$ python3 infer.py
Step 1: download trainset and unzip them into noisy/
and clean/
directories, respectively under traindata/
folder.
$ wget -c http://datashare.is.ed.ac.uk/bitstream/handle/10283/1942/noisy_trainset_wav.zip
$ wget -c http://datashare.is.ed.ac.uk/bitstream/handle/10283/1942/clean_trainset_wav.zip
$ python3 train.py
Step 3: use your optimal trained model to infer the test data (modify the model parameters configured path in config.py
)
$ python3 infer.py