README.md

---
license: apache-2.0
language:
- as
- bn
- bo
- gu
- hi
- kn
- ks
- kok
- mai
- ml
- mni
- mr
- ne
- or
- pa
- sa
- sat
- sd
- ta
- te
- ur
- doi
- raj
- en
base_model:
- SWivid/F5-TTS
library_name: transformers
pipeline_tag: text-to-speech
tags:
- text-to-speech
- tts
- multilingual
- indic-languages
widget:
- text: >-
    2026 లో ఈ project లో మా ultimate goal ఏంటంటే, ఒక Telugu speaker తన daily
    conversation లో naturally English words ఉపయోగించినప్పుడు, మన T T S system
    కూడా అదే style ని understand చేసి, unnecessary pauses లేకుండా చాలా smooth
    మరియు natural speech ని generate చేయాలి.
  example_title: CodeMix-Telugu
  output:
    url: examples/codemix_te.wav
- text: >-
    எங்களுடைய plan என்னவென்றால், இந்த project-ஐ பெரிய scale-க்கு கொண்டு செல்வது,
    அதற்காக 2026 முழுவதும் புதிய training data மற்றும் high-quality audio
    samples-ஐ தொடர்ந்து collect செய்வதாகும்.
  example_title: CodeMix-Tamil
  output:
    url: examples/codemix_ta.wav
- text: >-
    हम इस project को 2026 तक बड़े scale पर ले जाने की planning कर रहे हैं, इसलिए
    नए training data और high-quality audio samples लगातार collect कर रहे हैं।
  example_title: CodeMix-Hindi
  output:
    url: examples/codemix_hi.wav
- text: >-
    విద్య మన జీవితానికి వెలుగునిచ్చే శక్తి. అది జ్ఞానం, ఆత్మవిశ్వాసం, మంచి
    ఆలోచనా విధానాన్ని పెంచుతుంది. ప్రతి వ్యక్తి అభివృద్ధికి, సమాజ పురోగతికి
    విద్య ఎంతో ముఖ్యం
  example_title: Telugu1
  output:
    url: examples/example1_te.wav
- text: >-
    பள்ளிகூடத்திலேயே நம்மள தான் மாணவர்கள் எல்லாருக்கும் பிடிக்கும்னு எல்லாரும்
    பேசிக்குறாங்க.
  example_title: Tamil1
  output:
    url: examples/example1_ta.wav
- text: >-
    उन्होंने टाटा इंस्टीट्यूट ऑफ़ सोशल साइंसेज़, टाटा इंस्टीट्यूट ऑफ़ फ़ंडामेंटल
    रिसर्च और नेशनल सेंटर फ़ॉर परफ़ॉर्मिंग आर्ट्स की भी स्थापना की।
  example_title: Hindi1
  output:
    url: examples/example1_hi.wav
- text: >-
    The grain was of such excellent quality, that the likes of it had never been
    seen before.
  example_title: English1
  output:
    url: examples/example1_en.wav
datasets:
- ai4bharat/IndicVoices
- ai4bharat/Rasa
---
# SPRING_F5: Fine-tuned F5-TTS for 23 Indian Languages & English

**SPRING_F5** is a multilingual text-to-speech (TTS) model based on **F5-TTS**, fine-tuned to support **23 Indian Language & English**.

## Supported Languages

SPRING_F5 supports the following 24 languages:

- Assamese, Bengali, Bodo, Gujarati, Hindi, Kannada, Kashmiri, Konkani, Maithili, Malayalam, Manipuri, Marathi, Nepali, Odia, Punjabi, Sanskrit, Santali, Sindhi, Tamil, Telugu, Urdu, Dogri, Rajasthani and English
 
## 🚀 Installation

We recommend using a dedicated Conda environment.

```bash
conda create -n springf5 python=3.10 -y
conda activate springf5
pip install git+https://github.com/ArigalaAdarsh/SPRING_F5.git
```
 
 
```python
from transformers import AutoModel
import numpy as np
import soundfile as sf

# Load INF5 from Hugging Face
repo_id = "SPRINGLab/SPRING_F5"
model = AutoModel.from_pretrained(repo_id, trust_remote_code=True)

# Generate speech
audio = model(" 2026 లో ఈ project లో మా ultimate goal ఏంటంటే, ఒక Telugu speaker తన daily conversation లో naturally English words ఉపయోగించినప్పుడు, మన T T S system కూడా అదే style ని understand చేసి, unnecessary pauses లేకుండా చాలా smooth మరియు natural speech ని generate చేయాలి.", 
            ref_audio_path="prompt_audios/example1_te.wav",
            ref_text="వందేభారత్ రైలు ఆధునిక భారతదేశం యొక్క వైభవోపేతమైన చిత్రాలలో ఒకటిగా ఉంది",
            lang='te'  # Language ID is used for number-to-Indic word conversion.
                  )

# Normalize and save output
if audio.dtype == np.int16:
    audio = audio.astype(np.float32) / 32768.0
sf.write("generated_audios/outputs/example.wav", np.array(audio, dtype=np.float32), samplerate=24000)
```

## Training Details

SPRING_F5 was trained using **2 × NVIDIA H200 GPUs** for approximately **two weeks**.

We would like to sincerely thank **C-DAC** for providing the computational resources required to train this model.

We used the **F5-TTS Base configuration**, with approximately **330M parameters**, following the model configuration described in the original F5-TTS work.

### Training Data

The model was trained on approximately **3,220 hours of high-quality speech data** collected from the following datasets:

- **[Rasa](https://huggingface.co/datasets/ai4bharat/Rasa)**
- **[IndicTTS](https://www.iitm.ac.in/donlab/indictts/database)**
- **[IndicVoices-R](https://huggingface.co/datasets/ai4bharat/indicvoices_r)**

These datasets provide diverse multilingual speech data covering the Indian languages supported by SPRING_F5.

---

## References

We would like to express our sincere gratitude to the authors and contributors of **[F5-TTS](https://github.com/SWivid/F5-TTS)** for their valuable contributions to text-to-speech research and for providing the foundation for this work.

SPRING_F5 builds upon the F5-TTS architecture and training methodology and extends it toward multilingual speech synthesis for Indian languages.

```bibtex
@misc{spring_f5_2026,
  author       = {Adarsh Arigala},
  title        = {SPRING_F5},
  year         = {2026},
  url          = {https://github.com/arigalaadarsh/SPRING_F5},
}
```