Instructions to use datacrunch/tolkio-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use datacrunch/tolkio-models with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf datacrunch/tolkio-models
# Run inference directly in the terminal:
llama cli -hf datacrunch/tolkio-models

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf datacrunch/tolkio-models
# Run inference directly in the terminal:
llama cli -hf datacrunch/tolkio-models

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf datacrunch/tolkio-models
# Run inference directly in the terminal:
./build/bin/llama cli -hf datacrunch/tolkio-models

Use Docker

docker model run hf.co/datacrunch/tolkio-models

Llama new
LM Studio
Jan
Ollama
How to use datacrunch/tolkio-models with Ollama:
```
ollama run hf.co/datacrunch/tolkio-models
```
Unsloth Desktop

How to use datacrunch/tolkio-models with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf datacrunch/tolkio-models

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "datacrunch/tolkio-models"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use datacrunch/tolkio-models with Docker Model Runner:
```
docker model run hf.co/datacrunch/tolkio-models
```

Lemonade

How to use datacrunch/tolkio-models with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull datacrunch/tolkio-models

Run and chat with the model

lemonade run user.tolkio-models-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use datacrunch/tolkio-models with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf datacrunch/tolkio-models

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default datacrunch/tolkio-models

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use datacrunch/tolkio-models with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf datacrunch/tolkio-models

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "datacrunch/tolkio-models" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Tolkio models

The models the Tolkio app (an offline speech translator for Android) downloads. Each keeps its own license:

Folder	Model	License
asr/	Parakeet Redux (Moondream, from NVIDIA Parakeet TDT), converted to ONNX, 2-bit packed	CC BY 4.0
vad/	Silero VAD	MIT
mt/, mt13/	Index-Translate 2B (bilibili Index team), GGUF 4-bit / 8-bit	Apache 2.0
tts/	Supertonic 3 (Supertone); vocoder.onnx with 16-bit weights (cast back to 32-bit when loaded)	OpenRAIL-M (use restrictions in its license)
spk/	CAM++ (3D-Speaker, Apache 2.0; 16-bit weights), pyannote segmentation 3.0 (MIT), TIGER (Apache 2.0)	as listed
africa/asr/	Omnilingual ASR 300M (Meta), sherpa-onnx int8 export (African languages and Korean)	Apache 2.0
africa/mt*/	TranslatePsy-AfriSLM 0.8B / 2B (QVAC), GGUF 4-bit	Apache 2.0
voices/	VITS OpenBible voices (multilingual-tts), converted to ONNX, 16-bit weights	CC BY-SA 4.0
asia/ja/asr/	NVIDIA Parakeet TDT-CTC 0.6B ja (nvidia/parakeet-tdt_ctc-0.6b-ja), its CTC part as sherpa-onnx's int8 export; changed: matrix layers re-quantized to 4 bits (ONNX Runtime MatMulNBits, block 128) for this app (model.int4.onnx)	CC BY 4.0
asia/zh/asr/	X-ASR zh-en Zipformer transducer with punctuation (Gilgamesh-J/X-ASR), sherpa-onnx int8 export	Apache 2.0
asia/zh/voice/	Kokoro 82M v1.1-zh (hexgrad), sherpa-onnx int8 export, with its word lists; voices.bin holds the 10 voices the app uses	Apache 2.0
lid/	Silero Language Classifier 95 (Silero Team): which language is being spoken, to pick the right recogniser	MIT

Install from a folder instead of downloading: copy this folder to the phone and use "Import from a folder" in the app.

Downloads last month: 4,313

GGUF

Model size

2B params

Architecture

qwen35

Hardware compatibility

We're not able to determine the quantization variants.

View all variants