✍️ update env util

ebraraktas · May 16, 2021 · 44b6875 · 44b6875
1 parent f79b3f6
commit 44b6875
Show file tree

Hide file tree

Showing 34 changed files with 136 additions and 118 deletions.
diff --git a/examples/conformer/test.py b/examples/conformer/test.py
@@ -17,7 +17,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -63,13 +63,13 @@
 speech_featurizer = TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Use SentencePiece ...")
+    logger.info("Use SentencePiece ...")
     text_featurizer = SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Use subwords ...")
+    logger.info("Use subwords ...")
     text_featurizer = SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = CharFeaturizer(config.decoder_config)
 
 tf.random.set_seed(0)
@@ -96,7 +96,7 @@
         overwrite = input(f"Overwrite existing result file {filepath} ? (y/n): ").lower() == "y"
     if overwrite:
         results = conformer.predict(test_data_loader, verbose=1)
-        print(f"Saving result to {args.output} ...")
+        logger.info(f"Saving result to {args.output} ...")
         with open(filepath, "w") as openfile:
             openfile.write("PATH\tDURATION\tGROUNDTRUTH\tGREEDY\tBEAMSEARCH\n")
             progbar = tqdm(total=test_dataset.total_steps, unit="batch")

diff --git a/examples/conformer/tflite.py b/examples/conformer/tflite.py
@@ -16,7 +16,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 from tensorflow_asr.configs.config import Config

diff --git a/examples/conformer/train.py b/examples/conformer/train.py
@@ -17,7 +17,7 @@
 import argparse
 from tensorflow_asr.utils import env_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -64,13 +64,13 @@
 speech_featurizer = speech_featurizers.TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Loading SentencePiece model ...")
+    logger.info("Loading SentencePiece model ...")
     text_featurizer = text_featurizers.SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Loading subwords ...")
+    logger.info("Loading subwords ...")
     text_featurizer = text_featurizers.SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = text_featurizers.CharFeaturizer(config.decoder_config)
 
 if args.tfrecords:

diff --git a/examples/contextnet/test.py b/examples/contextnet/test.py
@@ -17,7 +17,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -63,13 +63,13 @@
 speech_featurizer = TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Use SentencePiece ...")
+    logger.info("Use SentencePiece ...")
     text_featurizer = SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Use subwords ...")
+    logger.info("Use subwords ...")
     text_featurizer = SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = CharFeaturizer(config.decoder_config)
 
 tf.random.set_seed(0)
@@ -96,7 +96,7 @@
         overwrite = input(f"Overwrite existing result file {filepath} ? (y/n): ").lower() == "y"
     if overwrite:
         results = contextnet.predict(test_data_loader, verbose=1)
-        print(f"Saving result to {args.output} ...")
+        logger.info(f"Saving result to {args.output} ...")
         with open(filepath, "w") as openfile:
             openfile.write("PATH\tDURATION\tGROUNDTRUTH\tGREEDY\tBEAMSEARCH\n")
             progbar = tqdm(total=test_dataset.total_steps, unit="batch")

diff --git a/examples/contextnet/tflite.py b/examples/contextnet/tflite.py
@@ -16,7 +16,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 from tensorflow_asr.configs.config import Config

diff --git a/examples/contextnet/train.py b/examples/contextnet/train.py
@@ -17,7 +17,7 @@
 import argparse
 from tensorflow_asr.utils import env_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -64,13 +64,13 @@
 speech_featurizer = speech_featurizers.TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Loading SentencePiece model ...")
+    logger.info("Loading SentencePiece model ...")
     text_featurizer = text_featurizers.SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Loading subwords ...")
+    logger.info("Loading subwords ...")
     text_featurizer = text_featurizers.SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = text_featurizers.CharFeaturizer(config.decoder_config)
 
 if args.tfrecords:

diff --git a/examples/deepspeech2/test.py b/examples/deepspeech2/test.py
@@ -17,7 +17,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -63,13 +63,13 @@
 speech_featurizer = TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Use SentencePiece ...")
+    logger.info("Use SentencePiece ...")
     text_featurizer = SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Use subwords ...")
+    logger.info("Use subwords ...")
     text_featurizer = SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = CharFeaturizer(config.decoder_config)
 
