Roshan S Sharma

Our pretrained model has been uploaded to the HF Repo [here](https://huggingface.co/espnet/roshansh_asr_base_sp_conformer_swbd)

The LM Config file is [here](https://github.com/espnet/espnet/blob/master/egs2/swbd/asr1/conf/tuning/train_lm_transformer2.yaml).

Thank you for opening this issue. This issue might take some time to resolve due to server issues. I am investigating alternative storage and will respond here when the issue...

The easiest thing might be to just start with skipping the batch, and we can explore changing the batching after I think. Or we can just go through the batch,...

It depends on what features would need to be extracted. Tools like Kaldi or Librosa can be used to generate speech features.

Please request the dataset using the data release form from the How2 data repository : https://github.com/srvk/how2-dataset. Apologies for the long delay in replying

Please wait for a day or two- I am in the process of setting up a PR to fix this and other issues with the recipe. Thanks!

You may refer to our PR https://github.com/espnet/espnet/pull/4805- this when merged should be used to prepare data from the downloaded dataset bz2 file. The modification made to espnet_model.py should fix the...

I am working on it. Will message after it is ready for review. On Sun, Mar 24, 2024 at 6:34 PM Shinji Watanabe ***@***.***> wrote: > We still have some...

@simpleoier - added fixes for (A) segmentation during KMeans PL extraction and (B) Data Cleaning in Stage 6 - welcome your comments. I also removed unnecessary files, but will make...