so
Upload model checkpoints
49907a3 verified
|
Raw
History Blame Contribute Delete
1.03 kB
metadata
license: cc-by-4.0
library_name: f5-tts
datasets:
  - SPRINGLab/IndicTTS-Hindi
  - SPRINGLab/IndicVoices-R_Hindi
language:
  - hi
pipeline_tag: text-to-speech
base_model:
  - SPRINGLab/F5-Hindi-24KHz
tags:
  - f5-tts
  - hindi
  - tts
  - fine-tuned

F5-TTS Hindi

Fine-tuned F5-TTS model for Hindi text-to-speech.

Training Details

  • Base model: SPRINGLab/F5-Hindi-24KHz (151M params, F5-TTS Small)
  • Datasets: IndicTTS-Hindi + IndicVoices-R_Hindi
  • Learning rate: 1e-05
  • Epochs: 10
  • Batch size: 200 frames/GPU
  • Architecture: DiT (dim=768, depth=18, heads=12) + ConvNeXt V2 (dim=512, layers=4)
  • Audio: 24kHz, 100-dim mel spectrogram, Vocos vocoder

Usage

from f5_tts.api import F5TTS
model = F5TTS(
    model_type="F5-TTS",
    ckpt_file="path/to/model_last.pt",  
    vocab_file="path/to/vocab.txt",
)
wav, sr, _ = model.infer(
    ref_file="reference.wav",
    ref_text="संदर्भ पाठ हिंदी में",
    gen_text="आपका दिन शुभ हो",
)