 tf.random.set_seed(0)
@@ -96,7 +96,7 @@
         overwrite = input(f"Overwrite existing result file {filepath} ? (y/n): ").lower() == "y"
     if overwrite:
         results = deepspeech2.predict(test_data_loader, verbose=1)
-        print(f"Saving result to {args.output} ...")
+        logger.info(f"Saving result to {args.output} ...")
         with open(filepath, "w") as openfile:
             openfile.write("PATH\tDURATION\tGROUNDTRUTH\tGREEDY\tBEAMSEARCH\n")
             progbar = tqdm(total=test_dataset.total_steps, unit="batch")

diff --git a/examples/deepspeech2/tflite.py b/examples/deepspeech2/tflite.py
@@ -16,7 +16,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 from tensorflow_asr.configs.config import Config

diff --git a/examples/deepspeech2/train.py b/examples/deepspeech2/train.py
@@ -16,7 +16,7 @@
 import argparse
 from tensorflow_asr.utils import env_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -62,13 +62,13 @@
 speech_featurizer = speech_featurizers.TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Loading SentencePiece model ...")
+    logger.info("Loading SentencePiece model ...")
     text_featurizer = text_featurizers.SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Loading subwords ...")
+    logger.info("Loading subwords ...")
     text_featurizer = text_featurizers.SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = text_featurizers.CharFeaturizer(config.decoder_config)
 
 if args.tfrecords:

diff --git a/examples/demonstration/conformer.py b/examples/demonstration/conformer.py
@@ -16,7 +16,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, math_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 parser = argparse.ArgumentParser(prog="Conformer non streaming")
@@ -52,10 +52,10 @@
 config = Config(args.config)
 speech_featurizer = TFSpeechFeaturizer(config.speech_config)
 if args.sentence_piece:
-    print("Loading SentencePiece model ...")
+    logger.info("Loading SentencePiece model ...")
     text_featurizer = SentencePieceFeaturizer.load_from_file(config.decoder_config, args.subwords)
 elif args.subwords and os.path.exists(args.subwords):
-    print("Loading subwords ...")
+    logger.info("Loading subwords ...")
     text_featurizer = SubwordFeaturizer.load_from_file(config.decoder_config, args.subwords)
 else:
     text_featurizer = CharFeaturizer(config.decoder_config)
@@ -74,14 +74,14 @@
 
 if args.beam_width:
     transcript = conformer.recognize_beam(features[None, ...], input_length[None, ...])
-    print("Transcript:", transcript[0].numpy().decode("UTF-8"))
+    logger.info("Transcript:", transcript[0].numpy().decode("UTF-8"))
 elif args.timestamp:
     transcript, stime, etime, _, _ = conformer.recognize_tflite_with_timestamp(
         signal, tf.constant(text_featurizer.blank, dtype=tf.int32), conformer.predict_net.get_initial_state())
-    print("Transcript:", transcript)
-    print("Start time:", stime)
-    print("End time:", etime)
+    logger.info("Transcript:", transcript)
+    logger.info("Start time:", stime)
+    logger.info("End time:", etime)
 else:
     transcript, _, _ = conformer.recognize_tflite(
         signal, tf.constant(text_featurizer.blank, dtype=tf.int32), conformer.predict_net.get_initial_state())
-    print("Transcript:", tf.strings.unicode_encode(transcript, "UTF-8").numpy().decode("UTF-8"))
+    logger.info("Transcript:", tf.strings.unicode_encode(transcript, "UTF-8").numpy().decode("UTF-8"))
diff --git a/examples/demonstration/rnn_transducer.py b/examples/demonstration/rnn_transducer.py
@@ -15,7 +15,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, math_util, data_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 parser = argparse.ArgumentParser(prog="Rnn Transducer non streaming")
@@ -51,10 +51,10 @@
 config = Config(args.config)
 speech_featurizer = TFSpeechFeaturizer(config.speech_config)
 if args.sentence_piece:
-    print("Loading SentencePiece model ...")
+    logger.info("Loading SentencePiece model ...")
     text_featurizer = SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Loading subwords ...")
+    logger.info("Loading subwords ...")
     text_featurizer = SubwordFeaturizer(config.decoder_config)
 else:
     text_featurizer = CharFeaturizer(config.decoder_config)
@@ -78,22 +78,22 @@
             inputs_length=input_length[None, ...]
         )
     )
-    print("Transcript:", transcript[0].numpy().decode("UTF-8"))
+    logger.info("Transcript:", transcript[0].numpy().decode("UTF-8"))
 elif args.timestamp:
     transcript, stime, etime, _, _, _ = rnnt.recognize_tflite_with_timestamp(
         signal=signal,
         predicted=tf.constant(text_featurizer.blank, dtype=tf.int32),
         encoder_states=rnnt.encoder.get_initial_state(),
         prediction_states=rnnt.predict_net.get_initial_state()
     )
-    print("Transcript:", transcript)
-    print("Start time:", stime)
-    print("End time:", etime)
+    logger.info("Transcript:", transcript)
+    logger.info("Start time:", stime)
+    logger.info("End time:", etime)
 else:
     transcript = rnnt.recognize(
         data_util.create_inputs(
             inputs=features[None, ...],
             inputs_length=input_length[None, ...]
         )
     )
-    print("Transcript:", transcript[0].numpy().decode("UTF-8"))
+    logger.info("Transcript:", transcript[0].numpy().decode("UTF-8"))
diff --git a/examples/demonstration/streaming_tflite_conformer.py b/examples/demonstration/streaming_tflite_conformer.py
@@ -39,7 +39,7 @@ def int_or_str(text):
 args, remaining = parser.parse_known_args()
 
