
+runtime:
+  distribution_strategy: 'tpu'
+  mixed_precision_dtype: 'bfloat16'
+task:
+  init_checkpoint: null
+  train_data:
+    global_batch_size: 256
+    parser:
+      aug_rand_hflip: true
+      aug_scale_min: 0.1
+      aug_scale_max: 2.0
+  losses:
+    l2_weight_decay: 0.00004
+  model:
+    anchor:
+      anchor_size: 3.0
+      num_scales: 3
+    min_level: 3
+    max_level: 7
+    input_size: [640, 640, 3]
+    backbone:
+      spinenet:
+        stochastic_depth_drop_rate: 0.2
+        model_id: '49'
+      type: 'spinenet'
+    decoder:
+      type: 'identity'
+    norm_activation:
+      norm_epsilon: 0.001
+      norm_momentum: 0.99
+      use_sync_bn: true
+    detection_generator:
+      pre_nms_top_k: 1000
+trainer:
+  train_steps: 231000
+  optimizer_config:
+    learning_rate:
+      type: 'stepwise'
+      stepwise:
+        boundaries: [219450, 226380]
+        values: [0.32, 0.032, 0.0032]
+    warmup:
+      type: 'linear'
+      linear:
+        warmup_steps: 2000
+        warmup_learning_rate: 0.0067

official/vision/beta/configs/experiments/maskrcnn/coco_spinenet96_tpu.yaml

+# Expect to reach: box mAP: 48.1%, mask mAP: 42.4% on COCO
+runtime:
+  distribution_strategy: 'tpu'
+  mixed_precision_dtype: 'bfloat16'
+task:
+  init_checkpoint: null
+  train_data:
+    global_batch_size: 256
+    parser:
+      aug_rand_hflip: true
+      aug_scale_min: 0.1
+      aug_scale_max: 2.0
+  losses:
+    l2_weight_decay: 0.00004
+  model:
+    anchor:
+      anchor_size: 3.0
+      num_scales: 3
+    min_level: 3
+    max_level: 7
+    input_size: [1024, 1024, 3]
+    backbone:
+      spinenet:
+        stochastic_depth_drop_rate: 0.2
+        model_id: '96'
+      type: 'spinenet'
+    decoder:
+      type: 'identity'
+    norm_activation:
+      norm_epsilon: 0.001
+      norm_momentum: 0.99
+      use_sync_bn: true
+    detection_generator:
+      pre_nms_top_k: 1000
+trainer:
+  train_steps: 231000
+  optimizer_config:
+    learning_rate:
+      type: 'stepwise'
+      stepwise:
+        boundaries: [219450, 226380]
+        values: [0.32, 0.032, 0.0032]
+    warmup:
+      type: 'linear'
+      linear:
+        warmup_steps: 2000
+        warmup_learning_rate: 0.0067

official/vision/beta/configs/experiments/video_classification/k400_3d-resnet50_tpu.yaml

