Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

KALDI_ASSERT: at train-ctc-parallel:AddMatMat:cuda-matrix.cc:570, failed: m == NumCols() #217

Open
icestoneking opened this issue Sep 22, 2019 · 0 comments

Comments

@icestoneking
Copy link

hi,alls
I already install essen and I am trying to run the aishell, but I got the failed:
train-ctc-parallel --report-step=160 --num-sequence=16 --frame-limit=1200 --learn-rate=0.003 --momentum=0.9 --verbose=1 'ark,s,cs:apply-cmvn --norm-vars=true --utt2spk=ark:data/train_nodev/utt2spk scp:data/train_nodev/cmvn.scp scp:exp/train_char_l5_c512/train.scp ark:- |' 'ark:gunzip -c exp/train_char_l5_c512/labels.tr.gz|' exp/train_char_l5_c512/nnet/nnet.iter0 exp/train_char_l5_c512/nnet/nnet.iter1
WARNING (train-ctc-parallel:SelectGpuId():cuda-device.cc:150) Suggestion: use 'nvidia-smi -c 1' to set compute exclusive mode
LOG (train-ctc-parallel:SelectGpuIdAuto():cuda-device.cc:262) Selecting from 4 GPUs
LOG (train-ctc-parallel:SelectGpuIdAuto():cuda-device.cc:277) cudaSetDevice(0): Tesla K80 free:11375M, used:64M, total:11439M, free/total:0.994373
LOG (train-ctc-parallel:SelectGpuIdAuto():cuda-device.cc:277) cudaSetDevice(1): Tesla K80 free:11375M, used:64M, total:11439M, free/total:0.994373
LOG (train-ctc-parallel:SelectGpuIdAuto():cuda-device.cc:277) cudaSetDevice(2): Tesla K80 free:11375M, used:64M, total:11439M, free/total:0.994373
LOG (train-ctc-parallel:SelectGpuIdAuto():cuda-device.cc:277) cudaSetDevice(3): Tesla K80 free:11375M, used:64M, total:11439M, free/total:0.994373
LOG (train-ctc-parallel:SelectGpuIdAuto():cuda-device.cc:310) Selected device: 0 (automatically)
LOG (train-ctc-parallel:FinalizeActiveGpu():cuda-device.cc:194) The active GPU is [0]: Tesla K80 free:11358M, used:81M, total:11439M, free/total:0.992887 version 3.7
LOG (train-ctc-parallel:PrintMemoryUsage():cuda-device.cc:334) Memory used: 0 bytes.
LOG (train-ctc-parallel:DisableCaching():cuda-device.cc:731) Disabling caching of GPU memory.
apply-cmvn --norm-vars=true --utt2spk=ark:data/train_nodev/utt2spk scp:data/train_nodev/cmvn.scp scp:exp/train_char_l5_c512/train.scp ark:-
LOG (train-ctc-parallel:main():train-ctc-parallel.cc:121) TRAINING STARTED
KALDI_ASSERT: at train-ctc-parallel:AddMatMat:cuda-matrix.cc:570, failed: m == NumCols()
Stack trace is:
eesen::KaldiGetStackTrace()
eesen::KaldiAssertFailure_(char const*, char const*, int, char const*)
eesen::CuMatrixBase::AddMatMat(float, eesen::CuMatrixBase const&, eesen::MatrixTransposeType, eesen::CuMatrixBase const&, eesen::MatrixTransposeType, float)
eesen::AffineTransform::BackpropagateFnc(eesen::CuMatrixBase const&, eesen::CuMatrixBase const&, eesen::CuMatrixBase const&, eesen::CuMatrixBase)
eesen::Layer::Backpropagate(eesen::CuMatrixBase const&, eesen::CuMatrixBase const&, eesen::CuMatrixBase const&, eesen::CuMatrix
)
eesen::Net::Backpropagate(eesen::CuMatrixBase const&, eesen::CuMatrix*)
train-ctc-parallel(main+0x10e7) [0x495678]
/usr/lib64/libc.so.6(__libc_start_main+0xf5) [0x7fde10ae53d5]
train-ctc-parallel() [0x4930a9]

How to fix it?
Thank you!
Looking forward to your reply~

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant