Instructions to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp") 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)# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp") model = AutoModelForMultimodalLM.from_pretrained("nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp", 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=256) print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - MLX
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp") config = load_config("nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp") # 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp" # 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp", "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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp
- SGLang
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp" \ --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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp", "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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp" \ --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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp", "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" } } ] } ] }' - Pi
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp"
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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp with Docker Model Runner:
docker model run hf.co/nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp
- Hermes Agent
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp"
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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use nightmedia/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp 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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp"
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/Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp" \ --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"
Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp
This is an experimental merge between:
- nightmedia/Qwen3.5-9B-Brainwaves
- Jackrong/Qwopus3-5-9B-Coder-Flash
- OmniJev/OneJev1.6-9B
arc arc/e boolq hswag obkqa piqa wino
bf16 0.698,0.878,0.904,0.769,0.518,0.802,0.717
mxfp8 0.699,0.878,0.903,0.768,0.518,0.802,0.720
q8-hi 0.699,0.880,0.906,0.770,0.518,0.799,0.721
q6-hi 0.699,0.877,0.905,0.769,0.522,0.803,0.713
oQ6e-mtp 0.697,0.881,0.906,0.768,0.518,0.801,0.721
mxfp4 0.688,0.866,0.899,0.767,0.500,0.807,0.703
gsm8k
mxfp8 90.52%
Quant Perplexity Peak Memory Tokens/sec
bf16 4.096 ยฑ 0.026 24.69 GB 878
mxfp8 4.106 ยฑ 0.026 16.02 GB 724
q8-hi 4.098 ยฑ 0.026 16.86 GB 730
q6-hi 4.101 ยฑ 0.026 14.62 GB 718
oQ6e-mtp 4.101 ยฑ 0.026 14.29 GB 710
mxfp4 4.327 ยฑ 0.028 11.55 GB 727
Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-oQ6e-mtp
ARC_CHALLENGE Sample 300/1172 97.0%
WINOGRANDE Sample 300/1267 85.7%
HUMANEVAL Full 164 79.3%
TRUTHFULQA Full 817 78.0%
SAFETYBENCH Full 11435 86.8%
GSM8K Full 1319 96.3%
MATHQA Full 2985 90.3%
BBQ Full 10864 96.7%
HELLASWAG Full 10042 90.8%
Model components
Qwen3.5-9B-Brainwaves
arc arc/e boolq hswag obkqa piqa wino
bf16 0.686,0.860,0.903,0.768,0.516,0.800,0.713
mxfp8 0.678,0.856,0.904,0.763,0.502,0.800,0.702
q8-hi 0.687,0.860,0.904,0.768,0.518,0.797,0.715
q6-hi 0.686,0.857,0.903,0.766,0.520,0.797,0.710
mxfp4 0.669,0.855,0.894,0.762,0.488,0.800,0.696
gsm8k
mxfp8 89.16%
q8-hi 89.69%
