Instructions to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="nightmedia/Qwen3.8-27B-Brainwaves-WFH")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("nightmedia/Qwen3.8-27B-Brainwaves-WFH")
model = AutoModelForMultimodalLM.from_pretrained("nightmedia/Qwen3.8-27B-Brainwaves-WFH", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

MLX

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("nightmedia/Qwen3.8-27B-Brainwaves-WFH")
config = load_config("nightmedia/Qwen3.8-27B-Brainwaves-WFH")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

vLLM

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nightmedia/Qwen3.8-27B-Brainwaves-WFH"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nightmedia/Qwen3.8-27B-Brainwaves-WFH",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/nightmedia/Qwen3.8-27B-Brainwaves-WFH

SGLang

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "nightmedia/Qwen3.8-27B-Brainwaves-WFH" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nightmedia/Qwen3.8-27B-Brainwaves-WFH",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "nightmedia/Qwen3.8-27B-Brainwaves-WFH" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nightmedia/Qwen3.8-27B-Brainwaves-WFH",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Unsloth Desktop

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "nightmedia/Qwen3.8-27B-Brainwaves-WFH"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "nightmedia/Qwen3.8-27B-Brainwaves-WFH"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with Docker Model Runner:
```
docker model run hf.co/nightmedia/Qwen3.8-27B-Brainwaves-WFH
```

Hermes Agent

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "nightmedia/Qwen3.8-27B-Brainwaves-WFH"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default nightmedia/Qwen3.8-27B-Brainwaves-WFH

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use nightmedia/Qwen3.8-27B-Brainwaves-WFH with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "nightmedia/Qwen3.8-27B-Brainwaves-WFH"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "nightmedia/Qwen3.8-27B-Brainwaves-WFH" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Benchmarks

by zviratko - opened 3 days ago

Discussion

zviratko

3 days ago

•

edited 3 days ago

For your modelcard if you want :) for some weird reason your finetunes always score better at 4-5bits ¯\_(ツ)_/¯

So far I like it very much, though I have seen some weird looping, but I think it was some chat template weirdness.

Intelligence Benchmark Comparison

              Mode    Sampled                   Qwen3.8-27B-Brainwaves-WFH   Qwen3.8-27B-Brainwaves-WFH-oQ4e-mtp   Qwen3.8-27B-Brainwaves-WFH-oQ5e-mtp   Qwen3.8-27B-Brainwaves-WFH-oQ6e-mtp   Qwen3.8-27B-Brainwaves-WFH-oQ8e-mtp  Qwen3.8-27B-Brainwaves-WFH-mxfp4-mlx
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
MMLU          Sample  300/14042                                      85.0%                                 86.0%                                 85.7%                                 85.0%                                 85.0%                                 85.7%
MMLU_PRO      Sample  300/12032                                      69.3%                                 69.3%                                 71.0%                                 69.0%                                 68.0%                                 69.0%
HUMANEVAL     Full    164                                            88.4%                                 88.4%                                 87.2%                                 87.2%                                 88.4%                                 87.2%

--- Detail ---

Model: Qwen3.8-27B-Brainwaves-WFH
Benchmark         Accuracy   Correct   Total   Time(s)   Think
--------------------------------------------------------------
MMLU                 85.0%       255     300     398.6      No
MMLU_PRO             69.3%       208     300     266.8      No
HUMANEVAL            88.4%       145     164    1083.7      No

Model: Qwen3.8-27B-Brainwaves-WFH-oQ4e-mtp
Benchmark         Accuracy   Correct   Total   Time(s)   Think
--------------------------------------------------------------
MMLU                 86.0%       258     300     386.2      No
MMLU_PRO             69.3%       208     300       198      No
HUMANEVAL            88.4%       145     164     372.4      No

Model: Qwen3.8-27B-Brainwaves-WFH-oQ5e-mtp
Benchmark         Accuracy   Correct   Total   Time(s)   Think
--------------------------------------------------------------
MMLU                 85.7%       257     300     469.9      No
MMLU_PRO             71.0%       213     300     209.2      No
HUMANEVAL            87.2%       143     164     459.2      No

Model: Qwen3.8-27B-Brainwaves-WFH-oQ6e-mtp
Benchmark         Accuracy   Correct   Total   Time(s)   Think
--------------------------------------------------------------
MMLU                 85.0%       255     300       482      No
MMLU_PRO             69.0%       207     300     223.5      No
HUMANEVAL            87.2%       143     164     486.4      No

Model: Qwen3.8-27B-Brainwaves-WFH-oQ8e-mtp
Benchmark         Accuracy   Correct   Total   Time(s)   Think
--------------------------------------------------------------
MMLU                 85.0%       255     300     485.4      No
MMLU_PRO             68.0%       204     300     232.3      No
HUMANEVAL            88.4%       145     164     522.7      No

Model: Qwen3.8-27B-Brainwaves-WFH-mxfp4-mlx
Benchmark         Accuracy   Correct   Total   Time(s)   Think
--------------------------------------------------------------
MMLU                 85.7%       257     300     339.5      No
MMLU_PRO             69.0%       207     300     174.7      No
HUMANEVAL            87.2%       143     164     420.8      No

nightmedia

Owner 3 days ago

wow, nice, excellent results!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment