---
license: apache-2.0
tags:
  - gguf
  - determinex
  - qwen2
---

# determinex-engineer

**C1 Engineer — Builder** in the [Determinex](https://github.com/DarthCeltic/determinex)
compiler-verified multi-agent coding system.

- **Base model:** Qwen2.5-Coder-1.5B-Instruct (1.5B)
- **Fine-tune:** LoRA on the Determinex Semantic DSL corpus, compiler-validated
- **Ollama tag:** `determinex-engineer-v11-dsl`
- **Size:** 1.53 GB (GGUF, quantized)

Determinex never trusts a model's output directly — every generation is
verified against a real compiler/test oracle before it's accepted. This
model is one of three specialists (Engineer / Observer / Sentinel) that
coordinate through a shared latent space (the Rosetta Stone).

## Usage

```bash
ollama create determinex-engineer-v11-dsl -f Modelfile
ollama run determinex-engineer-v11-dsl
```

## License

This model is a fine-tune of a base model licensed under Apache 2.0. Weights
here are released under the same Apache 2.0 terms. See `LICENSE` in this repo.