Anastasia Kuznetsova

Downstream ASR Custom dataset

1

Hi all, I would like to adapt SUPERB downstream ASR for Common Voice dataset. Are there any guidelines to follow in order to convert the dataset into expected format? Thanks!

Problem with hfst-compose

1

Hello everyone! I am trying to compose two transducers: https://dpaste.de/n6OU https://dpaste.de/V7jB First I convert from `.txt` to `.hfst`, then composin, then converting the composed transducer back to `.txt` ```bash hfst-txt2fst...

Running PESQ file

2

Hi, Jason I am trying to run MetricGAN.py But it fails while executing the `read_pesq` function. ``` def read_pesq(clean_root, enhanced_file, sr): f=enhanced_file.split('/')[-1] wave_name=f.split('_')[-1].split('@')[0] clean_file=clean_root+'Train_'+wave_name+'.wav' cmd = PESQ_path+'/PESQ {} {} +{}'.format(clean_file,...

Model storage

3

Hello Thomas! Using the library I did not figure out how to store the model for subsequent use and how to get embeddings from it. Could you help me out...