---
language:
  - as
license: cc-by-sa-4.0
library_name: f5-tts
tags:
  - text-to-speech
  - tts
  - f5-tts
  - open-bible
  - assamese
pipeline_tag: text-to-speech
base_model: SWivid/F5-TTS
datasets:
  - davidguzmanr/open-bible-resources
inference: false
---

# F5-TTS Open Bible — Assamese

A zero-shot text-to-speech model for **Assamese**, trained from scratch on
the [Open Bible](https://huggingface.co/datasets/davidguzmanr/open-bible-resources)
corpus using the [F5-TTS](https://github.com/SWivid/F5-TTS) architecture
(diffusion transformer with vocos vocoder, 24 kHz output).

The model takes a short reference audio clip (5–10 seconds) and a target text,
and synthesises the target text in the voice of the reference speaker. No
fine-tuning per voice is required.

## Files

| File | Purpose |
|------|---------|
| `model_last.pt` | Trained model weights. |
| `vocab.txt` | Character vocabulary built from the training transcripts. |
| `F5-TTS_OpenBible_Assamese.yaml` | Hydra training/inference config (architecture, mel spec settings, tokenizer). |

## Intended use

- Zero-shot TTS for Assamese, controlled by a user-supplied reference clip.
- Research on multilingual TTS, low-resource TTS evaluation, and listening
  studies on Open Bible–style read-speech.

## How to use

Install F5-TTS:

```bash
pip install git+https://github.com/SWivid/F5-TTS.git
```

Download the checkpoint and run inference:

```python
import torch
from huggingface_hub import hf_hub_download
from hydra.utils import get_class
from omegaconf import OmegaConf
from f5_tts.infer.utils_infer import infer_process, load_model, load_vocoder, preprocess_ref_audio_text

repo_id = "multilingual-tts/F5-TTS-OpenBible-Assamese"
ckpt   = hf_hub_download(repo_id, "model_last.pt")
vocab  = hf_hub_download(repo_id, "vocab.txt")
config = hf_hub_download(repo_id, "F5-TTS_OpenBible_Assamese.yaml")

device = "cuda" if torch.cuda.is_available() else "cpu"

model_cfg = OmegaConf.load(config)
model_cls = get_class(f"f5_tts.model.{model_cfg.model.backbone}")

vocoder = load_vocoder(vocoder_name="vocos", is_local=False, device=device)
model   = load_model(
    model_cls, model_cfg.model.arch, ckpt,
    mel_spec_type="vocos", vocab_file=vocab, use_ema=True, device=device,
)

# Supply your own clean reference clip — 5–10 s, single speaker and its transcription.
ref_audio = "/path/to/your-assamese-clip.wav"
ref_text  = "Exact transcription of the clip"
gen_text  = "..."   # text to synthesise in Assamese

ref_audio_proc, ref_text_proc = preprocess_ref_audio_text(ref_audio, ref_text)
wav, sr, _ = infer_process(
    ref_audio_proc, ref_text_proc, gen_text, model, vocoder,
    mel_spec_type="vocos", device=device,
)
```

## Training data

- **Source:** `davidguzmanr/open-bible-resources`, config `Assamese`
- **Size:** approximately 30,500 utterances
- **Speakers:** multispeaker; speaker identity is supplied at inference time
  via the reference clip, not by a fixed speaker id
- **Sample rate:** 24 kHz
- **Maximum utterance duration during training:** 15 s

## Training procedure

- Base architecture: F5-TTS v1 Base (DiT, 1024 dim, 22 layers, 16 heads,
  text dim 512, 4 convolutional layers).
- Tokenizer: custom character-level, built from the training transcripts.
- Vocoder: vocos.
- Mel spectrogram: 100 channels, hop 256, win 1024, n_fft 1024.
- Optimizer: AdamW, learning rate 7.5e-5, 20 000 warmup updates.
- Training budget: 500,000 optimizer updates on 4 GPUs with mixed precision
  (bf16), global batch ≈ 112,000 frames.

Audio preprocessing, vocab generation, and config sizing are reproducible via
the upstream
[open-bible-models](https://github.com/davidguzmanr/open-bible-models) repo.

## Evaluation

Evaluated alongside other Open-Bible TTS systems on character/word error rate
(via Meta's Omnilingual ASR) and UTMOSv2 naturalness scores. See the
[open-bible-models](https://github.com/davidguzmanr/open-bible-models) repository
for the evaluation pipeline and the
[open-bible-surveys](https://github.com/davidguzmanr/open-bible-surveys) repository
for the human-listening survey methodology.