From fe588bc508c0076bb007d6ed36c18ac8ecb341ac Mon Sep 17 00:00:00 2001
From: 王梦迪 <73778524+di-osc@users.noreply.github.com>
Date: 星期二, 20 五月 2025 16:10:59 +0800
Subject: [PATCH] Fsmn_vad支持多线程并发调用 (#2519)
---
runtime/python/onnxruntime/funasr_onnx/vad_bin.py | 7 +++----
runtime/python/onnxruntime/funasr_onnx/utils/frontend.py | 8 ++++----
2 files changed, 7 insertions(+), 8 deletions(-)
diff --git a/runtime/python/onnxruntime/funasr_onnx/utils/frontend.py b/runtime/python/onnxruntime/funasr_onnx/utils/frontend.py
index 7b38f8b..f4cc8ff 100644
--- a/runtime/python/onnxruntime/funasr_onnx/utils/frontend.py
+++ b/runtime/python/onnxruntime/funasr_onnx/utils/frontend.py
@@ -52,12 +52,12 @@
def fbank(self, waveform: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
waveform = waveform * (1 << 15)
- self.fbank_fn = knf.OnlineFbank(self.opts)
- self.fbank_fn.accept_waveform(self.opts.frame_opts.samp_freq, waveform.tolist())
- frames = self.fbank_fn.num_frames_ready
+ fbank_fn = knf.OnlineFbank(self.opts)
+ fbank_fn.accept_waveform(self.opts.frame_opts.samp_freq, waveform.tolist())
+ frames = fbank_fn.num_frames_ready
mat = np.empty([frames, self.opts.mel_opts.num_bins])
for i in range(frames):
- mat[i, :] = self.fbank_fn.get_frame(i)
+ mat[i, :] = fbank_fn.get_frame(i)
feat = mat.astype(np.float32)
feat_len = np.array(mat.shape[0]).astype(np.int32)
return feat, feat_len
diff --git a/runtime/python/onnxruntime/funasr_onnx/vad_bin.py b/runtime/python/onnxruntime/funasr_onnx/vad_bin.py
index 3f21004..af4663a 100644
--- a/runtime/python/onnxruntime/funasr_onnx/vad_bin.py
+++ b/runtime/python/onnxruntime/funasr_onnx/vad_bin.py
@@ -69,7 +69,7 @@
model_file, device_id, intra_op_num_threads=intra_op_num_threads
)
self.batch_size = batch_size
- self.vad_scorer = E2EVadModel(config["model_conf"])
+ self.vad_scorer_config = config["model_conf"]
self.max_end_sil = (
max_end_sil if max_end_sil is not None else config["model_conf"]["max_end_silence_time"]
)
@@ -90,10 +90,9 @@
waveform_list = self.load_data(audio_in, self.frontend.opts.frame_opts.samp_freq)
waveform_nums = len(waveform_list)
is_final = kwargs.get("kwargs", False)
-
segments = [[]] * self.batch_size
for beg_idx in range(0, waveform_nums, self.batch_size):
-
+ vad_scorer = E2EVadModel(self.vad_scorer_config)
end_idx = min(waveform_nums, beg_idx + self.batch_size)
waveform = waveform_list[beg_idx:end_idx]
feats, feats_len = self.extract_feat(waveform)
@@ -122,7 +121,7 @@
inputs.extend(in_cache)
scores, out_caches = self.infer(inputs)
in_cache = out_caches
- segments_part = self.vad_scorer(
+ segments_part = vad_scorer(
scores,
waveform_package,
is_final=is_final,
--
Gitblit v1.9.1