Merge pull request #2111 from yt605155624/rm_more_log

[CLI]replace logger.info with logger.debug in cli, change default log leve…

docs/requirements.txt

@@ -22,6 +22,7 @@ onnxruntime
 pandas
 paddlenlp
 paddlespeech_feat
+Pillow>=9.0.0
 praatio==5.0.0
 pypinyin
 pypinyin-dict

paddlespeech/cli/asr/infer.py

@@ -133,11 +133,11 @@ def _init_from_path(self,
         """
         Init model and other resources from a specific path.
         """
-        logger.info("start to init the model")
+        logger.debug("start to init the model")
         # default max_len: unit:second
         self.max_len = 50
         if hasattr(self, 'model'):
-            logger.info('Model had been initialized.')
+            logger.debug('Model had been initialized.')
             return
 
         if cfg_path is None or ckpt_path is None:
@@ -151,15 +151,15 @@ def _init_from_path(self,
             self.ckpt_path = os.path.join(
                 self.res_path,
                 self.task_resource.res_dict['ckpt_path'] + ".pdparams")
-            logger.info(self.res_path)
+            logger.debug(self.res_path)
 
         else:
             self.cfg_path = os.path.abspath(cfg_path)
             self.ckpt_path = os.path.abspath(ckpt_path + ".pdparams")
             self.res_path = os.path.dirname(
                 os.path.dirname(os.path.abspath(self.cfg_path)))
-        logger.info(self.cfg_path)
-        logger.info(self.ckpt_path)
+        logger.debug(self.cfg_path)
+        logger.debug(self.ckpt_path)
 
         #Init body.
         self.config = CfgNode(new_allowed=True)
@@ -216,7 +216,7 @@ def _init_from_path(self,
                 max_len = self.config.encoder_conf.max_len
 
             self.max_len = frame_shift_ms * max_len * subsample_rate
-            logger.info(
+            logger.debug(
                 f"The asr server limit max duration len: {self.max_len}")
 
     def preprocess(self, model_type: str, input: Union[str, os.PathLike]):
@@ -227,15 +227,15 @@ def preprocess(self, model_type: str, input: Union[str, os.PathLike]):
 
         audio_file = input
         if isinstance(audio_file, (str, os.PathLike)):
-            logger.info("Preprocess audio_file:" + audio_file)
+            logger.debug("Preprocess audio_file:" + audio_file)
 
         # Get the object for feature extraction
         if "deepspeech2" in model_type or "conformer" in model_type or "transformer" in model_type:
-            logger.info("get the preprocess conf")
+            logger.debug("get the preprocess conf")
             preprocess_conf = self.config.preprocess_config
             preprocess_args = {"train": False}
             preprocessing = Transformation(preprocess_conf)
-            logger.info("read the audio file")
+            logger.debug("read the audio file")
             audio, audio_sample_rate = soundfile.read(
                 audio_file, dtype="int16", always_2d=True)
             if self.change_format:
@@ -255,7 +255,7 @@ def preprocess(self, model_type: str, input: Union[str, os.PathLike]):
             else:
                 audio = audio[:, 0]
 
-            logger.info(f"audio shape: {audio.shape}")
+            logger.debug(f"audio shape: {audio.shape}")
             # fbank
             audio = preprocessing(audio, **preprocess_args)
 
@@ -264,19 +264,19 @@ def preprocess(self, model_type: str, input: Union[str, os.PathLike]):
 
             self._inputs["audio"] = audio
             self._inputs["audio_len"] = audio_len
-            logger.info(f"audio feat shape: {audio.shape}")
+            logger.debug(f"audio feat shape: {audio.shape}")
 
         else:
             raise Exception("wrong type")
 
-        logger.info("audio feat process success")
+        logger.debug("audio feat process success")
 
     @paddle.no_grad()
     def infer(self, model_type: str):
         """
         Model inference and result stored in self.output.
         """
-        logger.info("start to infer the model to get the output")
+        logger.debug("start to infer the model to get the output")
         cfg = self.config.decode
         audio = self._inputs["audio"]
         audio_len = self._inputs["audio_len"]
@@ -293,7 +293,7 @@ def infer(self, model_type: str):
             self._outputs["result"] = result_transcripts[0]
 
         elif "conformer" in model_type or "transformer" in model_type:
-            logger.info(
+            logger.debug(
                 f"we will use the transformer like model : {model_type}")
             try:
                 result_transcripts = self.model.decode(
@@ -352,7 +352,7 @@ def _check(self, audio_file: str, sample_rate: int, force_yes: bool):
                 logger.error("Please input the right audio file path")
                 return False
 
