Mirco Ravanelli comments

Results 148 comments of


                                            Mirco Ravanelli

Can't add new language to pre-trained spoken language recognition model: Model forgets other languages

That could play an important role. I think it is important to make sure there are data from different languages in each batch.

Integrating Pruned Fast RNNT with Transducer + new recipe for mTEDx dataset

Thank you @Anwarvic, could you please report there the current results? In particular, it is interesting to see the performance/speed comparison between torchaudio loss and the FastRNNT (at least with...

Integrating Pruned Fast RNNT with Transducer + new recipe for mTEDx dataset

Thank you @danpovey! @Anwarvic, did you try the warmup thing?

bug fix Pretrained.load_audio

@anautsch, any update on this PR?

Fix libresspech lm dataset

What should we do with this PR?

3rd party arpa LM to asr-crdnn-rnnlm-librispeech

I think Aku ***@***.***) can provide some examples. In general I think LMs and integration of ARPA LMs can be the object of a tutorial. What do you think about...

demo code in huggingface for 2 speaker separation is not working

That's weird. I tried a couple of days ago and everything was working. Could you please share the code snippet with us? Cem can take a looks as well. On...

AMI Diarization recipe databse folder structure

Any idea Nauman? On Wed, Aug 10, 2022, 1:27 PM Matt Nappo ***@***.***> wrote: > Did you end up finding a solution to this? I am having the same issue...

Brain's check_gradients seems to not do what name implies?

@pplantinga

Add Text-to-Speech functionality (Tacotron only)

Amazing work @flexthink! This looks really great. In the following, you can find several small comments that mainly go in the direction of making the recipe more compliant with the...