From f3cd90dcf21e2d4ca451abbfdc841ac6abfc68ee Mon Sep 17 00:00:00 2001
From: zhifu gao <zhifu.gzf@alibaba-inc.com>
Date: 星期二, 14 二月 2023 14:59:03 +0800
Subject: [PATCH] Merge pull request #105 from yufan-aslp/main
---
funasr/runtime/python/onnxruntime/paraformer/rapid_paraformer/paraformer_onnx.py | 20 ++++++--------------
1 files changed, 6 insertions(+), 14 deletions(-)
diff --git a/funasr/runtime/python/onnxruntime/paraformer/rapid_paraformer/paraformer_onnx.py b/funasr/runtime/python/onnxruntime/paraformer/rapid_paraformer/paraformer_onnx.py
index 1fc3582..e520008 100644
--- a/funasr/runtime/python/onnxruntime/paraformer/rapid_paraformer/paraformer_onnx.py
+++ b/funasr/runtime/python/onnxruntime/paraformer/rapid_paraformer/paraformer_onnx.py
@@ -9,10 +9,11 @@
import librosa
import numpy as np
-from .utils import (CharTokenizer, Hypothesis, ONNXRuntimeError,
- OrtInferSession, TokenIDConverter, WavFrontend, get_logger,
+from utils.utils import (CharTokenizer, Hypothesis, ONNXRuntimeError,
+ OrtInferSession, TokenIDConverter, get_logger,
read_yaml)
-from .postprocess_utils import sentence_postprocess
+from utils.postprocess_utils import sentence_postprocess
+from utils.frontend import WavFrontend
logging = get_logger()
@@ -65,7 +66,7 @@
wav_content: Union[str, np.ndarray, List[str]]) -> List:
def load_wav(path: str) -> np.ndarray:
waveform, _ = librosa.load(path, sr=None)
- return waveform[None, ...]
+ return waveform
if isinstance(wav_content, np.ndarray):
return [wav_content]
@@ -137,15 +138,6 @@
texts = sentence_postprocess(token)
text = texts[0]
# text = self.tokenizer.tokens2text(token)
- return text[:valid_token_num-1]
+ return text
-
-if __name__ == '__main__':
- project_dir = Path(__file__).resolve().parent.parent
- model_dir = "/home/zhifu.gzf/.cache/modelscope/hub/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
- model = Paraformer(model_dir)
-
- wav_file = os.path.join(model_dir, 'example/asr_example.wav')
- result = model(wav_file)
- print(result)
--
Gitblit v1.9.1