From 7ab14d064cf019d4e1b0b219b68604c9a697da81 Mon Sep 17 00:00:00 2001
From: 志浩 <neo.dzh@alibaba-inc.com>
Date: 星期一, 27 二月 2023 16:20:14 +0800
Subject: [PATCH] unit test modelscope for sd
---
funasr/bin/sond_inference.py | 14 ++++++++++++--
1 files changed, 12 insertions(+), 2 deletions(-)
diff --git a/funasr/bin/sond_inference.py b/funasr/bin/sond_inference.py
index 299de0d..4add1b5 100755
--- a/funasr/bin/sond_inference.py
+++ b/funasr/bin/sond_inference.py
@@ -33,6 +33,8 @@
from funasr.utils.types import str_or_none
from scipy.ndimage import median_filter
from funasr.utils.misc import statistic_model_parameters
+from funasr.datasets.iterable_dataset import load_bytes
+
class Speech2Diarization:
"""Speech2Xvector class
@@ -257,6 +259,9 @@
assert "sv_model_file" in param_dict, "sv_model_file must be provided in param_dict."
sv_train_config = param_dict["sv_train_config"]
sv_model_file = param_dict["sv_model_file"]
+ if "model_dir" in param_dict:
+ sv_train_config = os.path.join(param_dict["model_dir"], sv_train_config)
+ sv_model_file = os.path.join(param_dict["model_dir"], sv_model_file)
from funasr.bin.sv_inference import Speech2Xvector
speech2xvector_kwargs = dict(
sv_train_config=sv_train_config,
@@ -307,20 +312,25 @@
def _forward(
data_path_and_name_and_type: Sequence[Tuple[str, str, str]] = None,
- raw_inputs: List[List[Union[np.ndarray, torch.Tensor, str]]] = None,
+ raw_inputs: List[List[Union[np.ndarray, torch.Tensor, str, bytes]]] = None,
output_dir_v2: Optional[str] = None,
param_dict: Optional[dict] = None,
):
logging.info("param_dict: {}".format(param_dict))
if data_path_and_name_and_type is None and raw_inputs is not None:
if isinstance(raw_inputs, (list, tuple)):
+ if not isinstance(raw_inputs[0], List):
+ raw_inputs = [raw_inputs]
+
assert all([len(example) >= 2 for example in raw_inputs]), \
"The length of test case in raw_inputs must larger than 1 (>=2)."
def prepare_dataset():
for idx, example in enumerate(raw_inputs):
# read waveform file
- example = [soundfile.read(x)[0] if isinstance(example[0], str) else x
+ example = [load_bytes(x) if isinstance(x, bytes) else x
+ for x in example]
+ example = [soundfile.read(x)[0] if isinstance(x, str) else x
for x in example]
# convert torch tensor to numpy array
example = [x.numpy() if isinstance(example[0], torch.Tensor) else x
--
Gitblit v1.9.1