README.md

---
language: en
license: mit
tags:
  - aro
  - code-generation
  - dsl
  - mlx
  - 4-bit
  - lora
  - fine-tuned
base_model: mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit
pipeline_tag: text-generation
library_name: mlx
---

# ARO Coder

A fine-tuned code generation model specialised in the **ARO** (Action Result Object) programming language.

ARO is a domain-specific language where every statement follows the pattern:
`Verb the <Result> preposition [the] <Object>`.

| | |
|---|---|
| **Base model** | [mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit](https://huggingface.co/mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit) |
| **Quantization** | 4-bit (MLX) |
| **Language** | ARO |
| **Training samples** | 3630 |
| **Syntax pass rate** | 58% |
| **Source label** | distill_student |

## Links

- **Website**: [arolang.github.io/aro](https://arolang.github.io/aro/)
- **GitHub**: [github.com/arolang/aro](https://github.com/arolang/aro)
- **Documentation**: [Wiki](https://github.com/arolang/aro/wiki)
- **Language Guide (PDF)**: [Download](https://github.com/arolang/aro/releases/latest/download/ARO-Language-Guide.pdf)
- **Discussions**: [GitHub Discussions](https://github.com/arolang/aro/discussions)

## Quick Start

### MLX (Apple Silicon)

```python
from mlx_lm import load, generate

model, tokenizer = load("ARO-Lang/aro-coder-4bit")

messages = [
    {"role": "system", "content": "You are an expert ARO programmer."},
    {"role": "user", "content": "Write an ARO feature set that retrieves a user by ID and returns an OK response."},
]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
response = generate(model, tokenizer, prompt=prompt, max_tokens=500)
print(response)
```

### MLX Server (OpenAI-compatible API)

```bash
python -m mlx_lm.server --model ARO-Lang/aro-coder-4bit --port 8080

curl http://localhost:8080/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{"model": "aro-coder", "messages": [{"role": "user", "content": "Write hello world in ARO"}]}'
```

### Ollama

```bash
ollama run aro-coder
```

## Example Output

**Prompt:** *Write an ARO Application-Start that starts an HTTP server.*

```aro
(Application-Start: My API) {
    Log "Starting server..." to the <console>.
    Start the <http-server> with <contract>.
    Keepalive the <application> for the <events>.
    Return an <OK: status> for the <startup>.
}
```

## What is ARO?

ARO is a DSL for expressing business features as Action-Result-Object statements.
Every program is a directory of `.aro` files with event-driven feature sets:

```aro
(getUser: User API) {
    Extract the <id> from the <pathParameters: id>.
    Retrieve the <user> from the <user-repository> where id = <id>.
    Return an <OK: status> with <user>.
}
```

Key features:
- **Contract-first HTTP** — routes defined in `openapi.yaml`, feature sets match `operationId`
- **Event-driven** — feature sets triggered by events, not direct calls
- **Immutable bindings** — every transformation produces a new name
- **Happy-path only** — no error handling code; the runtime manages errors

## Training

This model was trained with the ARO training pipeline:

1. **Corpus collection** — 3630 samples from Examples, Book, Wiki, Proposals, and real-world ARO applications
2. **Supervised fine-tuning** — LoRA on all code generation, debugging, Q&A, and explanation tasks
3. **DPO preference training** — using `aro check` validation to build chosen/rejected pairs
4. **Iterative self-improvement** — multiple rounds of generate-validate-retrain

## License

This model and the ARO language are open source under the [MIT License](https://github.com/arolang/aro/blob/main/LICENSE).