Commit
·
8060d45
1
Parent(s):
df64b4e
Default log level to DEBUG, faster-whisper to match
Browse files- whisper_online.py +6 -1
- whisper_online_server.py +4 -10
whisper_online.py
CHANGED
@@ -106,7 +106,7 @@ class FasterWhisperASR(ASRBase):
|
|
106 |
|
107 |
def load_model(self, modelsize=None, cache_dir=None, model_dir=None):
|
108 |
from faster_whisper import WhisperModel
|
109 |
-
logging.getLogger("faster_whisper").setLevel(
|
110 |
if model_dir is not None:
|
111 |
logger.debug(f"Loading whisper model from model_dir {model_dir}. modelsize and cache_dir parameters are not used.")
|
112 |
model_size_or_path = model_dir
|
@@ -558,6 +558,7 @@ def add_shared_args(parser):
|
|
558 |
parser.add_argument('--vad', action="store_true", default=False, help='Use VAD = voice activity detection, with the default parameters.')
|
559 |
parser.add_argument('--buffer_trimming', type=str, default="segment", choices=["sentence", "segment"],help='Buffer trimming strategy -- trim completed sentences marked with punctuation mark and detected by sentence segmenter, or the completed segments returned by Whisper. Sentence segmenter must be installed for "sentence" option.')
|
560 |
parser.add_argument('--buffer_trimming_sec', type=float, default=15, help='Buffer trimming length threshold in seconds. If buffer length is longer, trimming sentence/segment is triggered.')
|
|
|
561 |
|
562 |
def asr_factory(args, logfile=sys.stderr):
|
563 |
"""
|
@@ -624,6 +625,10 @@ if __name__ == "__main__":
|
|
624 |
logger.error("No or one option from --offline and --comp_unaware are available, not both. Exiting.")
|
625 |
sys.exit(1)
|
626 |
|
|
|
|
|
|
|
|
|
627 |
audio_path = args.audio_path
|
628 |
|
629 |
SAMPLING_RATE = 16000
|
|
|
106 |
|
107 |
def load_model(self, modelsize=None, cache_dir=None, model_dir=None):
|
108 |
from faster_whisper import WhisperModel
|
109 |
+
logging.getLogger("faster_whisper").setLevel(logger.level)
|
110 |
if model_dir is not None:
|
111 |
logger.debug(f"Loading whisper model from model_dir {model_dir}. modelsize and cache_dir parameters are not used.")
|
112 |
model_size_or_path = model_dir
|
|
|
558 |
parser.add_argument('--vad', action="store_true", default=False, help='Use VAD = voice activity detection, with the default parameters.')
|
559 |
parser.add_argument('--buffer_trimming', type=str, default="segment", choices=["sentence", "segment"],help='Buffer trimming strategy -- trim completed sentences marked with punctuation mark and detected by sentence segmenter, or the completed segments returned by Whisper. Sentence segmenter must be installed for "sentence" option.')
|
560 |
parser.add_argument('--buffer_trimming_sec', type=float, default=15, help='Buffer trimming length threshold in seconds. If buffer length is longer, trimming sentence/segment is triggered.')
|
561 |
+
parser.add_argument("-l", "--log-level", dest="log_level", choices=['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'], help="Set the log level", default='DEBUG')
|
562 |
|
563 |
def asr_factory(args, logfile=sys.stderr):
|
564 |
"""
|
|
|
625 |
logger.error("No or one option from --offline and --comp_unaware are available, not both. Exiting.")
|
626 |
sys.exit(1)
|
627 |
|
628 |
+
if args.log_level:
|
629 |
+
logging.basicConfig(format='whisper-%(levelname)s:%(name)s: %(message)s',
|
630 |
+
level=getattr(logging, args.log_level))
|
631 |
+
|
632 |
audio_path = args.audio_path
|
633 |
|
634 |
SAMPLING_RATE = 16000
|
whisper_online_server.py
CHANGED
@@ -8,7 +8,6 @@ import logging
|
|
8 |
import numpy as np
|
9 |
|
10 |
logger = logging.getLogger(__name__)
|
11 |
-
print(__name__)
|
12 |
parser = argparse.ArgumentParser()
|
13 |
|
14 |
# server options
|
@@ -17,18 +16,13 @@ parser.add_argument("--port", type=int, default=43007)
|
|
17 |
parser.add_argument("--warmup-file", type=str, dest="warmup_file",
|
18 |
help="The path to a speech audio wav file to warm up Whisper so that the very first chunk processing is fast. It can be e.g. https://github.com/ggerganov/whisper.cpp/raw/master/samples/jfk.wav .")
|
19 |
|
20 |
-
parser.add_argument("-l", "--log-level", dest="log_level",
|
21 |
-
choices=['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'],
|
22 |
-
help="Set the log level",
|
23 |
-
default='INFO')
|
24 |
-
|
25 |
|
26 |
# options from whisper_online
|
27 |
add_shared_args(parser)
|
28 |
args = parser.parse_args()
|
29 |
|
30 |
if args.log_level:
|
31 |
-
logging.basicConfig(format='whisper-server-%(levelname)s: %(message)s',
|
32 |
level=getattr(logging, args.log_level))
|
33 |
|
34 |
# setting whisper object by args
|
@@ -47,11 +41,11 @@ if args.warmup_file:
|
|
47 |
if os.path.isfile(args.warmup_file):
|
48 |
a = load_audio_chunk(args.warmup_file,0,1)
|
49 |
asr.transcribe(a)
|
50 |
-
|
51 |
else:
|
52 |
-
|
53 |
else:
|
54 |
-
|
55 |
|
56 |
|
57 |
######### Server objects
|
|
|
8 |
import numpy as np
|
9 |
|
10 |
logger = logging.getLogger(__name__)
|
|
|
11 |
parser = argparse.ArgumentParser()
|
12 |
|
13 |
# server options
|
|
|
16 |
parser.add_argument("--warmup-file", type=str, dest="warmup_file",
|
17 |
help="The path to a speech audio wav file to warm up Whisper so that the very first chunk processing is fast. It can be e.g. https://github.com/ggerganov/whisper.cpp/raw/master/samples/jfk.wav .")
|
18 |
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
# options from whisper_online
|
21 |
add_shared_args(parser)
|
22 |
args = parser.parse_args()
|
23 |
|
24 |
if args.log_level:
|
25 |
+
logging.basicConfig(format='whisper-server-%(levelname)s:%(name)s: %(message)s',
|
26 |
level=getattr(logging, args.log_level))
|
27 |
|
28 |
# setting whisper object by args
|
|
|
41 |
if os.path.isfile(args.warmup_file):
|
42 |
a = load_audio_chunk(args.warmup_file,0,1)
|
43 |
asr.transcribe(a)
|
44 |
+
logger.info("Whisper is warmed up.")
|
45 |
else:
|
46 |
+
logger.warning("The warm up file is not available. "+msg)
|
47 |
else:
|
48 |
+
logger.warning(msg)
|
49 |
|
50 |
|
51 |
######### Server objects
|