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 preset_spk_num

Browse Source
This commit is contained in:
shixian.shi 2024-02-05 11:15:06 +08:00
parent 9c5f00794e
commit 94f7bd079c
2 changed files with 8 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
examples/industrial_data_pretraining/seaco_paraformer/demo.py
View File

@ -11,18 +11,22 @@ model = AutoModel(model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-com
vad_model_revision="v2.0.4", vad_model_revision="v2.0.4",
punc_model="damo/punc_ct-transformer_zh-cn-common-vocab272727-pytorch", punc_model="damo/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
punc_model_revision="v2.0.4", punc_model_revision="v2.0.4",
# spk_model="damo/speech_campplus_sv_zh-cn_16k-common", spk_model="damo/speech_campplus_sv_zh-cn_16k-common",
# spk_model_revision="v2.0.2", spk_model_revision="v2.0.2",
) )
# example1 # example1
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav", res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav",
hotword='达摩院 魔搭', hotword='达摩院 魔搭',
# preset_spk_num=2,
# sentence_timestamp=True, # return sentence level information when spk_model is not given # sentence_timestamp=True, # return sentence level information when spk_model is not given
) )
print(res) print(res)
'''
# tensor or numpy as input
# example2 # example2
import torchaudio import torchaudio
import os import os
@ -38,4 +42,4 @@ import soundfile
wav_file = os.path.join(model.model_path, "example/asr_example.wav") wav_file = os.path.join(model.model_path, "example/asr_example.wav")
speech, sample_rate = soundfile.read(wav_file) speech, sample_rate = soundfile.read(wav_file)
res = model.generate(input=[speech], batch_size_s=300, is_final=True) res = model.generate(input=[speech], batch_size_s=300, is_final=True)
'''

5
funasr/auto/auto_model.py
View File

@ -121,9 +121,6 @@ class AutoModel:
if spk_mode not in ["default", "vad_segment", "punc_segment"]: if spk_mode not in ["default", "vad_segment", "punc_segment"]:
logging.error("spk_mode should be one of default, vad_segment and punc_segment.") logging.error("spk_mode should be one of default, vad_segment and punc_segment.")
self.spk_mode = spk_mode self.spk_mode = spk_mode
self.preset_spk_num = kwargs.get("preset_spk_num", None)
if self.preset_spk_num:
logging.warning("Using preset speaker number: {}".format(self.preset_spk_num))
self.kwargs = kwargs self.kwargs = kwargs
self.model = model self.model = model
@ -391,7 +388,7 @@ class AutoModel:
if self.spk_model is not None: if self.spk_model is not None:
all_segments = sorted(all_segments, key=lambda x: x[0]) all_segments = sorted(all_segments, key=lambda x: x[0])
spk_embedding = result['spk_embedding'] spk_embedding = result['spk_embedding']
labels = self.cb_model(spk_embedding.cpu(), oracle_num=self.preset_spk_num) labels = self.cb_model(spk_embedding.cpu(), oracle_num=kwargs['preset_spk_num'])
del result['spk_embedding'] del result['spk_embedding']
sv_output = postprocess(all_segments, None, labels, spk_embedding.cpu()) sv_output = postprocess(all_segments, None, labels, spk_embedding.cpu())
if self.spk_mode == 'vad_segment': # recover sentence_list if self.spk_mode == 'vad_segment': # recover sentence_list