Internal change

PiperOrigin-RevId: 381320453

Author: Xhark

official/benchmark/base_benchmark.py

@@ -32,12 +32,16 @@
 from official.modeling import hyperparams
 
 
-def _get_benchmark_params(benchmark_models):
+def _get_benchmark_params(benchmark_models, eval_tflite=False):
   """Formats benchmark params into a list."""
   parameterized_benchmark_params = []
   for _, benchmarks in benchmark_models.items():
     for name, params in benchmarks.items():
-      for execution_mode in ['performance', 'accuracy']:
+      if eval_tflite:
+        execution_modes = ['performance', 'tflite_accuracy']
+      else:
+        execution_modes = ['performance', 'accuracy']
+      for execution_mode in execution_modes:
         benchmark_name = '{}.{}'.format(name, execution_mode)
         benchmark_params = (
             benchmark_name,  # First arg is used by ParameterizedBenchmark.
@@ -66,7 +70,8 @@ class BaseBenchmark(  # pylint: disable=undefined-variable
 
   _benchmark_parameters = _get_benchmark_params(
       benchmark_definitions.VISION_BENCHMARKS) + _get_benchmark_params(
-          benchmark_definitions.NLP_BENCHMARKS)
+          benchmark_definitions.NLP_BENCHMARKS) + _get_benchmark_params(
+              benchmark_definitions.QAT_BENCHMARKS, True)
 
   def __init__(self,
                output_dir=None,
@@ -144,7 +149,7 @@ def benchmark(self,
         execution_mode, params, self._get_model_dir(benchmark_name))
 
     metrics = []
-    if execution_mode == 'accuracy':
+    if execution_mode in ['accuracy', 'tflite_accuracy']:
       for metric_bound in metric_bounds:
         metric = {
             'name': metric_bound['name'],

official/benchmark/benchmark_definitions.py

@@ -51,3 +51,6 @@
 
 NLP_BENCHMARKS = {
 }
+
+QAT_BENCHMARKS = {
+}

official/benchmark/benchmark_lib.py

@@ -21,6 +21,7 @@
 from absl import logging
 import orbit
 import tensorflow as tf
+from official.benchmark import tflite_utils
 from official.common import distribute_utils
 from official.core import config_definitions
 from official.core import task_factory
@@ -37,15 +38,18 @@ def run_benchmark(
   """Runs benchmark for a specific experiment.
 
   Args:
-    execution_mode: A 'str', specifying the mode. Can be 'accuracy', or
-      'performance'.
+    execution_mode: A 'str', specifying the mode. Can be 'accuracy',
+      'performance', or 'tflite_accuracy'.
     params: ExperimentConfig instance.
     model_dir: A 'str', a path to store model checkpoints and summaries.
     distribution_strategy: A tf.distribute.Strategy to use. If specified,
      it will be used instead of inferring the strategy from params.
 
   Returns:
     benchmark_data: returns benchmark data in dict format.
+
+  Raises:
+    NotImplementedError: If try to use unsupported setup.
   """
 
   # For GPU runs, allow option to set thread mode
@@ -77,7 +81,7 @@ def run_benchmark(
     trainer.initialize()
 
   steps_per_loop = params.trainer.steps_per_loop if (
-      execution_mode == 'accuracy') else 100
+      execution_mode in ['accuracy', 'tflite_accuracy']) else 100
   controller = orbit.Controller(
       strategy=strategy,
       trainer=trainer,
@@ -105,6 +109,10 @@ def run_benchmark(
       benchmark_data = {'metrics': eval_logs}
     elif execution_mode == 'performance':
       benchmark_data = {}
+    elif execution_mode == 'tflite_accuracy':
+      eval_logs = tflite_utils.train_and_evaluate(
+          params, task, trainer, controller)
+      benchmark_data = {'metrics': eval_logs}
     else:
       raise NotImplementedError(
           'The benchmark execution mode is not implemented: %s' %

official/benchmark/tflite_utils.py

@@ -0,0 +1,29 @@
+# Copyright 2020 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+"""TFLite utils."""
+import orbit
+from official.core import base_task
+from official.core import base_trainer
+from official.core import config_definitions
+
+
+def train_and_evaluate(
+    params: config_definitions.ExperimentConfig,
+    task: base_task.Task,
+    trainer: base_trainer.Trainer,
+    controller: orbit.Controller):
+  """Train and evaluate on TFLite."""
+  raise NotImplementedError('train_and_evaluate on tflite_utils is not '
+                            'implemented yet.')