M-AILABS Speech German Dataset

M-AILABS Speech German Dataset
Format: .opus 32 bits (Sound Directory for GoldenDict PC -not Android)
Headwords: 117,000 sentences
Size: 2.8 GiB
Source: [size=2]https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset/[/size]

Description: German phrases pronounced by native speakers mainly from Librivox.org . The data is ready to be used on GoldenDict PC (not Android) and the “Search Bar” needs to be visible in order to show all the results.



  • Decompress the .7zip file in one folder and load them as “Sound Directory” on GoldenDict (GD).
  • Open GD Menu > Edit > Dictionaries > Sources > Sound Directories > Press the button ADD
  • Choose the folder where the decompressed audios are located, click the “Recursive” option. Press Apply.
  • The collection of audios should appear as a Dictionary available on GoldenDict ! Enjoy :smiley:

PS. File names were produced on Linux. There might be compatibility issues with Windows and Mac. A method to solve the issue is to compile .dsl or .mdx files from the audios. Please write in the forum if there is any problem.

It would be desirable to convert the data into .mdx format, but not really necessary. Any conversion would be welcome :smiley:

1 Like

Also check:

1 Like