CallWhiz/F5-TTS-TELUGU

F5-TTS Telugu 🗣️

Fine-tuned F5-TTS model for Telugu language text-to-speech synthesis with voice cloning capabilities.

Model Description

This is a fine-tuned version of F5-TTS specifically trained for Telugu language. The model has been trained for 250,000 steps and includes an extended vocabulary to support Telugu script characters.

Demo

Files

• model_250000.pt - The fine-tuned model checkpoint (5.1GB)
• vocab.txt - Extended vocabulary file with Telugu characters (2640 tokens)
• f5-tts-clone.mp4 - Voice cloning demonstration

Sample Telugu Text

Common Phrases

• నమస్కారం (namaskāram) - Hello/Greetings
• ధన్యవాదాలు (dhanyavādālu) - Thank you
• మీరు ఎలా ఉన్నారు? (mīru elā unnāru?) - How are you?

Sample Sentences

1. తెలుగు భాష చాలా మధురమైనది మరియు అందమైనది.
   "Telugu language is very sweet and beautiful."

2. కృత్రిమ మేధస్సు సాంకేతికత వేగంగా అభివృద్ధి చెందుతోంది.
   "Artificial intelligence technology is developing rapidly."

3. ఆంధ్రప్రదేశ్ మరియు తెలంగాణ రాష్ట్రాల్లో తెలుగు మాట్లాడతారు.
   "Telugu is spoken in Andhra Pradesh and Telangana states."

Usage

For detailed usage instructions and inference code, please refer to the original F5-TTS repository.

Training Details

• Base Model: F5-TTS
• Language: Telugu
• Training Steps: 250,000
• Vocabulary Size: 2640 tokens

License

CC-BY-NC-4.0

Acknowledgments

Based on the original F5-TTS model architecture.