-# 3D ResNet-50 video classification on Kinetics-400. 75.3% top-1 and 91.2% top-5 accuracy.
+# 3D ResNet-50 video classification on Kinetics-400.
 #
 # --experiment_type=video_classification_kinetics400
-# Expected accuracy on TPU 8x8: 75.1%
-# Updated: 2020-12-16
+# Expected accuracy: 77.0% top-1, 93.0% top-5.
 runtime:
   distribution_strategy: 'tpu'
   mixed_precision_dtype: 'bfloat16'
@@ -15,45 +14,63 @@ task:
       resnet_3d:
         block_specs: !!python/tuple
         - temporal_kernel_sizes: !!python/tuple
-          - 3
-          - 3
-          - 3
+          - 1
+          - 1
+          - 1
           temporal_strides: 1
-          use_self_gating: true
+          use_self_gating: false
         - temporal_kernel_sizes: !!python/tuple
-          - 3
           - 1
-          - 3
+          - 1
+          - 1
           - 1
           temporal_strides: 1
-          use_self_gating: true
+          use_self_gating: false
         - temporal_kernel_sizes: !!python/tuple
           - 3
-          - 1
           - 3
-          - 1
           - 3
-          - 1
+          - 3
+          - 3
+          - 3
           temporal_strides: 1
-          use_self_gating: true
+          use_self_gating: false
         - temporal_kernel_sizes: !!python/tuple
-          - 1
           - 3
-          - 1
+          - 3
+          - 3
           temporal_strides: 1
-          use_self_gating: true
+          use_self_gating: false
         model_id: 50
         stem_conv_temporal_kernel_size: 5
         stem_conv_temporal_stride: 2
-        stem_pool_temporal_stride: 2
+        stem_pool_temporal_stride: 1
   train_data:
     name: kinetics400
+    feature_shape: !!python/tuple
+    - 32
+    - 224
+    - 224
+    - 3
+    temporal_stride: 2
     global_batch_size: 1024
     dtype: 'bfloat16'
     shuffle_buffer_size: 1024
+    aug_max_area_ratio: 1.0
+    aug_max_aspect_ratio: 2.0
+    aug_min_area_ratio: 0.08
+    aug_min_aspect_ratio: 0.5
   validation_data:
     name: kinetics400
-    global_batch_size: 32
+    feature_shape: !!python/tuple
+    - 32
+    - 256
+    - 256
+    - 3
+    temporal_stride: 2
+    num_test_clips: 10
+    num_test_crops: 3
+    global_batch_size: 64
     dtype: 'bfloat16'
     drop_remainder: false
 trainer:
@@ -61,11 +78,11 @@ trainer:
     learning_rate:
       cosine:
         initial_learning_rate: 0.8
-        decay_steps: 42000
+        decay_steps: 42104
     warmup:
       linear:
-        warmup_steps: 1050
-  train_steps: 42000
+        warmup_steps: 1053
+  train_steps: 42104
   steps_per_loop: 500
   summary_interval: 500
   validation_interval: 500

official/vision/beta/configs/experiments/video_classification/k400_slowonly16x4_tpu.yaml

+# SlowOnly 16x4 video classification on Kinetics-400.
+#
+# --experiment_type=video_classification_kinetics400
+# Expected accuracy: 75.6% top-1, 92.1% top-5.
+runtime:
+  distribution_strategy: 'tpu'
+  mixed_precision_dtype: 'bfloat16'
+task:
+  model:
+    dropout_rate: 0.5
+    norm_activation:
+      use_sync_bn: false
+    backbone:
+      resnet_3d:
+        block_specs: !!python/tuple
+        - temporal_kernel_sizes: !!python/tuple
+          - 1
+          - 1
+          - 1
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 1
+          - 1
+          - 1
+          - 1
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 3
+          - 3
+          - 3
+          - 3
+          - 3
+          - 3
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 3
+          - 3
+          - 3
+          temporal_strides: 1
+          use_self_gating: false
+        model_id: 50
+        stem_conv_temporal_kernel_size: 1
+        stem_conv_temporal_stride: 1
+        stem_pool_temporal_stride: 1
+  train_data:
+    name: kinetics400
+    feature_shape: !!python/tuple
+    - 16
+    - 224
+    - 224
+    - 3
+    temporal_stride: 4
+    global_batch_size: 1024
+    dtype: 'bfloat16'
+    shuffle_buffer_size: 1024
+    aug_max_area_ratio: 1.0
+    aug_max_aspect_ratio: 2.0
+    aug_min_area_ratio: 0.08
+    aug_min_aspect_ratio: 0.5
+  validation_data:
+    name: kinetics400
+    feature_shape: !!python/tuple
+    - 16
+    - 256
+    - 256
+    - 3
+    temporal_stride: 4
+    num_test_clips: 10
+    num_test_crops: 3
+    global_batch_size: 64
+    dtype: 'bfloat16'
+    drop_remainder: false
+trainer:
+  optimizer_config:
+    learning_rate:
+      cosine:
+        initial_learning_rate: 0.8
+        decay_steps: 42104
+    warmup:
+      linear:
+        warmup_steps: 1053
+  train_steps: 42104
+  steps_per_loop: 500
+  summary_interval: 500
+  validation_interval: 500

official/vision/beta/configs/experiments/video_classification/k400_slowonly8x8_tpu.yaml

