speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

update timestamp infer param name

Browse Source
This commit is contained in:
shixian.shi 2023-03-08 17:02:50 +08:00
parent 31cda51dfa
commit 1582bad4b5
2 changed files with 23 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
funasr/bin/tp_inference.py
View File

@ -100,9 +100,9 @@ def time_stamp_lfr6_advance(us_alphas, us_cif_peak, char_list):
class SpeechText2Timestamp: class SpeechText2Timestamp:
def __init__( def __init__(
self, self,
tp_train_config: Union[Path, str] = None, timestamp_infer_config: Union[Path, str] = None,
tp_model_file: Union[Path, str] = None, timestamp_model_file: Union[Path, str] = None,
tp_cmvn_file: Union[Path, str] = None, timestamp_cmvn_file: Union[Path, str] = None,
device: str = "cpu", device: str = "cpu",
dtype: str = "float32", dtype: str = "float32",
**kwargs, **kwargs,
@ -110,11 +110,11 @@ class SpeechText2Timestamp:
assert check_argument_types() assert check_argument_types()
# 1. Build ASR model # 1. Build ASR model
tp_model, tp_train_args = ASRTask.build_model_from_file( tp_model, tp_train_args = ASRTask.build_model_from_file(
tp_train_config, tp_model_file, device timestamp_infer_config, timestamp_model_file, device
) )
frontend = None frontend = None
if tp_train_args.frontend is not None: if tp_train_args.frontend is not None:
frontend = WavFrontend(cmvn_file=tp_cmvn_file, **tp_train_args.frontend_conf) frontend = WavFrontend(cmvn_file=timestamp_cmvn_file, **tp_train_args.frontend_conf)
logging.info("tp_model: {}".format(tp_model)) logging.info("tp_model: {}".format(tp_model))
logging.info("tp_train_args: {}".format(tp_train_args)) logging.info("tp_train_args: {}".format(tp_train_args))
@ -178,9 +178,9 @@ def inference(
ngpu: int, ngpu: int,
log_level: Union[int, str], log_level: Union[int, str],
data_path_and_name_and_type, data_path_and_name_and_type,
tp_train_config: Optional[str], timestamp_infer_config: Optional[str],
tp_model_file: Optional[str], timestamp_model_file: Optional[str],
tp_cmvn_file: Optional[str] = None, timestamp_cmvn_file: Optional[str] = None,
raw_inputs: Union[np.ndarray, torch.Tensor] = None, raw_inputs: Union[np.ndarray, torch.Tensor] = None,
key_file: Optional[str] = None, key_file: Optional[str] = None,
allow_variable_data_keys: bool = False, allow_variable_data_keys: bool = False,
@ -194,9 +194,9 @@ def inference(
batch_size=batch_size, batch_size=batch_size,
ngpu=ngpu, ngpu=ngpu,
log_level=log_level, log_level=log_level,
tp_train_config=tp_train_config, timestamp_infer_config=timestamp_infer_config,
tp_model_file=tp_model_file, timestamp_model_file=timestamp_model_file,
tp_cmvn_file=tp_cmvn_file, timestamp_cmvn_file=timestamp_cmvn_file,
key_file=key_file, key_file=key_file,
allow_variable_data_keys=allow_variable_data_keys, allow_variable_data_keys=allow_variable_data_keys,
output_dir=output_dir, output_dir=output_dir,
@ -213,9 +213,9 @@ def inference_modelscope(
ngpu: int, ngpu: int,
log_level: Union[int, str], log_level: Union[int, str],
# data_path_and_name_and_type, # data_path_and_name_and_type,
tp_train_config: Optional[str], timestamp_infer_config: Optional[str],
tp_model_file: Optional[str], timestamp_model_file: Optional[str],
tp_cmvn_file: Optional[str] = None, timestamp_cmvn_file: Optional[str] = None,
# raw_inputs: Union[np.ndarray, torch.Tensor] = None, # raw_inputs: Union[np.ndarray, torch.Tensor] = None,
key_file: Optional[str] = None, key_file: Optional[str] = None,
allow_variable_data_keys: bool = False, allow_variable_data_keys: bool = False,
@ -246,9 +246,9 @@ def inference_modelscope(
# 2. Build speech2vadsegment # 2. Build speech2vadsegment
speechtext2timestamp_kwargs = dict( speechtext2timestamp_kwargs = dict(
tp_train_config=tp_train_config, timestamp_infer_config=timestamp_infer_config,
tp_model_file=tp_model_file, timestamp_model_file=timestamp_model_file,
tp_cmvn_file=tp_cmvn_file, timestamp_cmvn_file=timestamp_cmvn_file,
device=device, device=device,
dtype=dtype, dtype=dtype,
) )
@ -365,17 +365,17 @@ def get_parser():
group = parser.add_argument_group("The model configuration related") group = parser.add_argument_group("The model configuration related")
group.add_argument( group.add_argument(
"--tp_train_config", "--timestamp_infer_config",
type=str, type=str,
help="VAD infer configuration", help="VAD infer configuration",
) )
group.add_argument( group.add_argument(
"--tp_model_file", "--timestamp_model_file",
type=str, type=str,
help="VAD model parameter file", help="VAD model parameter file",
) )
group.add_argument( group.add_argument(
"--tp_cmvn_file", "--timestamp_cmvn_file",
type=str, type=str,
help="Global cmvn file", help="Global cmvn file",
) )

6
funasr/bin/tp_inference_launch.py
View File

@ -76,17 +76,17 @@ def get_parser():
group = parser.add_argument_group("The model configuration related") group = parser.add_argument_group("The model configuration related")
group.add_argument( group.add_argument(
"--tp_train_config", "--timestamp_infer_config",
type=str, type=str,
help="VAD infer configuration", help="VAD infer configuration",
) )
group.add_argument( group.add_argument(
"--tp_model_file", "--timestamp_model_file",
type=str, type=str,
help="VAD model parameter file", help="VAD model parameter file",
) )
group.add_argument( group.add_argument(
"--tp_cmvn_file", "--timestamp_cmvn_file",
type=str, type=str,
help="Global CMVN file", help="Global CMVN file",
) )