search

mravanelli / SincNet

SincNet is a neural architecture for efficiently processing raw audio samples.
MIT License
1.11k stars 260 forks source link

The error when run speaker_id.py #86

Open BLack-yzf opened 4 years ago

BLack-yzf commented 4 years ago
  1. hello，Mr.mravanelli I have run the speaker_id.py, but i got this error. I have checked the code in line 52. I have found that one of the signal's "snt_len" in the batchsize is “2470” ,therefore the value of "snt_len-wlen-1" is negtive which causes this error. I look forward to your answer and thank you very much!

Traceback (most recent call last): File "speaker_id.py", line 235, in [inp,lab]=create_batches_rnd(batch_size,data_folder,wav_lst_tr,snt_tr,wlen,lab_dict,0.2) File "speaker_id.py", line 52, in create_batches_rnd snt_beg=np.random.randint(snt_len-wlen-1) #randint(0, snt_len-2*wlen-1)## File "mtrand.pyx", line 745, in numpy.random.mtrand.RandomState.randint File "_bounded_integers.pyx", line 1363, in numpy.random._bounded_integers._rand_int32 ValueError: low >= high

BLack-yzf commented 4 years ago

I found more details. Is there any chance the prepared data is not exactly right? Hope you can reply. Thanks. train/dr5/fbjl0/sx202.wav 14940 train/dr5/mmdm1/sx333.wav 56333 train/dr4/mtrc0/sx93.wav 20220 train/dr5/mmab1/sx414.wav 36862 train/dr8/mbcg0/sx327.wav 60940 train/dr1/fsma0/sx361.wav 33339 train/dr5/flmk0/sx315.wav 35970 train/dr4/fjxp0/sx42.wav 25120 train/dr2/msat0/sx446.wav 2470