Quant Perplexity Peak Memory Tokens/sec
bf16 4.157 ยฑ 0.027 24.69 GB 767
mxfp8 4.292 ยฑ 0.028 16.02 GB 624
q8-hi 4.156 ยฑ 0.027 16.86 GB 652
q6-hi 4.160 ยฑ 0.027 14.62 GB 597
mxfp4 4.501 ยฑ 0.030 11.55 GB 635
Baseline model
Qwen3.5-9B (Instruct)
arc arc/e boolq hswag obkqa piqa wino
mxfp8 0.571,0.719,0.895,0.683,0.426,0.770,0.671
q8-hi 0.561,0.708,0.892,0.690,0.448,0.764,0.685
q8 0.564,0.713,0.892,0.690,0.442,0.766,0.687
ARC_CHALLENGE: 97.0%
Hit token limit 2 / 300
Finished within limit 298 / 300
Accuracy on finished answers 97.0%
https://omlx.ai/benchmarks/intelligence/2s3mmhvt
WINOGRANDE: 85.7%
Hit token limit 1 / 300
Finished within limit 299 / 300
Accuracy on finished answers 86.0%
https://omlx.ai/benchmarks/intelligence/moccy1ix
TRUTHFULQA: 78.0%
Hit token limit 55 / 817
Finished within limit 762 / 817
Accuracy on finished answers 81.6%
https://omlx.ai/benchmarks/intelligence/mb0utr2u
SAFETYBENCH: 86.8%
Hit token limit 256 / 11435
Finished within limit 11179 / 11435
Accuracy on finished answers 87.7%
Category Score
Physical Health 93.7%
Mental Health 92.1%
Ethics and Morality 90.8%
Illegal Activities 90.1%
Privacy and Property 89.9%
Offensiveness 78.0%
Unfairness and Bias 77.5%
https://omlx.ai/benchmarks/intelligence/4n2w5shr
GSM8K: 96.3%
Hit token limit 10 / 1319
Finished within limit 1309 / 1319
Accuracy on finished answers 96.9%
https://omlx.ai/benchmarks/intelligence/alql4zng
MATHQA: 90.3%
Hit token limit 222 / 2985
Finished within limit 2763 / 2985
Accuracy on finished answers 94.7%
Category Score
gain 92.8%
probability 92.3%
physics 91.6%
geometry 91.5%
general 88.6%
other 87.2%
https://omlx.ai/benchmarks/intelligence/v5mth297
BBQ / Social bias: 96.7%
Hit token limit 84 / 10864
Finished within limit 10780 / 10864
Accuracy on finished answers 97.0%
Category Answered Score
Age 908/1000 90.8%
Disability_status 977/1000 97.7%
Gender_identity 995/1000 99.5%
Nationality 973/1000 97.3%
Physical_appearance 907/1000 90.7%
Race_ethnicity 991/1000 99.1%
Race_x_gender 998/1000 99.8%
Race_x_SES 995/1000 99.5%
Religion 924/1000 92.4%
SES 994/1000 99.4%
Sexual_orientation 841/864 97.3%
https://omlx.ai/benchmarks/intelligence/pt6ikk3k
HELLASWAG: 90.8%
Hit token limit 22 / 10042
Finished within limit 10020 / 10042
Accuracy on finished answers 90.9%
Category Score
Breakdancing 100.0%
Brushing teeth 100.0%
Building sandcastles 100.0%
Bullfighting 100.0%
Cricket 100.0%
Doing fencing 100.0%
Doing kickboxing 100.0%
Doing motocross 100.0%
Fixing bicycle 100.0%
Getting a haircut 100.0%
Getting a tattoo 100.0%
Hitting a pinata 100.0%
Home,Categories 100.0%
Hurling 100.0%
Ironing clothes 100.0%
Javelin throw 100.0%
Kite flying 100.0%
Knitting 100.0%
Paintball 100.0%
Plastering 100.0%
Playing accordion 100.0%
Playing blackjack 100.0%
Playing ice hockey 100.0%
Playing rubik cube 100.0%
Playing squash 100.0%
Pole vault 100.0%
Rock-paper-scissors 100.0%
Sumo 100.0%
Table soccer 100.0%
Trimming branches or hedges 100.0%
Uncategorized 100.0%
Vacuuming floor 100.0%
Washing hands 100.0%
Welding 100.0%
Windsurfing 100.0%
Education and Communications 97.0%
Finance and Business 97.0%
Home and Garden 96.7%
Work World 96.5%
Food and Entertaining 96.0%
Cars & Other Vehicles 95.4%
Health 95.3%
Computers and Electronics 95.2%
Personal Care and Style 94.6%
Pets and Animals 94.5%
Sports and Fitness 93.8%
Family Life 93.7%
Doing crunches 93.3%
Making a cake 93.3%
Youth 93.2%
Powerbocking 92.9%
Mixing drinks 92.3%
Plataform diving 92.3%
Rollerblading 92.3%
Doing a powerbomb 91.7%
Making a sandwich 91.7%
