Merge pull request #251 from TensorSpeech/refactor

Refactor: add support tf2.8 + update imports + update examples + add helpers

examples/conformer/README.md

@@ -10,11 +10,11 @@ Go to [config.yml](./config.yml)
 
 ## Usage
 
-Training, see `python examples/conformer/train_*.py --help`
+Training, see `python examples/conformer/train.py --help`
 
-Testing, see `python examples/conformer/test_*.py --help`
+Testing, see `python examples/conformer/test.py --help`
 
-TFLite Conversion, see `python examples/conformer/tflite_*.py --help`
+TFLite Conversion, see `python examples/conformer/inference/gen_tflite_model.py --help`
 
 ## Conformer Subwords - Results on LibriSpeech
 

examples/conformer/config.yml

@@ -31,7 +31,7 @@ decoder_config:
   beam_width: 0
   norm_score: True
   corpus_files:
-    - /mnt/Data/MLDL/Datasets/ASR/Raw/LibriSpeech/train-clean-100/transcripts.tsv
+    - H:/MLDL/Datasets/ASR/Raw/LibriSpeech/train-clean-100/transcripts.tsv
 
 model_config:
   name: conformer
@@ -75,8 +75,8 @@ learning_config:
           num_masks: 1
           mask_factor: 27
     data_paths:
-      - /mnt/Data/MLDL/Datasets/ASR/Raw/LibriSpeech/train-clean-100/transcripts.tsv
-    tfrecords_dir: /mnt/Data/MLDL/Datasets/ASR/Raw/LibriSpeech/tfrecords_1030
+      - H:/MLDL/Datasets/ASR/Raw/LibriSpeech/train-clean-100/transcripts.tsv
+    tfrecords_dir: null
     shuffle: True
     cache: True
     buffer_size: 100
@@ -85,8 +85,9 @@ learning_config:
 
   eval_dataset_config:
     use_tf: True
-    data_paths: null
-    tfrecords_dir: /mnt/Data/MLDL/Datasets/ASR/Raw/LibriSpeech/tfrecords_1030
+    data_paths:
+      - H:/MLDL/Datasets/ASR/Raw/LibriSpeech/dev-clean/transcripts.tsv
+    tfrecords_dir: null
     shuffle: False
     cache: True
     buffer_size: 100
@@ -95,7 +96,8 @@ learning_config:
 
   test_dataset_config:
     use_tf: True
-    data_paths: null
+    data_paths:
+      - H:/MLDL/Datasets/ASR/Raw/LibriSpeech/test-clean/transcripts.tsv
     tfrecords_dir: null
     shuffle: False
     cache: True
@@ -113,13 +115,13 @@ learning_config:
     batch_size: 2
     num_epochs: 50
     checkpoint:
-      filepath: /mnt/Miscellanea/Models/local/conformer/checkpoints/{epoch:02d}.h5
+      filepath: D:/Models/local/conformer/checkpoints/{epoch:02d}.h5
       save_best_only: False
       save_weights_only: True
       save_freq: epoch
-    states_dir: /mnt/Miscellanea/Models/local/conformer/states
+    states_dir: D:/Models/local/conformer/states
     tensorboard:
-      log_dir: /mnt/Miscellanea/Models/local/conformer/tensorboard
+      log_dir: D:/Models/local/conformer/tensorboard
       histogram_freq: 1
       write_graph: True
       write_images: True

examples/conformer/inference/gen_saved_model.py

@@ -12,8 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import argparse
 import os
+import fire
 
 from tensorflow_asr.utils import env_util
 
@@ -22,71 +22,58 @@
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
 
-tf.keras.backend.clear_session()
-
-parser = argparse.ArgumentParser(prog="Conformer Testing")
-
-parser.add_argument("--config", type=str, default=DEFAULT_YAML, help="The file path of model configuration file")
-
-parser.add_argument("--h5", type=str, default=None, help="Path to saved h5 weights")
-
-parser.add_argument("--sentence_piece", default=False, action="store_true", help="Whether to use `SentencePiece` model")
-
-parser.add_argument("--subwords", default=False, action="store_true", help="Use subwords")
-
-parser.add_argument("--output_dir", type=str, default=None, help="Output directory for saved model")
-
-args = parser.parse_args()
-
-assert args.h5
-assert args.output_dir
 
 from tensorflow_asr.configs.config import Config
-from tensorflow_asr.featurizers.speech_featurizers import TFSpeechFeaturizer
-from tensorflow_asr.featurizers.text_featurizers import CharFeaturizer, SentencePieceFeaturizer, SubwordFeaturizer
+from tensorflow_asr.helpers import featurizer_helpers
 from tensorflow_asr.models.transducer.conformer import Conformer
 