-        logger.info("checking the audio file format......")
+        logger.debug("checking the audio file format......")
         try:
             audio, audio_sample_rate = soundfile.read(
                 audio_file, dtype="int16", always_2d=True)
@@ -374,7 +374,7 @@ def _check(self, audio_file: str, sample_rate: int, force_yes: bool):
                  sox input_audio.xx --rate 8k --bits 16 --channels 1 output_audio.wav \n \
                  ")
             return False
-        logger.info("The sample rate is %d" % audio_sample_rate)
+        logger.debug("The sample rate is %d" % audio_sample_rate)
         if audio_sample_rate != self.sample_rate:
             logger.warning("The sample rate of the input file is not {}.\n \
                             The program will resample the wav file to {}.\n \
@@ -383,28 +383,28 @@ def _check(self, audio_file: str, sample_rate: int, force_yes: bool):
                         ".format(self.sample_rate, self.sample_rate))
             if force_yes is False:
                 while (True):
-                    logger.info(
+                    logger.debug(
                         "Whether to change the sample rate and the channel. Y: change the sample. N: exit the prgream."
                     )
                     content = input("Input(Y/N):")
                     if content.strip() == "Y" or content.strip(
                     ) == "y" or content.strip() == "yes" or content.strip(
                     ) == "Yes":
-                        logger.info(
+                        logger.debug(
                             "change the sampele rate, channel to 16k and 1 channel"
                         )
                         break
                     elif content.strip() == "N" or content.strip(
                     ) == "n" or content.strip() == "no" or content.strip(
                     ) == "No":
-                        logger.info("Exit the program")
+                        logger.debug("Exit the program")
                         return False
                     else:
                         logger.warning("Not regular input, please input again")
 
             self.change_format = True
         else:
-            logger.info("The audio file format is right")
+            logger.debug("The audio file format is right")
             self.change_format = False
 
         return True

paddlespeech/cli/cls/infer.py

@@ -92,7 +92,7 @@ def _init_from_path(self,
             Init model and other resources from a specific path.
         """
         if hasattr(self, 'model'):
-            logger.info('Model had been initialized.')
+            logger.debug('Model had been initialized.')
             return
 
         if label_file is None or ckpt_path is None:
@@ -135,14 +135,14 @@ def preprocess(self, audio_file: Union[str, os.PathLike]):
             Input content can be a text(tts), a file(asr, cls) or a streaming(not supported yet).
         """
         feat_conf = self._conf['feature']
-        logger.info(feat_conf)
+        logger.debug(feat_conf)
         waveform, _ = load(
             file=audio_file,
             sr=feat_conf['sample_rate'],
             mono=True,
             dtype='float32')
         if isinstance(audio_file, (str, os.PathLike)):
-            logger.info("Preprocessing audio_file:" + audio_file)
+            logger.debug("Preprocessing audio_file:" + audio_file)
 
         # Feature extraction
         feature_extractor = LogMelSpectrogram(

paddlespeech/cli/download.py

@@ -61,7 +61,7 @@ def _get_unique_endpoints(trainer_endpoints):
             continue
         ips.add(ip)
         unique_endpoints.add(endpoint)
-    logger.info("unique_endpoints {}".format(unique_endpoints))
+    logger.debug("unique_endpoints {}".format(unique_endpoints))
     return unique_endpoints
 
 
@@ -96,7 +96,7 @@ def get_path_from_url(url,
     # data, and the same ip will only download data once.
     unique_endpoints = _get_unique_endpoints(ParallelEnv().trainer_endpoints[:])
     if osp.exists(fullpath) and check_exist and _md5check(fullpath, md5sum):
-        logger.info("Found {}".format(fullpath))
+        logger.debug("Found {}".format(fullpath))
     else:
         if ParallelEnv().current_endpoint in unique_endpoints:
             fullpath = _download(url, root_dir, md5sum, method=method)
@@ -118,7 +118,7 @@ def _get_download(url, fullname):
     try:
         req = requests.get(url, stream=True)
     except Exception as e:  # requests.exceptions.ConnectionError
-        logger.info("Downloading {} from {} failed with exception {}".format(
+        logger.debug("Downloading {} from {} failed with exception {}".format(
             fname, url, str(e)))
         return False
 
@@ -190,7 +190,7 @@ def _download(url, path, md5sum=None, method='get'):
     fullname = osp.join(path, fname)
     retry_cnt = 0
 
-    logger.info("Downloading {} from {}".format(fname, url))
+    logger.debug("Downloading {} from {}".format(fname, url))
     while not (osp.exists(fullname) and _md5check(fullname, md5sum)):
         if retry_cnt < DOWNLOAD_RETRY_LIMIT:
             retry_cnt += 1
@@ -209,16 +209,16 @@ def _md5check(fullname, md5sum=None):
     if md5sum is None:
         return True
 
-    logger.info("File {} md5 checking...".format(fullname))
+    logger.debug("File {} md5 checking...".format(fullname))
     md5 = hashlib.md5()
     with open(fullname, 'rb') as f:
         for chunk in iter(lambda: f.read(4096), b""):
             md5.update(chunk)
     calc_md5sum = md5.hexdigest()
 
     if calc_md5sum != md5sum:
-        logger.info("File {} md5 check failed, {}(calc) != "
-                    "{}(base)".format(fullname, calc_md5sum, md5sum))
+        logger.debug("File {} md5 check failed, {}(calc) != "
+                     "{}(base)".format(fullname, calc_md5sum, md5sum))
         return False
     return True
 