 if args.list_devices:
-    print(sd.query_devices())
+    logger.info(sd.query_devices())
     parser.exit(0)
 
 parser.add_argument('filename', metavar='FILENAME',
@@ -113,7 +113,7 @@ def recognize(signal, lastid, states):
             data = Q.get()
             text, lastid, states = recognize(data, lastid, states)
             transcript += text
-            print(transcript, flush=True)
+            logger.info(transcript, flush=True)
         except queue.Empty:
             pass
 
@@ -126,14 +126,14 @@ def send(q, Q, E):
     def callback(outdata, frames, time, status):
         assert frames == args.blocksize
         if status.output_underflow:
-            print('Output underflow: increase blocksize?', file=sys.stderr)
+            logger.info('Output underflow: increase blocksize?', file=sys.stderr)
             raise sd.CallbackAbort
         assert not status
         try:
             data = q.get_nowait()
             Q.put(np.frombuffer(data, dtype=np.float32))
         except queue.Empty as e:
-            print('Buffer is empty: increase buffersize?', file=sys.stderr)
+            logger.info('Buffer is empty: increase buffersize?', file=sys.stderr)
             raise sd.CallbackAbort from e
         if len(data) < len(outdata):
             outdata[:len(data)] = data

diff --git a/examples/demonstration/tflite_conformer.py b/examples/demonstration/tflite_conformer.py
@@ -59,4 +59,4 @@
 tflitemodel.invoke()
 hyp = tflitemodel.get_tensor(output_details[0]["index"])
 
-print("".join([chr(u) for u in hyp]))
+logger.info("".join([chr(u) for u in hyp]))
diff --git a/examples/jasper/test.py b/examples/jasper/test.py
@@ -17,7 +17,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 DEFAULT_YAML = os.path.join(os.path.abspath(os.path.dirname(__file__)), "config.yml")
@@ -63,13 +63,13 @@
 speech_featurizer = TFSpeechFeaturizer(config.speech_config)
 
 if args.sentence_piece:
-    print("Use SentencePiece ...")
+    logger.info("Use SentencePiece ...")
     text_featurizer = SentencePieceFeaturizer(config.decoder_config)
 elif args.subwords:
-    print("Use subwords ...")
+    logger.info("Use subwords ...")
     text_featurizer = SubwordFeaturizer(config.decoder_config)
 else:
-    print("Use characters ...")
+    logger.info("Use characters ...")
     text_featurizer = CharFeaturizer(config.decoder_config)
 
 tf.random.set_seed(0)
@@ -96,7 +96,7 @@
         overwrite = input(f"Overwrite existing result file {filepath} ? (y/n): ").lower() == "y"
     if overwrite:
         results = jasper.predict(test_data_loader, verbose=1)
-        print(f"Saving result to {args.output} ...")
+        logger.info(f"Saving result to {args.output} ...")
         with open(filepath, "w") as openfile:
             openfile.write("PATH\tDURATION\tGROUNDTRUTH\tGREEDY\tBEAMSEARCH\n")
             progbar = tqdm(total=test_dataset.total_steps, unit="batch")

diff --git a/examples/jasper/tflite.py b/examples/jasper/tflite.py
@@ -16,7 +16,7 @@
 import argparse
 from tensorflow_asr.utils import env_util, file_util
 
-env_util.setup_environment()
+logger = env_util.setup_environment()
 import tensorflow as tf
 
 from tensorflow_asr.configs.config import Config