Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
PaddlePaddle
DeepSpeech
提交
2c9dc0c8
D
DeepSpeech
项目概览
PaddlePaddle
/
DeepSpeech
12 个月 前同步成功
通知
205
Star
8425
Fork
1598
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
245
列表
看板
标记
里程碑
合并请求
3
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
D
DeepSpeech
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
245
Issue
245
列表
看板
标记
里程碑
合并请求
3
合并请求
3
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
提交
2c9dc0c8
编写于
3月 26, 2022
作者:
X
xiongxinlei
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
add some vector cli comments, test=doc
上级
ef1bc5e8
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
8 addition
and
4 deletion
+8
-4
paddlespeech/cli/vector/infer.py
paddlespeech/cli/vector/infer.py
+8
-4
未找到文件。
paddlespeech/cli/vector/infer.py
浏览文件 @
2c9dc0c8
...
...
@@ -68,12 +68,13 @@ class VectorExecutor(BaseExecutor):
self
.
parser
=
argparse
.
ArgumentParser
(
prog
=
"paddlespeech.vector"
,
add_help
=
True
)
self
.
parser
.
add_argument
(
"--model"
,
type
=
str
,
default
=
"ecapatdnn_voxceleb12"
,
choices
=
[
"ecapatdnn_voxceleb12"
],
help
=
"Choose model type of
as
r task."
)
help
=
"Choose model type of
vecto
r task."
)
self
.
parser
.
add_argument
(
"--task"
,
type
=
str
,
...
...
@@ -81,7 +82,7 @@ class VectorExecutor(BaseExecutor):
choices
=
[
"spk"
],
help
=
"task type in vector domain"
)
self
.
parser
.
add_argument
(
"--input"
,
type
=
str
,
default
=
None
,
help
=
"Audio file to
recognize
."
)
"--input"
,
type
=
str
,
default
=
None
,
help
=
"Audio file to
extract embedding
."
)
self
.
parser
.
add_argument
(
"--sample_rate"
,
type
=
int
,
...
...
@@ -186,7 +187,7 @@ class VectorExecutor(BaseExecutor):
sample_rate (int, optional): model sample rate. Defaults to 16000.
config (os.PathLike, optional): yaml config. Defaults to None.
ckpt_path (os.PathLike, optional): pretrained model path. Defaults to None.
device (
_type_,
optional): paddle running host device. Defaults to paddle.get_device().
device (optional): paddle running host device. Defaults to paddle.get_device().
Returns:
dict: return the audio embedding and the embedding shape
...
...
@@ -216,6 +217,7 @@ class VectorExecutor(BaseExecutor):
def
_get_pretrained_path
(
self
,
tag
:
str
)
->
os
.
PathLike
:
"""get the neural network path from the pretrained model list
we stored all the pretained mode in the variable `pretrained_models`
Args:
tag (str): model tag in the pretrained model list
...
...
@@ -332,6 +334,7 @@ class VectorExecutor(BaseExecutor):
logger
.
info
(
f
"embedding size:
{
embedding
.
shape
}
"
)
# stage 2: put the embedding and dim info to _outputs property
# the embedding type is numpy.array
self
.
_outputs
[
"embedding"
]
=
embedding
def
postprocess
(
self
)
->
Union
[
str
,
os
.
PathLike
]:
...
...
@@ -356,6 +359,7 @@ class VectorExecutor(BaseExecutor):
logger
.
info
(
f
"Preprocess audio file:
{
audio_file
}
"
)
# stage 1: load the audio sample points
# Note: this process must match the training process
waveform
,
sr
=
load_audio
(
audio_file
)
logger
.
info
(
f
"load the audio sample points, shape is:
{
waveform
.
shape
}
"
)
...
...
@@ -397,7 +401,7 @@ class VectorExecutor(BaseExecutor):
sample_rate (int): the desired model sample rate
Returns:
bool: return if the audio sample rate matches the
model sample rate
bool: return if the audio sample rate matches the model sample rate
"""
self
.
sample_rate
=
sample_rate
if
self
.
sample_rate
!=
16000
and
self
.
sample_rate
!=
8000
:
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录