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
633d5126
D
DeepSpeech
项目概览
PaddlePaddle
/
DeepSpeech
大约 1 年 前同步成功
通知
206
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
633d5126
编写于
6月 07, 2022
作者:
Y
Yang Zhou
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
add opt in script,test=doc
上级
2296e015
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
18 addition
and
2 deletion
+18
-2
speechx/examples/ds2_ol/aishell/run.sh
speechx/examples/ds2_ol/aishell/run.sh
+8
-0
speechx/examples/ds2_ol/aishell/run_fbank.sh
speechx/examples/ds2_ol/aishell/run_fbank.sh
+10
-2
未找到文件。
speechx/examples/ds2_ol/aishell/run.sh
浏览文件 @
633d5126
...
...
@@ -89,6 +89,7 @@ if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
--model_path
=
$model_dir
/avg_1.jit.pdmodel
\
--param_path
=
$model_dir
/avg_1.jit.pdiparams
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--nnet_decoder_chunk
=
8
\
--dict_file
=
$vocb_dir
/vocab.txt
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/result
...
...
@@ -96,6 +97,7 @@ if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
>
$exp
/
${
wer
}
echo
"ctc-prefix-beam-search-decoder-ol without lm has finished!!!"
echo
"please checkout in
${
exp
}
/
${
wer
}
"
tail
-n
7
$exp
/
${
wer
}
fi
if
[
${
stage
}
-le
3
]
&&
[
${
stop_stage
}
-ge
3
]
;
then
...
...
@@ -106,6 +108,7 @@ if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
--model_path
=
$model_dir
/avg_1.jit.pdmodel
\
--param_path
=
$model_dir
/avg_1.jit.pdiparams
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--nnet_decoder_chunk
=
8
\
--dict_file
=
$vocb_dir
/vocab.txt
\
--lm_path
=
$lm
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/result_lm
...
...
@@ -114,6 +117,7 @@ if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
_lm
>
$exp
/
${
wer
}
.lm
echo
"ctc-prefix-beam-search-decoder-ol with lm test has finished!!!"
echo
"please checkout in
${
exp
}
/
${
wer
}
.lm"
tail
-n
7
$exp
/
${
wer
}
.lm
fi
wfst
=
$data
/wfst/
...
...
@@ -138,6 +142,7 @@ if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
--word_symbol_table
=
$wfst
/words.txt
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--graph_path
=
$wfst
/TLG.fst
--max_active
=
7500
\
--nnet_decoder_chunk
=
8
\
--acoustic_scale
=
1.2
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/result_tlg
...
...
@@ -145,6 +150,7 @@ if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
_tlg
>
$exp
/
${
wer
}
.tlg
echo
"wfst-decoder-ol have finished!!!"
echo
"please checkout in
${
exp
}
/
${
wer
}
.tlg"
tail
-n
7
$exp
/
${
wer
}
.tlg
fi
if
[
${
stage
}
-le
5
]
&&
[
${
stop_stage
}
-ge
5
]
;
then
...
...
@@ -156,6 +162,7 @@ if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
--model_path
=
$model_dir
/avg_1.jit.pdmodel
\
--param_path
=
$model_dir
/avg_1.jit.pdiparams
\
--word_symbol_table
=
$wfst
/words.txt
\
--nnet_decoder_chunk
=
8
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--graph_path
=
$wfst
/TLG.fst
--max_active
=
7500
\
--acoustic_scale
=
1.2
\
...
...
@@ -165,4 +172,5 @@ if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
_recognizer
>
$exp
/
${
wer
}
.recognizer
echo
"recognizer test have finished!!!"
echo
"please checkout in
${
exp
}
/
${
wer
}
.recognizer"
tail
-n
7
$exp
/
${
wer
}
.recognizer
fi
speechx/examples/ds2_ol/aishell/run_fbank.sh
浏览文件 @
633d5126
...
...
@@ -90,11 +90,13 @@ if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
--param_path
=
$model_dir
/avg_5.jit.pdiparams
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--model_cache_shapes
=
"5-1-2048,5-1-2048"
\
--nnet_decoder_chunk
=
8
\
--dict_file
=
$vocb_dir
/vocab.txt
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/result_fbank
cat
$data
/split
${
nj
}
/
*
/result_fbank
>
$exp
/
${
label_file
}
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
>
$exp
/
${
wer
}
tail
-n
7
$exp
/
${
wer
}
fi
if
[
${
stage
}
-le
3
]
&&
[
${
stop_stage
}
-ge
3
]
;
then
...
...
@@ -105,13 +107,15 @@ if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
--model_path
=
$model_dir
/avg_5.jit.pdmodel
\
--param_path
=
$model_dir
/avg_5.jit.pdiparams
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--model_cache_shapes
=
"5-1-2048,5-1-2048"
\
--model_cache_shapes
=
"5-1-2048,5-1-2048"
\
--nnet_decoder_chunk
=
8
\
--dict_file
=
$vocb_dir
/vocab.txt
\
--lm_path
=
$lm
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/fbank_result_lm
cat
$data
/split
${
nj
}
/
*
/fbank_result_lm
>
$exp
/
${
label_file
}
_lm
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
_lm
>
$exp
/
${
wer
}
.lm
tail
-n
7
$exp
/
${
wer
}
.lm
fi
wfst
=
$data
/wfst_fbank/
...
...
@@ -135,7 +139,8 @@ if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
--param_path
=
$model_dir
/avg_5.jit.pdiparams
\
--word_symbol_table
=
$wfst
/words.txt
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--model_cache_shapes
=
"5-1-2048,5-1-2048"
\
--model_cache_shapes
=
"5-1-2048,5-1-2048"
\
--nnet_decoder_chunk
=
8
\
--graph_path
=
$wfst
/TLG.fst
--max_active
=
7500
\
--acoustic_scale
=
1.2
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/result_tlg
...
...
@@ -144,6 +149,7 @@ if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
_tlg
>
$exp
/
${
wer
}
.tlg
echo
"wfst-decoder-ol have finished!!!"
echo
"please checkout in
${
exp
}
/
${
wer
}
.tlg"
tail
-n
7
$exp
/
${
wer
}
.tlg
fi
if
[
${
stage
}
-le
5
]
&&
[
${
stop_stage
}
-ge
5
]
;
then
...
...
@@ -157,6 +163,7 @@ if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
--word_symbol_table
=
$wfst
/words.txt
\
--model_output_names
=
softmax_0.tmp_0,tmp_5,concat_0.tmp_0,concat_1.tmp_0
\
--model_cache_shapes
=
"5-1-2048,5-1-2048"
\
--nnet_decoder_chunk
=
8
\
--graph_path
=
$wfst
/TLG.fst
--max_active
=
7500
\
--acoustic_scale
=
1.2
\
--result_wspecifier
=
ark,t:
$data
/split
${
nj
}
/JOB/result_fbank_recognizer
...
...
@@ -165,4 +172,5 @@ if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
utils/compute-wer.py
--char
=
1
--v
=
1
$text
$exp
/
${
label_file
}
_recognizer
>
$exp
/
${
wer
}
.recognizer
echo
"recognizer test have finished!!!"
echo
"please checkout in
${
exp
}
/
${
wer
}
.recognizer"
tail
-n
7
$exp
/
${
wer
}
.recognizer
fi
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