Huang Kuan Po

Where is the code for choosing hard class-m ?

11

Hi, I have read your paper. In section 4.3 Hard task (HT) meta-batch, it says that "We choose the failure class-m from each task by ranking the class-level accuracies ..."...

Extremely slow when evaluating

8

Hi, I am running stage 3 of [egs/librimix/ConvTasNet/run.sh](https://github.com/asteroid-team/asteroid/blob/master/egs/librimix/ConvTasNet/run.sh#L119). I used ```--compute_wer 1 --eval_mode max``` to evaluate WER. However, it is running extremely slow. ``` 2%|█▉ | 58/3000 [46:02

artists images link unavailable

1

this link: https://hcicloud.iwr.uni-heidelberg.de/index.php/s/NcJj2oLBTYuT1tf is currently unavailable Is there any alternative way to obtain the artists' images ? thanks

Hi, I am trying to reproduce the results of `base_noise_pt_noise_ft_30h.pt` by fine-tuning the pre-trained checkpoint `base_noise_pt_lrs3_vox_iter5.pt`. By directly decoding `base_noise_pt_noise_ft_30h.pt` on the clean set, I get WER 4%. But by...