search

SeanNaren / deepspeech.torch

Speech Recognition using DeepSpeech2 network and the CTC activation function.
MIT License
259 stars 73 forks source link

Issue with configuring training #76

Closed eranfain closed 7 years ago

eranfain commented 7 years ago

Hello, I am using Mac OS X 10.10.2 with CUDA 8.0 and cudnn V5. When I run the following command (as part of the AN4 setup): th MakeLMDB.lua -rootPath prepare_datasets/an4_dataset -lmdbPath prepare_datasets/an4_lmdb -windowSize 0.02 -stride 0.01 -sampleRate 16000

I get the following exception:

/Users/admin/torch/install/bin/luajit: /Users/admin/torch/install/share/lua/5.1/trepl/init.lua:389: /Users/admin/torch/install/share/lua/5.1/lmdb/ffi.lua:175: dlopen(liblmdb.dylib, 5): image not found stack traceback: [C]: in function 'error' /Users/admin/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require' MakeLMDB.lua:7: in main chunk [C]: in function 'dofile' ...dmin/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk [C]: at 0x010b83bd50

Any help would be appreciated.

suhaspillai commented 7 years ago

Check your lmdb installation What happens when you do 'require lmdb' ? Just go step by step get the audio load the audio calculate spectrogram and then create lmdb data just for that sample.

eranfain commented 7 years ago

Ok. The problem is solved. I probably had something wrong with the lmdb installation.

Thank you!

SeanNaren commented 7 years ago

Thanks @suhaspillai for your help and glad your problem is fixed @eranfain