No description provided.

Hello. When this will be merged, we will be able to get full language together with it's label when detecting language, am I correct?

SpeechBrain org

Hello. When this will be merged, we will be able to get full language together with it's label when detecting language, am I correct?

Hello,

This PR only modify the README.md .

The model has been trained to recognize 45 languages from short speech recordings. It does not handle machine translation. (if I have understood what you meant)

Best.

SpeechBrain org

This PR lgtm.

Many thanks. 🙂

Adel-Moumen changed pull request status to merged

Hello. When this will be merged, we will be able to get full language together with it's label when detecting language, am I correct?

Hello,

This PR only modify the README.md .

The model has been trained to recognize 45 languages from short speech recordings. It does not handle machine translation. (if I have understood what you meant)

Best.

No... I meant that I want to be able to get short lang name, e.g. 'en': 'English', 'es': 'Spanish', etc.

SpeechBrain org

Hello. When this will be merged, we will be able to get full language together with it's label when detecting language, am I correct?

Hello,

This PR only modify the README.md .

The model has been trained to recognize 45 languages from short speech recordings. It does not handle machine translation. (if I have understood what you meant)

Best.

No... I meant that I want to be able to get short lang name, e.g. 'en': 'English', 'es': 'Spanish', etc.

Hello,

No. You will need to parse it yourself. You can create a dict of association {"English":"en", ...} or you can modify the file label_encoder.txt that you will get when loading the model.

Best.

Hello. When this will be merged, we will be able to get full language together with it's label when detecting language, am I correct?

Hello,

This PR only modify the README.md .

The model has been trained to recognize 45 languages from short speech recordings. It does not handle machine translation. (if I have understood what you meant)

Best.

No... I meant that I want to be able to get short lang name, e.g. 'en': 'English', 'es': 'Spanish', etc.

Hello,

No. You will need to parse it yourself. You can create a dict of association {"English":"en", ...} or you can modify the file label_encoder.txt that you will get when loading the model.

Best.

That's what I've done for now :)

Sign up or log in to comment