The implementation of the paper SNR-Based Progressive Learning of Deep Neural Network for Speech Enhancement.
Unofficial implementation, with some differences from the original paper.
- Data pre-processing for training, validation and test
- PL DNN models
- Training data
- Validation data
- Training logic
- validation logic
- visualization for validation set (waveform, audio, metrics)
- Config parameters
- Integrate pre-processing scripts in the project
- Global Normalization
- Test script
- Single-SNR
- Multi-SNR
- Baseline DNN model
- tqdm
- pypesq
- pystoi
- librosa
- pytorch==1.0
- matplotlib
- tensorboardX
😅 waiting for integration into the project.
# specify training configuration (-C), and GPU devices (-D).
python train.py -C config/train/<name>.json -D 0
# Resume expriment (-R)
python train.py -C config/train/<name>.json -D 0 -RConfiguration for Training:
{
# 实验名，每次实验都不能重复
"name": "basic_1000",
# 实验使用的 GPU 数量，需要配合 train.py 的 -D 选项
"n_gpu": 1,
# 是否使用 Cudnn 加快速度，使用时无法保证实验的可重复性
"use_cudnn": true,
# 损失函数，见 models/loss.py
"loss_func": "mse_loss",
# Checkpoints，logs 等目录所在的位置，可以指定在另外的数据盘中
"save_location": "/media/imucs/DataDisk/haoxiang/Experiment/DNN",
# 本次实验的描述信息，后续会打印在 tensorboardX 中
"description": "修复 epoch 显示错误的问题",
# 可视化评价指标的频率
"visualize_metrics_period": 10,
# 优化器参数
"optimizer": {
"lr": 0.01
},
# 训练过程的可配置参数
"trainer": {
"epochs": 1000,
# 模型存储的频率
"save_period": 3
},
# 所有关于训练的数据集
"train_dataset": {
"mixture_dataset": "/home/imucs/Center/timit_single_snr_limit_500/train/dB-5.npy",
"clean_dataset": "/home/imucs/Center/timit_single_snr_limit_500/train/clean.npy",
"target_1_dataset": "/home/imucs/Center/timit_single_snr_limit_500/train/dB5.npy",
"target_2_dataset": "/home/imucs/Center/timit_single_snr_limit_500/train/dB15.npy"
},
# 所有关于验证的数据集
"valid_dataset": {
"mixture_dataset": "/home/imucs/Center/timit_waveform_192/test/dB-5.npy",
"clean_dataset": "/home/imucs/Center/timit_waveform_192/test/clean.npy"
},
# 训练数据集中的参数
"train_data": {
"batch_size": 20000,
"shuffle": true,
"num_workers": 100
},
# 验证数据集中的参数
"valid_data": {
"limit": 20,
"offset": 1,
"batch_size": 1,
"num_workers": 1
}
}
tensorborad --logdir <training_config:save_location>/logs --port <port>Similar to the validation part in trainer/trainer.py.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent