1.41 GB
76 files
Updated 20 days ago
Name
Size
checkpoints
f5_tts
prompts
samples
.gitattributes2.09 kB
xet
README.md3.32 kB
xet
config.json350 Bytes
xet
model.py5.77 kB
xet
model.safetensors1.4 GB
xet
README.md

IndicF5: High-Quality Text-to-Speech for Indian Languages

We release IndicF5, a near-human polyglot Text-to-Speech (TTS) model trained on 1417 hours of high-quality speech from Rasa, IndicTTS, LIMMITS, and IndicVoices-R.

IndicF5 supports 11 Indian languages:
Assamese, Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, Telugu.


๐Ÿš€ Installation

conda create -n indicf5 python=3.10 -y
conda activate indicf5
pip install git+https://github.com/ai4bharat/IndicF5.git

๐ŸŽ™ Usage

To generate speech, you need to provide three inputs:

  1. Text to synthesize โ€“ The content you want the model to speak.
  2. A reference prompt audio โ€“ An example speech clip that guides the modelโ€™s prosody and speaker characteristics.
  3. Text spoken in the reference prompt audio โ€“ The transcript of the reference prompt audio.
from transformers import AutoModel
import numpy as np
import soundfile as sf

# Load IndicF5 from Hugging Face
repo_id = "ai4bharat/IndicF5"
model = AutoModel.from_pretrained(repo_id, trust_remote_code=True)

# Generate speech
audio = model(
    "เคจเคฎเคธเฅเคคเฅ‡! เคธเค‚เค—เฅ€เคค เค•เฅ€ เคคเคฐเคน เคœเฅ€เคตเคจ เคญเฅ€ เค–เฅ‚เคฌเคธเฅ‚เคฐเคค เคนเฅ‹เคคเคพ เคนเฅˆ, เคฌเคธ เค‡เคธเฅ‡ เคธเคนเฅ€ เคคเคพเคฒ เคฎเฅ‡เค‚ เคœเฅ€เคจเคพ เค†เคจเคพ เคšเคพเคนเคฟเค.",
    ref_audio_path="prompts/PAN_F_HAPPY_00001.wav",
    ref_text="เจญเจนเฉฐเจชเฉ€ เจตเจฟเฉฑเจš เจธเจฎเจพเจฐเจ•เจพเจ‚ เจฆเฉ‡ เจญเจตเจจ เจจเจฟเจฐเจฎเจพเจฃ เจ•เจฒเจพ เจฆเฉ‡ เจตเฉ‡เจฐเจตเฉ‡ เจ—เฉเฉฐเจเจฒเจฆเจพเจฐ เจ…เจคเฉ‡ เจนเฉˆเจฐเจพเจจ เจ•เจฐเจจ เจตเจพเจฒเฉ‡ เจนเจจ, เจœเฉ‹ เจฎเฉˆเจจเฉ‚เฉฐ เจ–เฉเจธเจผ เจ•เจฐเจฆเฉ‡  เจนเจจเฅค"
)

# Normalize and save output
if audio.dtype == np.int16:
    audio = audio.astype(np.float32) / 32768.0
sf.write("namaste.wav", np.array(audio, dtype=np.float32), samplerate=24000)
print("Audio saved succesfully.")

You can find example prompt audios used here.

Terms of Use

By using this model, you agree to only clone voices for which you have explicit permission. Unauthorized voice cloning is strictly prohibited. Any misuse of this model is the responsibility of the user.

References

We would like to extend our gratitude to the authors of F5-TTS for their invaluable contributions and inspiration to this work. Their efforts have played a crucial role in advancing the field of text-to-speech synthesis.

๐Ÿ“– Citation

If you use IndicF5 in your research or projects, please consider citing it:

๐Ÿ”น BibTeX

@misc{AI4Bharat_IndicF5_2025,
  author       = {Praveen S V and Srija Anand and Soma Siddhartha and Mitesh M. Khapra},
  title        = {IndicF5: High-Quality Text-to-Speech for Indian Languages},
  year         = {2025},
  url          = {https://github.com/AI4Bharat/IndicF5},
}
Total size
1.41 GB
Files
76
Last updated
Jul 23
Pre-warmed CDN
US EU US EU

Contributors