-
Input: {"sentence": "what is i miss you"}
Trace:
```
Traceback (most recent call last):
File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"__main__", mod_spec)
File …
-
After I trained the lstm model using speech_demo in mxnet, I want to decode for eval data.
But it occurs an error. The following is the one of the log file for decoding.
```
-----------------------…
-
Hello. We have obtained more than 90% (token) accuracy on our validation set. If I understood correctly, tokens are basically phonemes, plus disambiguation symbols and blanks (please, correct me if I'…
-
here the log file：
# /home/m/CNTK_V2.0_Fianl/CNTK/build/release/bin/cntk configFile=cntk_config/CNTK2_write.cntk DeviceNumber=-1 modelName=/home/m/CNTK_V2.0_Fianl/CNTK/Examples/Speech/BYLW/timit/cn…
JJMEI updated
7 years ago
-
Hi,
I'm trying to use CNTK with the Kaldi plug-in. I have trained a DNN of 6 hidden layers with CNTK (without any error using wsj5k data) and I'm trying now to use "decode_cntk2.sh" to decode the test…
-
I saw that 30% of the utterances in fisher data were discarded as they were less than the required length. We have to use steps/cleanup/combine_short_segments.py stage in the chain recipe. For exampl…
-
Enable the user to rescore word lattices in addition to n-best lists.
-
Hi
I'm trying to make a prediction from an lm learned.
I have used the PennTreebank example to train an lstm language model.
Now i try to write the output of the PosteriorProb layer in order to have t…