Skip to content
Audio

Turaco-nano-tts-wes

lightweight TTS model for Pidgin English

2 min read

Built by Clevaway, Turaco-nano-tts-wes generates spoken Cameroonian Pidgin English (WES) from text prompts. Based on the Kyutai Pocket TTS architecture, this tiny language model packs around 109.5 million parameters and weighs just 438 MB in FP32. The release is built to run locally on standard CPUs, integrate into mobile apps, or run right in the browser via WebAssembly.

Key features

  • Compact footprint: 14.9x smaller than the previous Turaco CSM version, making integration much easier.
  • Pilot voice: Comes with a precomputed voice state for fast generation without needing the original dataset.
  • Browser-ready: Supports WebAssembly execution via the Phonon runtime for running directly in a web UI.
  • Targeted training: Fine-tuned on the Common Voice 27.0 corpus to handle this low-resource African dialect.

What you can do

  • Local speech synthesis: Generate Cameroonian Pidgin English audio locally on a CPU using the Python library.
  • In-browser demo: Test the model directly in your browser through a dedicated WebAssembly runtime without installing heavy software.
  • App integration: Embed the lightweight weights into mobile apps and edge devices.

Example

from pocket_tts import TTSModel
model = TTSModel.load_model(config="hf://fotiecodes/turaco-nano-tts-wes/wes.yaml")
state = model.get_state_for_audio_prompt("hf://fotiecodes/turaco-nano-tts-wes/voice.safetensors")
audio = model.generate_audio(state, "I di go for ma wok.")
Read moreDetails
Base model: Kyutai Pocket TTS
Parameters: 109.5M
Weights size: 438 MB (FP32)
Browser GGUF size: ~177 MB
Training: 200 optimizer steps on a Tesla T4 GPU
Discuss Telegram 0

Will it run?

Runs in F32

Fits without quantisation.

  • F32weights 0.4 GB · ready files0.9 GB
  • BF16weights 0.2 GB0.7 GB
  • FP8 / INT8recommendedweights 0.1 GB0.6 GB
  • 4-bitweights 0.1 GB0.6 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 2060 · F32 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # fotiecodes/turaco-nano-tts-wes
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download fotiecodes/turaco-nano-tts-wes --local-dir ~/models/turaco-nano-tts-wes
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to turaco-nano-tts-wes
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of pocket-tts (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face

About the model

fotiecodes/turaco-nano-tts-wes
Author
fotiecodes
Category
Audio
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments