Give Feedback
0 of 0
About this mod
Fastpitch models that can be used with xVA Synth for voice types FemaleChild01, FemaleChild02, MaleChild01 and MaleChild02
- Requirements
- Permissions and credits
- Changelogs
These are PyTorch state_dict files of Fastpitch models trained withe the dialogue lines from Fallout New Vegasf or FemaleChild01, FemaleChild02, MaleChild01 and MaleChild02 voice types.
You can use these with the scripts provided on the Fastpitch website (check 6: Start inference/predictions) or with the NVVA Synth tool (drop the files in:
xVASynth\resources\app\models\skyrim).
The author of xVASynth stated that they will not release their models for child voice types, so I decided to train the models myself. I think the results are reasonable (except FemaleChild02 which is not as good because the are too few dialogue lines).
Here are some examples (the generated audio begins obviously after the intro line):
No preview video for FemaleChild02 because I could not figure out why my dialogue was not added to the game.
If you like this mod, you might be interested in the child voices I trained for Skyrim or Fallout 4 Shaun.
You can use these with the scripts provided on the Fastpitch website (check 6: Start inference/predictions) or with the NVVA Synth tool (drop the files in:
xVASynth\resources\app\models\skyrim).
The author of xVASynth stated that they will not release their models for child voice types, so I decided to train the models myself. I think the results are reasonable (except FemaleChild02 which is not as good because the are too few dialogue lines).
Here are some examples (the generated audio begins obviously after the intro line):
No preview video for FemaleChild02 because I could not figure out why my dialogue was not added to the game.
If you like this mod, you might be interested in the child voices I trained for Skyrim or Fallout 4 Shaun.