-
Hello,
I tried to decode CE model with LM, but I got an error like below. (When I decode CE model without LM, it works well.)
#
- MODEL=wsj_paper7
- PART=test
- BEAM_SIZE=20
- LM=lm
- LM_PATH=data…
-
I have come across few tasks (e.g. `egs/tedlium/`) which do not provide a README. This makes it difficult for experimenters to start working on this task. It would be great if we start adding the nece…
-
Hello,
I have an issue when I tried to decode the model for TLE.
export THEANO_FLAGS=mode=FAST_RUN,device=gpu3,floatX=float32; $LVSR/exp/wsj/decode_tle.sh wsj_reward6 test 200
Here is the log:
- M…
-
Dear Dzmitry
I have a problem installing lvcsr.
I followed the instruction installed all the dependencies. But while I am runnng python setup.py install in the attention-lvcsr/ it gives a lot of conf…
-
In the large data recipes in Kaldi many single job scripts in steps/\* and utils/*, which were very fast for normal sized LVCSR tasks, consume several hours.
e.g. on fisher_swbd/aspire the `utils/data…
-
Hello all folks:
I and @dataf3l wanted to base a new product on your LVCSR engine, but needed Spanish support, could you tell us me how much would be needed in hours-man or money to do this?
I cannot …
-
Hi All
Just to let anyone interested to know, that after 8 months of in depth research I managed to convert HTK 3.5.2 models to Julius DNN format and received really good results. Example WER reduc…
-
ubuntu 14.04.
use thi command(.setup.py install) to setup kaldi-pthon, I have set $KALDI_ROOT already
the errors are as follows:
/usr/include/python2.7/numpy/npy_1_7_deprecated_api.h:15:2: warning:…
-
### 本周工作
1.改D25论文3.3节,调整了3.1节和3.2节的结构
2.看了几篇早期的论文。
《improving deep neural networks for LVCSR using rectified linear units and dropout》
《Dropout: A Simple Way to Prevent Neural Networks from overfi…
-
$LVSR/exp/wsj/decode.sh wsj_paper6
in decode.sh, three parameters should be given.....while only 2 parameters appear in command.