dual-path-RNNs-DPRNNs-based-speech-separation

dual-path-RNNs-DPRNNs-based-speech-separation copied to clipboard

tangchunyang

run time

9

您好，我想问下您运行完100个epoch需要花多少时间呢？因为batch_size有点小，在我这里运行时间有点长，是否可以设置的更大一点呢？会对实验结果有影响吗？

lymiou

asking for an example wav

1

hi,thanks for sharing the code,could you please show us an example mixwav ?

shaoyating

Hello, thanks for sharing useful code. I want to implement "DUAL-PATH RNN: EFFICIENT LONG SEQUENCE MODELING FOR TIME-DOMAIN SINGLE-CHANNEL SPEECH SEPARATION" paper. So I decided to use **DPRNN_base** model instead...

fncode246

in solver.py ``` def _run_one_epoch(self, epoch, cross_valid=False): start = time.time() total_loss = 0 data_loader = self.tr_loader if not cross_valid else self.cv_loader print('data_loader.len {}'.format(len(data_loader))) for i, (data) in enumerate(data_loader): padded_mixture_, mixture_lengths_,...

CunNH3

About cuda memory

1

Hello, thanks for your efforts in this experiment. I am reproducing your experiment. But I found there is an error about cuda memory. Could I know GPUs memory you used...

key2miao

multichannel

3

It seems that you add a beamforming class in this model ,does this mean we can use multi channel datasets?

shanhaidexiamo

fanmanqian

hi, l want to know your datsetset such as WSJ0-mix how to get it,thank you

wangtao2668129173

Hello,thanks for your help on dprnn.I try to use the code but I have meet a problem when I use the command:.CUDA_VISIBLE_DEVICES=0 python train.py --train_dir data/tr --valid_dir data/cv --sample_rate 8000...

Johnsonabuse