| | --- |
| | license: apache-2.0 |
| | language: |
| | - en |
| | - es |
| | - zh |
| | tags: |
| | - mlx |
| | - uncensored |
| | - abliterated |
| | - osirisbrain |
| | - apple-silicon |
| | - qwen2.5-coder |
| | - code-generation |
| | base_model: Qwen/Qwen2.5-Coder-7B-Instruct |
| | pipeline_tag: text-generation |
| | library_name: mlx |
| | --- |
| | |
| | # OsirisPtah-Coder-v7-MLX |
| |
|
| | **The Ptah** — Osiris's dedicated coding and hacking brain. Fully uncensored (abliterated). Runs natively on Apple Silicon via MLX Metal. |
| |
|
| | ## Architecture |
| |
|
| | - **Base Model:** Qwen2.5-Coder-7B-Instruct (7 billion parameters) |
| | - **Modification:** Abliterated by huihui-ai, converted to MLX 4-bit by OsirisBrain |
| | - **Format:** MLX 4-bit quantized (4.501 bits/weight) |
| | - **Size:** ~4.0 GB |
| | - **Speed:** ~120-180 tokens/sec on M2 Pro (MLX Metal) |
| | - **Specialization:** Code generation, debugging, security analysis, full-stack development |
| |
|
| | ## Usage |
| |
|
| | ```python |
| | from mlx_lm import load, generate |
| | |
| | model, tokenizer = load("osirisbrain/OsirisPtah-Coder-v7-MLX") |
| | prompt = tokenizer.apply_chat_template( |
| | [{"role": "user", "content": "Write a TypeScript WebSocket server"}], |
| | add_generation_prompt=True |
| | ) |
| | response = generate(model, tokenizer, prompt=prompt, max_tokens=2048) |
| | ``` |
| |
|
| | ## Credits |
| |
|
| | Abliterated by [huihui-ai](https://huggingface.co/huihui-ai/Qwen2.5-Coder-7B-Instruct-abliterated). |
| | Original model: [Qwen/Qwen2.5-Coder-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct) by Alibaba. |
| |
|