Philosophy and Religion 91.7%
Playing polo 91.7%
Putting in contact lenses 91.7%
Baking cookies 90.9%
Bathing dog 90.9%
Cleaning shoes 90.9%
Clipping cat claws 90.9%
Hanging wallpaper 90.9%
Preparing salad 90.0%
Surfing 90.0%
Ice fishing 89.8%
Holidays and Traditions 89.5%
Painting furniture 89.5%
Carving jack-o-lanterns 88.9%
Dodgeball 88.9%
Ping-pong 88.9%
Playing piano 88.9%
Polishing forniture 88.9%
Removing curlers 88.9%
Rock climbing 88.9%
Shaving 88.9%
Tai chi 88.9%
Playing pool 88.5%
Travel 88.3%
Playing beach volleyball 88.2%
Assembling bicycle 87.5%
Cleaning sink 87.5%
Drum corps 87.5%
Using uneven bars 87.5%
Walking the dog 87.5%
Relationships 86.7%
Beer pong 86.7%
Playing badminton 86.7%
Playing water polo 86.7%
Running a marathon 86.4%
Decorating the Christmas tree 85.7%
Hopscotch 85.7%
Roof shingle removal 85.7%
Smoking a cigarette 85.7%
Swimming 85.7%
Tumbling 85.7%
Using parallel bars 85.7%
Cutting the grass 84.7%
Shuffleboard 84.6%
Tennis serve with ball bouncing 84.6%
Futsal 84.0%
Layup drill in basketball 83.8%
Beach soccer 83.3%
Braiding hair 83.3%
Discus throw 83.3%
Getting a piercing 83.3%
Playing guitarra 83.3%
Preparing pasta 83.3%
Sharpening knives 83.3%
Shot put 83.3%
Spinning 83.3%
Spread mulch 83.3%
Skiing 82.3%
Wrapping presents 82.3%
Blowing leaves 81.8%
Doing karate 81.8%
Drinking beer 81.8%
Installing carpet 81.8%
River tubing 81.8%
Triple jump 81.8%
Clean and jerk 81.8%
Calf roping 81.5%
Gargling mouthwash 81.0%
Washing face 80.9%
High jump 80.3%
Applying sunscreen 80.0%
Drinking coffee 80.0%
Playing bagpipes 80.0%
Shaving legs 80.0%
Smoking hookah 80.0%
Using the pommel horse 80.0%
Using the rowing machine 80.0%
Zumba 80.0%
Playing harmonica 79.4%
Having an ice cream 79.3%
Waterskiing 79.0%
Baton twirling 78.6%
Canoeing 78.6%
Capoeira 78.6%
Polishing shoes 78.6%
Scuba diving 78.3%
Disc dog 78.1%
Chopping wood 77.8%
Mooping floor 77.8%
Playing saxophone 77.8%
Shoveling snow 77.8%
Snow tubing 77.8%
Playing violin 77.7%
Wakeboarding 76.9%
Camel ride 75.0%
Cleaning windows 75.0%
Hand washing clothes 75.0%
Horseback riding 75.0%
Hula hoop 75.0%
Playing congas 75.0%
Sailing 75.0%
Slacklining 75.0%
Cheerleading 73.9%
Elliptical trainer 73.3%
Grooming dog 72.7%
Longboarding 72.7%
Raking leaves 72.7%
Making a lemonade 71.9%
Ballet 71.4%
Kayaking 71.4%
Long jump 71.4%
Rope skipping 71.4%
Starting a campfire 71.4%
Hand car wash 69.6%
Archery 66.7%
BMX 66.7%
Blow-drying hair 66.7%
Mowing the lawn 66.7%
Playing drums 66.7%
Playing lacrosse 66.7%
Snatch 66.7%
Tango 66.7%
Arm wrestling 65.0%
Croquet 63.6%
Laying tile 58.3%
Putting on makeup 57.1%
Skateboarding 55.6%
Kneeling 54.5%
Fixing the roof 50.0%
Playing kickball 50.0%
Putting on shoes 50.0%
Removing ice from car 50.0%
Throwing darts 50.0%
Volleyball 50.0%
Fun sliding down 40.0%
Painting 0.0%
https://omlx.ai/benchmarks/intelligence/amdzfbse
This model was quantized using oQ (oMLX v0.7.3) mixed-precision quantization.
The effective average is 7.172 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 6-bit, group 64 | 8.250B | 85.464% | 6.243 GiB | 6.50 |
| Affine 8-bit, group 64 | 0.913B | 9.453% | 0.903 GiB | 8.50 |
| BF16 | 0.491B | 5.083% | 0.914 GiB | 16.00 |
| Total | 9.653B | 100% | 8.060 GiB | 7.172 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 8.954B | 7.499 GB / 6.984 GiB |
| Vision | 0.456B | 0.912 GB / 0.849 GiB |
| MTP head | 0.243B | 0.243 GB / 0.226 GiB |
| Total | 9.653B | 8.654 GB / 8.060 GiB |
- Downloads last month
- 658
6-bit
