. Voice Separate Models for speech application
├── README.md
├── code: core model for speech separate
├── __init__.py
├── __pycache__
├── conf.yml
├── conv_tasnet.py
├── data.py
├── dprnn-train-deploy.py
├── inference.py
├── parser_utils.py
├── requirements.txt
├── wham_dataset_no_sf.py
└── wsj0_mix.py
├── conf.yml: config file for model
├── helper: helper function for application
├── __pycache__
└── utils.py
├── local_experiment.ipynb: local experiment
├── train_and_deploy.ipynb: implementation in Sagemaker
├── mix_both.json: config for mix application
├── result: sample results
├── complete_split_s1.wav
└── complete_split_s2.wav
├── split_weight: best model for speech application
└── best_model.pth
├── test_sample: test sample for speech application
└── audio_raw_1-20_clip_0.wav
└── train_sample_v6: sample for train/validation
├── cv
├── tr
└── tt
11 directories, 20 files