Unsigned
22
3
voice-tts
:Voice TTS — Kokoro
Text-to-speech model for the voice AI pipeline. Uses Kokoro, an 82M parameter model that produces remarkably natural speech.
Model Details
| Property | Value |
|---|---|
| Model | Kokoro v1.0 |
| Parameters | 82M |
| Size | ~300MB |
| Output | 24kHz audio |
| Default voice | af_heart |
| Framework | PyTorch |
KitOps Usage
# Pack
kit pack . -t jozu.ml/arnabchat2001/voice-tts:v1.0.0
# Push
kit push jozu.ml/arnabchat2001/voice-tts:v1.0.0
# Swap voice model (champion/challenger)
kit pack . -t jozu.ml/arnabchat2001/voice-tts:v2.0.0
kit push jozu.ml/arnabchat2001/voice-tts:v2.0.0
kit diff jozu.ml/arnabchat2001/voice-tts:v1.0.0 jozu.ml/arnabchat2001/voice-tts:v2.0.0
Local Usage
from src.tts_service import TTSService
tts = TTSService()
audio = tts.synthesize("Thank you for calling. How can I help you today?")