-config = Config(args.config)
-speech_featurizer = TFSpeechFeaturizer(config.speech_config)
-
-if args.sentence_piece:
-    logger.info("Use SentencePiece ...")
-    text_featurizer = SentencePieceFeaturizer(config.decoder_config)
-elif args.subwords:
-    logger.info("Use subwords ...")
-    text_featurizer = SubwordFeaturizer(config.decoder_config)
-else:
-    logger.info("Use characters ...")
-    text_featurizer = CharFeaturizer(config.decoder_config)
-
-tf.random.set_seed(0)
-
-# build model
-conformer = Conformer(**config.model_config, vocabulary_size=text_featurizer.num_classes)
-conformer.make(speech_featurizer.shape)
-conformer.load_weights(args.h5, by_name=True)
-conformer.summary(line_length=100)
-conformer.add_featurizers(speech_featurizer, text_featurizer)
-
-
-class ConformerModule(tf.Module):
-    def __init__(self, model: Conformer, name=None):
-        super().__init__(name=name)
-        self.model = model
-        self.num_rnns = config.model_config["prediction_num_rnns"]
-        self.rnn_units = config.model_config["prediction_rnn_units"]
-        self.rnn_nstates = 2 if config.model_config["prediction_rnn_type"] == "lstm" else 1
-
-    @tf.function(input_signature=[tf.TensorSpec(shape=[None], dtype=tf.float32)])
-    def pred(self, signal):
-        predicted = tf.constant(0, dtype=tf.int32)
-        states = tf.zeros([self.num_rnns, self.rnn_nstates, 1, self.rnn_units], dtype=tf.float32)
-        features = self.model.speech_featurizer.tf_extract(signal)
-        encoded = self.model.encoder_inference(features)
-        hypothesis = self.model._perform_greedy(encoded, tf.shape(encoded)[0], predicted, states, tflite=False)
-        transcript = self.model.text_featurizer.indices2upoints(hypothesis.prediction)
-        return transcript
-
 
-module = ConformerModule(model=conformer)
-tf.saved_model.save(module, export_dir=args.output_dir, signatures=module.pred.get_concrete_function())
+def main(
+    config: str = DEFAULT_YAML,
+    h5: str = None,
+    sentence_piece: bool = False,
+    subwords: bool = False,
+    output_dir: str = None,
+):
+    assert h5 and output_dir
+    config = Config(config)
+    tf.random.set_seed(0)
+    tf.keras.backend.clear_session()
+
+    speech_featurizer, text_featurizer = featurizer_helpers.prepare_featurizers(
+        config=config,
+        subwords=subwords,
+        sentence_piece=sentence_piece,
+    )
+
+    # build model
+    conformer = Conformer(**config.model_config, vocabulary_size=text_featurizer.num_classes)
+    conformer.make(speech_featurizer.shape)
+    conformer.load_weights(h5, by_name=True)
+    conformer.summary(line_length=100)
+    conformer.add_featurizers(speech_featurizer, text_featurizer)
+
+    class ConformerModule(tf.Module):
+        def __init__(self, model: Conformer, name=None):
+            super().__init__(name=name)
+            self.model = model
+            self.num_rnns = config.model_config["prediction_num_rnns"]
+            self.rnn_units = config.model_config["prediction_rnn_units"]
+            self.rnn_nstates = 2 if config.model_config["prediction_rnn_type"] == "lstm" else 1
+
+        @tf.function(input_signature=[tf.TensorSpec(shape=[None], dtype=tf.float32)])
+        def pred(self, signal):
+            predicted = tf.constant(0, dtype=tf.int32)
+            states = tf.zeros([self.num_rnns, self.rnn_nstates, 1, self.rnn_units], dtype=tf.float32)
+            features = self.model.speech_featurizer.tf_extract(signal)
+            encoded = self.model.encoder_inference(features)
+            hypothesis = self.model._perform_greedy(encoded, tf.shape(encoded)[0], predicted, states, tflite=False)
+            transcript = self.model.text_featurizer.indices2upoints(hypothesis.prediction)
+            return transcript
+
+    module = ConformerModule(model=conformer)
+    tf.saved_model.save(module, export_dir=output_dir, signatures=module.pred.get_concrete_function())
+
+
+if __name__ == "__main__":
+    fire.Fire(main)

