Can Riva TTS checkpoints in .ckpt be converted to .tlt?

For speech synthesis and finetuning we need pretrained checkpoints in .tlt format. can we create them in lj speech or any other dataset by simply converting the existing checkpoints?

Please provide the following information when requesting support.

Hardware - GPU (A100/A30/T4/V100)
Hardware - CPU
Operating System
Riva Version
TLT Version (if relevant)
How to reproduce the issue ? (This is for errors. Please share the command and the detailed log here)

Hi @pragya.chaturvedi

Thanks for your interest in Riva,

can you kindly let me know what is the format of your current model checkpoints

For Riva TTS Deployment, we can deploy pre-trained models (including .nemo , .riva , .tlt and .rmir assets).


.ckpt or .nemo