Skip to content
Computer vision

Qwen3.8-27B-opencode-MLX-adapter

faster local coding on Apple Silicon

1 min read

This LoRA adapter (a lightweight add-on for a base model) brings Qwen/Qwen3.8-27B to the mlx-lm format for Apple Silicon chips. Trained with the AsyncGRPO reinforcement learning algorithm, it targets software engineering and terminal workflows inside the OpenCode agent. It skips unnecessary reasoning output to deliver faster, more accurate results. It is a solid fit for developers running local models on a Mac to write and debug code.

Key features

  • Built specifically for coding tasks and terminal workflows.
  • Trained using AsyncGRPO within the TRL framework.
  • Uses the mlx-lm format with rank 32 and scale 2.0 for seamless MLX integration.
  • Compact adapter weights at just 638 MB.

What you can do

  • Pair the adapter with a base or quantized Qwen3.8-27B model for local code generation.
  • Run it as the backend engine for the OpenCode agent on a Mac.
  • Spin up a local server with mlx_lm.server to handle API requests.

Example

pip install -U "mlx-lm>=0.32.2"
mlx_lm.generate --model mlx-community/Qwen3.8-27B-4bit --adapter-path netprojects/Qwen3.8-27B-opencode-MLX-adapter \ --chat-template-config '{"enable_thinking": false}' --prompt "Write a bash one-liner that counts lines in all .py files." --max-tokens 300
Read moreDetails
• SWE-bench Pro (hard 51): improved from 35.3% to 51.0%
• Terminal-Bench 2.1: improved from 46.4% to 52.4%
• License: Apache-2.0
• Requirements: compatible mlx-lm build and the base Qwen3.8-27B model
Discuss Telegram 0

Will it run?

Working out the memory needs

This is a LoRA: it only runs on top of the base model harness-hub/Qwen3.8-27B-opencode-adapted. The base model's size is still being collected; the estimate will appear within a day.

Massed ComputeOpen Massed Compute

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # netprojects/Qwen3.8-27B-opencode-MLX-adapter
    # ===== Tokens: paste yours between the quotes =====
    export HF_TOKEN=""        # huggingface.co/settings/tokens, for gated models
    # ==========================================
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download harness-hub/Qwen3.8-27B-opencode-adapted --local-dir ~/models/Qwen3.8-27B-opencode-adapted
    hf download netprojects/Qwen3.8-27B-opencode-MLX-adapter --local-dir ~/models/Qwen3.8-27B-opencode-MLX-adapter
    # The model is in ~/models. How to run it is in the model card (link above).
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

Requires the base model

harness-hub/Qwen3.8-27B-opencode-adapted

This model is a LoRA / adapter. Download the base model first, then load this one on top.

  1. Qwen3.8-27Bfamily root
  2. Qwen3.8-27B-opencode-adapted
  3. Qwen3.8-27B-opencode-MLX-adapter

New posts about versions of Qwen3.8-27B (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 2All on Hugging Face

About the model

netprojects/Qwen3.8-27B-opencode-MLX-adapter
Author
netprojects
Category
Computer vision
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments