Instructions to use cagataydev/tin-4b-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use cagataydev/tin-4b-4bit with MLX:
# Make sure mlx-vlm is installed # pip install --upgrade mlx-vlm from mlx_vlm import load, generate from mlx_vlm.prompt_utils import apply_chat_template from mlx_vlm.utils import load_config # Load the model model, processor = load("cagataydev/tin-4b-4bit") config = load_config("cagataydev/tin-4b-4bit") # Prepare input image = ["http://images.cocodataset.org/val2017/000000039769.jpg"] prompt = "Describe this image." # Apply chat template formatted_prompt = apply_chat_template( processor, config, prompt, num_images=1 ) # Generate output output = generate(model, processor, formatted_prompt, image) print(output) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Pi
How to use cagataydev/tin-4b-4bit with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "cagataydev/tin-4b-4bit"
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "cagataydev/tin-4b-4bit" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent
How to use cagataydev/tin-4b-4bit with Hermes Agent:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "cagataydev/tin-4b-4bit"
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 cagataydev/tin-4b-4bit
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use cagataydev/tin-4b-4bit with OpenClaw:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "cagataydev/tin-4b-4bit"
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 "cagataydev/tin-4b-4bit" \ --custom-provider-id mlx-lm \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
Tin 4B Β· MLX 4-bit
A small model that knows Tin, the typed programming language where agents and tools are part of the syntax. It writes Tin programs, fixes them from the compiler's errors, and runs a check-and-fix loop with tin_check / tin_run tools. Made by ΓaΔatay ΓalΔ± (@cagataycali).
The bf16 weights for further tuning are cagataydev/tin-4b. Post-tuned from cagataydev/loopl-4b (Qwen3.5-4B).
Scores
| probe | Tin 4B | Qwen3.5 4B |
|---|---|---|
| Says it is Tin | 100 | 60.4 |
| Tin knowledge | 76.1 | 39.4 |
| Writes Tin | 68.3 | 3.5 |
| Fixes Tin | 97.2 | 80.4 |
| Fixes stacked bugs | 93.9 | 37.8 |
| Agent loop Β· native | 65 | 15 |
% pass on a held-out set, graded by the real Tin compiler. "Writes Tin" means the program compiles and prints exactly the expected output on the first try.
Run it
curl -fsSL https://tin.ad/install.sh | bash
tin model pull tin:4b # then any agent without a model: line uses it
Or directly with mlx-lm:
pip install -U mlx-lm
mlx_lm.generate --model cagataydev/tin-4b-4bit --prompt "Write a Tin program with a guarded tool fn deleteFile."
All sizes
| size | writes Tin | fixes Tin | agent loop | phone build |
|---|---|---|---|---|
| 0.8B | 37.5 | 88.8 | 32.5 | tin-0.8b-4bit |
| 2B | 54.1 | 92.5 | 55 | tin-2b-v3-4bit |
| 4B | 68.3 | 97.2 | 65 | tin-4b-4bit |
Limits
- About 32% of first tries fail
tin checkor print the wrong thing, so always run the compiler. - It knows Tin as of the training commit; newer std functions may be missing.
- Small model: details like flag names and error codes are often wrong.
Licence
Apache-2.0, like the base Qwen3.5 model.
- Downloads last month
- 19
4-bit
Model tree for cagataydev/tin-4b-4bit
Base model
Qwen/Qwen3.5-4B-Base