支持24k音频超分48k采样率

5 months ago · ffbf205ae4
parent 00bdc01113
commit ffbf205ae4
1 changed files with 3 additions and 0 deletions
--- a/tools/AP_BWE_main/24kto48k/readme.txt
+++ b/tools/AP_BWE_main/24kto48k/readme.txt
@ -1,3 +1,6 @@
 For the inference of the v3 model, if you find that the generated audio sounds somewhat muffled, you can try using this audio super-resolution model.
 对于v3模型的推理，如果你发现生成的音频比较闷，可以尝试这个音频超分模型。
 put g_24kto48k.zip and config.json in this folder
 把g_24kto48k.zip and config.json下到这个文件夹