@@ -227,7 +227,7 @@ def _decompress(fname):
     """
     Decompress for zip and tar file
     """
-    logger.info("Decompressing {}...".format(fname))
+    logger.debug("Decompressing {}...".format(fname))
 
     # For protecting decompressing interupted,
     # decompress to fpath_tmp directory firstly, if decompress

paddlespeech/cli/kws/infer.py

@@ -88,7 +88,7 @@ def _init_from_path(self,
             Init model and other resources from a specific path.
         """
         if hasattr(self, 'model'):
-            logger.info('Model had been initialized.')
+            logger.debug('Model had been initialized.')
             return
 
         if ckpt_path is None:
@@ -141,7 +141,7 @@ def preprocess(self, audio_file: Union[str, os.PathLike]):
         assert os.path.isfile(audio_file)
         waveform, _ = load(audio_file)
         if isinstance(audio_file, (str, os.PathLike)):
-            logger.info("Preprocessing audio_file:" + audio_file)
+            logger.debug("Preprocessing audio_file:" + audio_file)
 
         # Feature extraction
         waveform = paddle.to_tensor(waveform).unsqueeze(0)

paddlespeech/cli/log.py

@@ -49,7 +49,7 @@ def __init__(self, name: str=None):
         self.handler.setFormatter(self.format)
 
         self.logger.addHandler(self.handler)
-        self.logger.setLevel(logging.DEBUG)
+        self.logger.setLevel(logging.INFO)
         self.logger.propagate = False
 
     def __call__(self, log_level: str, msg: str):

paddlespeech/cli/st/infer.py

@@ -110,7 +110,7 @@ def _set_kaldi_bins(self) -> os.PathLike:
         """
         decompressed_path = download_and_decompress(self.kaldi_bins, MODEL_HOME)
         decompressed_path = os.path.abspath(decompressed_path)
-        logger.info("Kaldi_bins stored in: {}".format(decompressed_path))
+        logger.debug("Kaldi_bins stored in: {}".format(decompressed_path))
         if "LD_LIBRARY_PATH" in os.environ:
             os.environ["LD_LIBRARY_PATH"] += f":{decompressed_path}"
         else:
@@ -128,7 +128,7 @@ def _init_from_path(self,
             Init model and other resources from a specific path.
         """
         if hasattr(self, 'model'):
-            logger.info('Model had been initialized.')
+            logger.debug('Model had been initialized.')
             return
 
         if cfg_path is None or ckpt_path is None:
@@ -140,8 +140,8 @@ def _init_from_path(self,
             self.ckpt_path = os.path.join(
                 self.task_resource.res_dir,
                 self.task_resource.res_dict['ckpt_path'])
-            logger.info(self.cfg_path)
-            logger.info(self.ckpt_path)
+            logger.debug(self.cfg_path)
+            logger.debug(self.ckpt_path)
             res_path = self.task_resource.res_dir
         else:
             self.cfg_path = os.path.abspath(cfg_path)
@@ -192,7 +192,7 @@ def preprocess(self, wav_file: Union[str, os.PathLike], model_type: str):
             Input content can be a file(wav).
         """
         audio_file = os.path.abspath(wav_file)
-        logger.info("Preprocess audio_file:" + audio_file)
+        logger.debug("Preprocess audio_file:" + audio_file)
 
         if "fat_st" in model_type:
             cmvn = self.config.cmvn_path

paddlespeech/cli/text/infer.py

@@ -98,7 +98,7 @@ def _init_from_path(self,
             Init model and other resources from a specific path.
         """
         if hasattr(self, 'model'):
-            logger.info('Model had been initialized.')
+            logger.debug('Model had been initialized.')
             return
 
         self.task = task

paddlespeech/cli/tts/infer.py

@@ -173,7 +173,7 @@ def _init_from_path(
         Init model and other resources from a specific path.
         """
         if hasattr(self, 'am_inference') and hasattr(self, 'voc_inference'):
-            logger.info('Models had been initialized.')
+            logger.debug('Models had been initialized.')
             return
 
         # am
@@ -200,9 +200,9 @@ def _init_from_path(
             # must have phones_dict in acoustic
             self.phones_dict = os.path.join(
                 self.am_res_path, self.task_resource.res_dict['phones_dict'])
-            logger.info(self.am_res_path)
-            logger.info(self.am_config)
-            logger.info(self.am_ckpt)
+            logger.debug(self.am_res_path)
+            logger.debug(self.am_config)
+            logger.debug(self.am_ckpt)
         else:
             self.am_config = os.path.abspath(am_config)
             self.am_ckpt = os.path.abspath(am_ckpt)
@@ -248,9 +248,9 @@ def _init_from_path(
             self.voc_stat = os.path.join(
                 self.voc_res_path,
                 self.task_resource.voc_res_dict['speech_stats'])
-            logger.info(self.voc_res_path)
-            logger.info(self.voc_config)
-            logger.info(self.voc_ckpt)
+            logger.debug(self.voc_res_path)
+            logger.debug(self.voc_config)
+            logger.debug(self.voc_ckpt)
         else:
             self.voc_config = os.path.abspath(voc_config)
             self.voc_ckpt = os.path.abspath(voc_ckpt)