Skip to content
Computer vision

Qwen3.8-27B-hermes-adapted-MLX-adapter

A LoRA adapter for programming tasks on Apple Silicon

2 min readupdated

This LoRA adapter (a small plugin built on top of a base model) ports the model from harness-hub into the mlx-lm library format so it can run on Apple Silicon devices. The model was trained using reinforcement learning (RL) to operate within the Hermes Agent for software engineering and terminal tasks. It handles coding and terminal commands much more efficiently, using fewer tokens and steps. The adapter is useful for developers and Mac users who want to run powerful local agents for writing code.

Key features

  • Specialization in software engineering and terminal tasks
  • Optimization for the mlx-lm framework on Mac
  • Improved metrics on SWE-bench Pro and Terminal-Bench tasks
  • Reduced token generation compared to the base model

What you can do

  • Apply it in Hermes Agent to automate engineering tasks
  • Run it via the mlx_lm.server command line for local development
  • Integrate it into Python scripts using mlx_lm
  • Use the adapter on top of quantized versions of the model on Mac

Example

from huggingface_hub import snapshot_download
from mlx_lm import load, generate

adapter = snapshot_download("netprojects/Qwen3.8-27B-hermes-adapted-MLX-adapter")
model, tokenizer = load("mlx-community/Qwen3.8-27B-4bit", adapter_path=adapter)
prompt = tokenizer.apply_chat_template([{"role": "user", "content": "Hello"}], add_generation_prompt=True,
                                       enable_thinking=False, tokenize=False)
print(generate(model, tokenizer, prompt=prompt, max_tokens=200))
Read moreDetails
• SWE-bench Pro (hard 51): 29.4% → 43.1%
• Terminal-Bench 2.1: 44.7% → 58.8%
• Adapter size: 638 MB
• License: Apache-2.0
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-hermes-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-hermes-adapted-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-hermes-adapted --local-dir ~/models/Qwen3.8-27B-hermes-adapted
    hf download netprojects/Qwen3.8-27B-hermes-adapted-MLX-adapter --local-dir ~/models/Qwen3.8-27B-hermes-adapted-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-hermes-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-hermes-adapted
  3. Qwen3.8-27B-hermes-adapted-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-hermes-adapted-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