chazo1994
chazo1994
I don't have GPU, so I need to train tacotron 2 model with cpu. How can't i train a model with multiple thread in cpu (or multiple core) ?.
I try to train a acoustic model with magphase vocoder and use mp.analysis_lossless to extract full lossless features of magphase. Here is the script I used. ` file_name_token = os.path.basename(os.path.splitext(wav_file)[0])...
In my speech synthesis system built from Merlin toolkit, it take long time to generate speech from text. Most of the time used by World Vocoder and DNN generation module....
I changed output_feature_normalisation to MINMAX and got following error: `MINMAX.dat Traceback (most recent call last): File "/data/thinhnv20/Projects/merlin-master-magphase/src/run_merlin.py", line 1270, in main_function(cfg) File "/data/thinhnv20/Projects/merlin-master-magphase/src/run_merlin.py", line 725, in main_function feature_std_vector = numpy.array(global_std_vector[:,feature_index:feature_index+cfg.out_dimension_dict[feature_name]],...
I'm training a DNN model with Magphase vocoder and get error "index out of bound" with my question file that have only QS question type. I debugged file "label_normalisation.py" and...
I use python wrapper of phonetisaurus to run the trained model. It work well with single process, but with multiple processing the following problems occurs: `File "/home/thinhnv13/anaconda2/envs/vits/lib/python3.6/multiprocessing/reduction.py", line 60, in...
I have tried to train a lot of model with sampling rate 22050 but, it can not reproduce quality of 16000 hz model. Can you explain why you use 16000...
Could you please provide the pretrain of Discriminator which corresponding to the generator model you provided (LJS and VCTK)
I would like to say thank you for your great work. To archive faster training, could you provide pretrain model.
I test SqueezeWave and Waveglow with same mel input on GPU Gefore RTX 2080. The results show that, SqueezeWave is two times slower than Waveglow (with short sentence.)