-# SlowOnly video classification on Kinetics-400. Expected performance to be updated.
+# SlowOnly 8x8 video classification on Kinetics-400.
+#
+# --experiment_type=video_classification_kinetics400
+# Expected accuracy: 74.1% top-1, 91.4% top-5.
 runtime:
   distribution_strategy: 'tpu'
   mixed_precision_dtype: 'bfloat16'
@@ -53,6 +56,10 @@ task:
     global_batch_size: 1024
     dtype: 'bfloat16'
     shuffle_buffer_size: 1024
+    aug_max_area_ratio: 1.0
+    aug_max_aspect_ratio: 2.0
+    aug_min_area_ratio: 0.08
+    aug_min_aspect_ratio: 0.5
   validation_data:
     name: kinetics400
     feature_shape: !!python/tuple
@@ -61,8 +68,9 @@ task:
     - 256
     - 3
     temporal_stride: 8
-    num_test_clips: 1
-    global_batch_size: 32
+    num_test_clips: 10
+    num_test_crops: 3
+    global_batch_size: 64
     dtype: 'bfloat16'
     drop_remainder: false
 trainer:

official/vision/beta/configs/experiments/video_classification/k600_3d-resnet50_tpu.yaml

+# 3D ResNet-50 video classification on Kinetics-600.
+#
+# --experiment_type=video_classification_kinetics600
+# Expected accuracy: 79.5% top-1, 94.8% top-5.
+runtime:
+  distribution_strategy: 'tpu'
+  mixed_precision_dtype: 'bfloat16'
+task:
+  model:
+    dropout_rate: 0.5
+    norm_activation:
+      use_sync_bn: false
+    backbone:
+      resnet_3d:
+        block_specs: !!python/tuple
+        - temporal_kernel_sizes: !!python/tuple
+          - 1
+          - 1
+          - 1
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 1
+          - 1
+          - 1
+          - 1
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 3
+          - 3
+          - 3
+          - 3
+          - 3
+          - 3
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 3
+          - 3
+          - 3
+          temporal_strides: 1
+          use_self_gating: false
+        model_id: 50
+        stem_conv_temporal_kernel_size: 5
+        stem_conv_temporal_stride: 2
+        stem_pool_temporal_stride: 1
+  train_data:
+    name: kinetics600
+    feature_shape: !!python/tuple
+    - 32
+    - 224
+    - 224
+    - 3
+    temporal_stride: 2
+    global_batch_size: 1024
+    dtype: 'bfloat16'
+    shuffle_buffer_size: 1024
+    aug_max_area_ratio: 1.0
+    aug_max_aspect_ratio: 2.0
+    aug_min_area_ratio: 0.08
+    aug_min_aspect_ratio: 0.5
+  validation_data:
+    name: kinetics600
+    feature_shape: !!python/tuple
+    - 32
+    - 256
+    - 256
+    - 3
+    temporal_stride: 2
+    num_test_clips: 10
+    num_test_crops: 3
+    global_batch_size: 64
+    dtype: 'bfloat16'
+    drop_remainder: false
+trainer:
+  optimizer_config:
+    learning_rate:
+      cosine:
+        initial_learning_rate: 0.8
+        decay_steps: 71488
+    warmup:
+      linear:
+        warmup_steps: 1787
+  train_steps: 71488
+  steps_per_loop: 500
+  summary_interval: 500
+  validation_interval: 500

official/vision/beta/configs/experiments/video_classification/k600_slowonly8x8_tpu.yaml

