V1 Checkpoints released

With this initial release, we are providing the checkpoints for a variant of CLARIS trained on the following datasets:

1. Whispers from 44 speakers of wTIMIT dataset
2. Dysarthric speech of all 8 speakers in TORGO dataset
3. Updated Code HiFi-GAN Vocoder checkpoint (previously uploaded checkpoint was corrupted).