examples/conformer/inference/gen_tflite_model.py

@@ -13,58 +13,45 @@
 # limitations under the License.
 
 import os
-import argparse
-from tensorflow_asr.utils import env_util, file_util
+import fire
+from tensorflow_asr.utils import env_util
 
 logger = env_util.setup_environment()
 import tensorflow as tf
 
 from tensorflow_asr.configs.config import Config
-from tensorflow_asr.featurizers.speech_featurizers import TFSpeechFeaturizer
-from tensorflow_asr.featurizers.text_featurizers import SubwordFeaturizer, CharFeaturizer
+from tensorflow_asr.helpers import exec_helpers, featurizer_helpers
 from tensorflow_asr.models.transducer.conformer import Conformer
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
 
-tf.keras.backend.clear_session()
-tf.compat.v1.enable_control_flow_v2()
 
-parser = argparse.ArgumentParser(prog="Conformer TFLite")
+def main(
+    config: str = DEFAULT_YAML,
+    h5: str = None,
+    subwords: bool = False,
+    sentence_piece: bool = False,
+    output: str = None,
+):
+    assert h5 and output
+    tf.keras.backend.clear_session()
+    tf.compat.v1.enable_control_flow_v2()
 
-parser.add_argument("--config", type=str, default=DEFAULT_YAML, help="The file path of model configuration file")
+    config = Config(config)
+    speech_featurizer, text_featurizer = featurizer_helpers.prepare_featurizers(
+        config=config,
+        subwords=subwords,
+        sentence_piece=sentence_piece,
+    )
 
-parser.add_argument("--h5", type=str, default=None, help="Path to saved model")
+    conformer = Conformer(**config.model_config, vocabulary_size=text_featurizer.num_classes)
+    conformer.make(speech_featurizer.shape)
+    conformer.load_weights(h5, by_name=True)
+    conformer.summary(line_length=100)
+    conformer.add_featurizers(speech_featurizer, text_featurizer)
 
-parser.add_argument("--subwords", default=False, action="store_true", help="Use subwords")
+    exec_helpers.convert_tflite(model=conformer, output=output)
 
-parser.add_argument("output", type=str, default=None, help="TFLite file path to be exported")
 
-args = parser.parse_args()
-
-assert args.h5 and args.output
-
-config = Config(args.config)
-speech_featurizer = TFSpeechFeaturizer(config.speech_config)
-
-if args.subwords:
-    text_featurizer = SubwordFeaturizer(config.decoder_config)
-else:
-    text_featurizer = CharFeaturizer(config.decoder_config)
-
-# build model
-conformer = Conformer(**config.model_config, vocabulary_size=text_featurizer.num_classes)
-conformer.make(speech_featurizer.shape)
-conformer.load_weights(args.h5, by_name=True)
-conformer.summary(line_length=100)
-conformer.add_featurizers(speech_featurizer, text_featurizer)
-
-concrete_func = conformer.make_tflite_function().get_concrete_function()
-converter = tf.lite.TFLiteConverter.from_concrete_functions([concrete_func])
-converter.experimental_new_converter = True
-converter.optimizations = [tf.lite.Optimize.DEFAULT]
-converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS, tf.lite.OpsSet.SELECT_TF_OPS]
-tflite_model = converter.convert()
-
-args.output = file_util.preprocess_paths(args.output)
-with open(args.output, "wb") as tflite_out:
-    tflite_out.write(tflite_model)
+if __name__ == "__main__":
+    fire.Fire(main)

examples/conformer/inference/run_saved_model.py

@@ -12,8 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import argparse
 import os
+import fire
 
 from tensorflow_asr.utils import env_util
 
@@ -22,21 +22,23 @@
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
 
-tf.keras.backend.clear_session()
 