+# SlowOnly 8x8 video classification on Kinetics-600.
+#
+# --experiment_type=video_classification_kinetics600
+# Expected accuracy: 77.3% top-1, 93.6% top-5.
+runtime:
+  distribution_strategy: 'tpu'
+  mixed_precision_dtype: 'bfloat16'
+task:
+  model:
+    dropout_rate: 0.5
+    norm_activation:
+      use_sync_bn: false
+    backbone:
+      resnet_3d:
+        block_specs: !!python/tuple
+        - temporal_kernel_sizes: !!python/tuple
+          - 1
+          - 1
+          - 1
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 1
+          - 1
+          - 1
+          - 1
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 3
+          - 3
+          - 3
+          - 3
+          - 3
+          - 3
+          temporal_strides: 1
+          use_self_gating: false
+        - temporal_kernel_sizes: !!python/tuple
+          - 3
+          - 3
+          - 3
+          temporal_strides: 1
+          use_self_gating: false
+        model_id: 50
+        stem_conv_temporal_kernel_size: 1
+        stem_conv_temporal_stride: 1
+        stem_pool_temporal_stride: 1
+  train_data:
+    name: kinetics600
+    feature_shape: !!python/tuple
+    - 8
+    - 224
+    - 224
+    - 3
+    temporal_stride: 8
+    global_batch_size: 1024
+    dtype: 'bfloat16'
+    shuffle_buffer_size: 1024
+    aug_max_area_ratio: 1.0
+    aug_max_aspect_ratio: 2.0
+    aug_min_area_ratio: 0.08
+    aug_min_aspect_ratio: 0.5
+  validation_data:
+    name: kinetics600
+    feature_shape: !!python/tuple
+    - 8
+    - 256
+    - 256
+    - 3
+    temporal_stride: 8
+    num_test_clips: 10
+    num_test_crops: 3
+    global_batch_size: 64
+    dtype: 'bfloat16'
+    drop_remainder: false
+trainer:
+  optimizer_config:
+    learning_rate:
+      cosine:
+        initial_learning_rate: 0.8
+        decay_steps: 71488
+    warmup:
+      linear:
+        warmup_steps: 1787
+  train_steps: 71488
+  steps_per_loop: 500
+  summary_interval: 500
+  validation_interval: 500

official/vision/beta/configs/image_classification.py

@@ -35,7 +35,7 @@ class DataConfig(cfg.DataConfig):
   shuffle_buffer_size: int = 10000
   cycle_length: int = 10
   aug_policy: Optional[str] = None  # None, 'autoaug', or 'randaug'
-  file_type: str = 'tfrecord'  # tfrecord, or sstable
+  file_type: str = 'tfrecord'
 
 
 @dataclasses.dataclass

official/vision/beta/configs/maskrcnn.py

@@ -31,11 +31,13 @@ from official.vision.beta.configs import decoders
 @dataclasses.dataclass
 class TfExampleDecoder(hyperparams.Config):
   regenerate_source_id: bool = False
+  mask_binarize_threshold: Optional[float] = None
 
 
 @dataclasses.dataclass
 class TfExampleDecoderLabelMap(hyperparams.Config):
   regenerate_source_id: bool = False
+  mask_binarize_threshold: Optional[float] = None
   label_map: str = ''
 
 
@@ -73,6 +75,7 @@ class DataConfig(cfg.DataConfig):
   decoder: DataDecoder = DataDecoder()
   parser: Parser = Parser()
   shuffle_buffer_size: int = 10000
+  file_type: str = 'tfrecord'
 
 
 @dataclasses.dataclass

official/vision/beta/configs/retinanet.py

@@ -68,6 +68,7 @@ class DataConfig(cfg.DataConfig):
   decoder: DataDecoder = DataDecoder()
   parser: Parser = Parser()
   shuffle_buffer_size: int = 10000
+  file_type: str = 'tfrecord'
 
 
 @dataclasses.dataclass

official/vision/beta/configs/semantic_segmentation.py

@@ -51,7 +51,7 @@ class DataConfig(cfg.DataConfig):
   aug_scale_max: float = 1.0
   aug_rand_hflip: bool = True
   drop_remainder: bool = True
-  file_type: str = 'tfrecord'  # tfrecord, or sstable
+  file_type: str = 'tfrecord'
 
 
 @dataclasses.dataclass