speaker-recognition-pytorch issues

small typo in dvector create file at the end

1

``` np.save('test_sequence',train_sequence) np.save('test_cluster_id',train_cluster_id) ``` should be "train sequence" in the first line - you save the train_sequence right? so should print train sequence too.

erjan

can I apply SpeechEmbedder to TDNN-LSTM?

As I mentioned, I wonder if I apply your SpeechEmbedder to TDNN-LSTM. TDNN-LSTM requires the following things. delay: the delay to be used in the recurrence of LSTMs decay-time: an...

laetokang

运行报错，求助！

1

您好，下载好TIMIT完整数据集后，请问是否应该先python dvector_create.py，将音频文件.WAV转换为标准格式，然后再python data_preprocess.py，特征提取得到音频的.npy文件？我在运行dvector_create.py时，报错utils.py中的from constants import c，得到结果ModuleNotFoundError: No module named 'constants'，然后我conda install constants，又报错ImportError: cannot import name 'c'，请问该如何解决呢？

LittleMaWen

pre trained weights file download issue

4

Hello, I couldnt raise an issue for the ocr , so i would like to ask a question regarding https://github.com/Aurora11111/CHINESE-OCR here. I'm not able to download the pre trained models...

SreenijaK