Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Specified dimension not compatible with data

See original GitHub issue

Hi, I am implementing a base-model for DNN based TTS system using merlin frontend. Currently I’m using 50 wav files from arctic database and I have 280 binary features. However, I’m unable to proceed with output feature composition with the following error-

2016-09-17 13:32:25,526 INFO      acoustic_comp: processing file    1 of   50 : /home/neo/work/speech_iiitd/dnn_test2/data/nn_mgc_lf0_vuv_bap_199/arctic_a0001.cmp
Traceback (most recent call last):
  File "../dnn/dnn_tts/run_dnn.py", line 970, in <module>
    main_function(cfg)
  File "../dnn/dnn_tts/run_dnn.py", line 635, in main_function
    acoustic_worker.prepare_nn_data(in_file_list_dict, nn_cmp_file_list, cfg.in_dimension_dict, cfg.out_dimension_dict)
  File "/home/neo/work/speech_iiitd/dnn/dnn_tts/frontend/acoustic_base.py", line 122, in prepare_nn_data
    self.prepare_data(in_file_list_dict, out_file_list, in_dimension_dict, out_dimension_dict)
  File "/home/neo/work/speech_iiitd/dnn/dnn_tts/frontend/acoustic_composition.py", line 126, in prepare_data
    features, frame_number = io_funcs.load_binary_file_frame(in_file_name, in_feature_dim)
  File "/home/neo/work/speech_iiitd/dnn/dnn_tts/io_funcs/binary_io.py", line 67, in load_binary_file_frame
    assert features.size % float(dimension) == 0.0,'specified dimension %s not compatible with data'%(dimension)
AssertionError: specified dimension 5 not compatible with data

I’m new to dnn based speech synthesis. Kindly give some suggestions. Also, why do we normalize the features with [0,1]. Thanks.

My conf file-


[Labels]
question_file_name:path/to/data/resources/questions_dnn_unilex-rpx_quinphone.hed 

[Outputs]
mgc    : 60
dmgc   : 180
bap    : 5  
dbap   : 15
lf0    : 1  
dlf0   : 3  

[Data]
train_file_number: 30
valid_file_number: 10
test_file_number : 10
buffer_size: 200000

[Processes]
NORMLAB  : False
MAKECMP  : True
NORMCMP  : True
TRAINDNN : False
DNNGEN   : False
GENWAV   : False
CALMCD   : False

Issue Analytics

State:
Created 7 years ago
Comments:15 (9 by maintainers)

Top GitHub Comments

2reactions

ronankicommented, Sep 29, 2016

If you specify the order 5, while extracting the coefficients with the given scripts, you will end up with extracting 6 (5+1) coefficients. So, either use 6 in configuration file or change the order to 4 while extraction.

2reactions

gillesdegottexcommented, Sep 26, 2016

Did you clear the data folder before ? There are nasty codes that skip the computation of files if the file already exists …