jzlianglu / pykaldi2

Yet another speech toolkit based on Kaldi and PyTorch

Repository from Github https://github.comjzlianglu/pykaldi2

jzlianglu/pykaldi2 Issues

no module named 'kaldi.base._kaldi_error'
Closed 4 years ago
Decoding with train_transformer_ce.py
Closed 5 years ago2
minibatches for LFMMI
Updated 6 years ago1
SETraining is extremely slow
Updated 6 years ago4
Problems about HCLG.fst needed by "train_transformer_se.py"
Closed 6 years ago2
How to use CE regularizer in chain-model training?
Updated 6 years ago1
Why is the chain loss computation so slow?
Closed 6 years ago8
Why negative grad_input of LF_MMI loss?
Closed 6 years ago1
Convolve vs. BlockConvolve for RIR Augmentation
Updated 6 years ago1
incompatible pytorch and torchvision version
Closed 6 years ago1
Targeted simulation
Updated 6 years ago
Selective data simulation
Updated 6 years ago
Physical data format
Updated 6 years ago
RIR format
Updated 6 years ago