-parser = argparse.ArgumentParser()
+from tensorflow_asr.featurizers.speech_featurizers import read_raw_audio
 
-parser.add_argument("--saved_model", type=str, default=None, help="The file path of saved model")
 
-parser.add_argument("filename", type=str, default=None, help="Audio file path")
+def main(
+    saved_model: str = None,
+    filename: str = None,
+):
+    tf.keras.backend.clear_session()
 
-args = parser.parse_args()
+    module = tf.saved_model.load(export_dir=saved_model)
 
-from tensorflow_asr.featurizers.speech_featurizers import read_raw_audio
+    signal = read_raw_audio(filename)
+    transcript = module.pred(signal)
 
-module = tf.saved_model.load(export_dir=args.saved_model)
+    print("Transcript: ", "".join([chr(u) for u in transcript]))
 
-signal = read_raw_audio(args.filename)
-transcript = module.pred(signal)
 
-print("Transcript: ", "".join([chr(u) for u in transcript]))
+if __name__ == "__main__":
+    fire.Fire(main)

examples/conformer/inference/run_tflite_model.py

@@ -12,39 +12,36 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import argparse
+import fire
 import tensorflow as tf
 
 from tensorflow_asr.featurizers.speech_featurizers import read_raw_audio
 
-parser = argparse.ArgumentParser()
 
-parser.add_argument("filename", metavar="FILENAME", help="Audio file to be played back")
+def main(
+    filename: str,
+    tflite: str = None,
+    blank: int = 0,
+    num_rnns: int = 1,
+    nstates: int = 2,
+    statesize: int = 320,
+):
+    tflitemodel = tf.lite.Interpreter(model_path=tflite)
 
-parser.add_argument("--tflite", type=str, default=None, help="Path to conformer tflite")
+    signal = read_raw_audio(filename)
 
-parser.add_argument("--blank", type=int, default=0, help="Blank index")
+    input_details = tflitemodel.get_input_details()
+    output_details = tflitemodel.get_output_details()
+    tflitemodel.resize_tensor_input(input_details[0]["index"], signal.shape)
+    tflitemodel.allocate_tensors()
+    tflitemodel.set_tensor(input_details[0]["index"], signal)
+    tflitemodel.set_tensor(input_details[1]["index"], tf.constant(blank, dtype=tf.int32))
+    tflitemodel.set_tensor(input_details[2]["index"], tf.zeros([num_rnns, nstates, 1, statesize], dtype=tf.float32))
+    tflitemodel.invoke()
+    hyp = tflitemodel.get_tensor(output_details[0]["index"])
 
-parser.add_argument("--num_rnns", type=int, default=1, help="Number of RNN layers in prediction network")
+    print("".join([chr(u) for u in hyp]))
 
-parser.add_argument("--nstates", type=int, default=2, help="Number of RNN states in prediction network")
 
-parser.add_argument("--statesize", type=int, default=320, help="Size of RNN state in prediction network")
-
-args = parser.parse_args()
-
-tflitemodel = tf.lite.Interpreter(model_path=args.tflite)
-
-signal = read_raw_audio(args.filename)
-
-input_details = tflitemodel.get_input_details()
-output_details = tflitemodel.get_output_details()
-tflitemodel.resize_tensor_input(input_details[0]["index"], signal.shape)
-tflitemodel.allocate_tensors()
-tflitemodel.set_tensor(input_details[0]["index"], signal)
-tflitemodel.set_tensor(input_details[1]["index"], tf.constant(args.blank, dtype=tf.int32))
-tflitemodel.set_tensor(input_details[2]["index"], tf.zeros([args.num_rnns, args.nstates, 1, args.statesize], dtype=tf.float32))
-tflitemodel.invoke()
-hyp = tflitemodel.get_tensor(output_details[0]["index"])
-
-print("".join([chr(u) for u in hyp]))
+if __name__ == "__main__":
+    fire.Fire(main)