From 14a9e017811b14bdddc1301715e356843293ac30 Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期一, 05 二月 2024 11:18:41 +0800
Subject: [PATCH] Merge branch 'main' of github.com:alibaba-damo-academy/FunASR merge
---
funasr/auto/auto_model.py | 5 +----
examples/industrial_data_pretraining/seaco_paraformer/demo.py | 10 +++++++---
2 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/examples/industrial_data_pretraining/seaco_paraformer/demo.py b/examples/industrial_data_pretraining/seaco_paraformer/demo.py
index 85d989e..804acdd 100644
--- a/examples/industrial_data_pretraining/seaco_paraformer/demo.py
+++ b/examples/industrial_data_pretraining/seaco_paraformer/demo.py
@@ -11,18 +11,22 @@
vad_model_revision="v2.0.4",
punc_model="damo/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
punc_model_revision="v2.0.4",
- # spk_model="damo/speech_campplus_sv_zh-cn_16k-common",
- # spk_model_revision="v2.0.2",
+ spk_model="damo/speech_campplus_sv_zh-cn_16k-common",
+ spk_model_revision="v2.0.2",
)
# example1
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav",
hotword='杈炬懇闄� 榄旀惌',
+ # preset_spk_num=2,
# sentence_timestamp=True, # return sentence level information when spk_model is not given
)
print(res)
+
+'''
+# tensor or numpy as input
# example2
import torchaudio
import os
@@ -38,4 +42,4 @@
wav_file = os.path.join(model.model_path, "example/asr_example.wav")
speech, sample_rate = soundfile.read(wav_file)
res = model.generate(input=[speech], batch_size_s=300, is_final=True)
-
+'''
\ No newline at end of file
diff --git a/funasr/auto/auto_model.py b/funasr/auto/auto_model.py
index ae47d35..3829bca 100644
--- a/funasr/auto/auto_model.py
+++ b/funasr/auto/auto_model.py
@@ -121,9 +121,6 @@
if spk_mode not in ["default", "vad_segment", "punc_segment"]:
logging.error("spk_mode should be one of default, vad_segment and punc_segment.")
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.model = model
@@ -391,7 +388,7 @@
if self.spk_model is not None:
all_segments = sorted(all_segments, key=lambda x: x[0])
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']
sv_output = postprocess(all_segments, None, labels, spk_embedding.cpu())
if self.spk_mode == 'vad_segment': # recover sentence_list
--
Gitblit v1.9.1