This is somali TTS model using microsoft/speecht5_tts as base model this was trained on over 67 hrs total audio dataset!