- This topic has 2 replies, 2 voices, and was last updated 8 years, 6 months ago by .
Viewing 1 reply thread
Viewing 1 reply thread
- You must be logged in to reply to this topic.
› Forums › Speech Synthesis › Merlin › Different results with the same scripts
When I follow the instructions with msc_demo_02 for DNN training for my own data, when running the same scripts, sometimes there are errors and sometimes it goes alright, which seems very mysterious.
1. After the first step (training using train.py), in the folder “dur_data”, sometimes only parts of the data get their respective dur_data files, and the rest are missing. Sometimes when I run this script for the same data, all the dur_data files are there.
2. When I train acoustic model and duration model using run_dnn.py with respective conf files, sometimes an error message occurs saying that it cannot find the log file. But when I redo this step, it works.
3. When synthesis, an error sometimes occurs saying:
AttributeError: ‘NNDurationPredictor’ object has no attribute ‘model’
and no wav file is generated.
But sometimes I do not get this message, and the generated wav file seems fine.
A possible explanation is filesystem / network glitches – are you working on the iMacs in your Documents folder? If so, try making a copy of everything on to the machine’s own local disk and running there.
Simon
I am working on a Dice machine and have my own copy of Ossian…
Some forums are only available if you are logged in. Searching will only return results from those forums if you log in.
Copyright © 2024 · Balance Child Theme on Genesis Framework · WordPress · Log in