Instructions to use rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4 with Transformers:

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

pipe = pipeline("text-generation", model="rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4")
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("rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4")
model = AutoModelForMultimodalLM.from_pretrained("rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4", 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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4

SGLang

How to use rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4 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 "rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4" \
    --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": "rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4" \
        --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": "rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4 with Docker Model Runner:
```
docker model run hf.co/rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4
```

edit: with vllm, use --language-model-only , have not figured this one out yet.

Qwen3.5-397B-A17B — REAP 28% Pruned, NVFP4

A personal experiment in aggressive MoE pruning. The goal: fit Qwen3.5-397B on 2× 96GB Blackwell GPUs with usable KV cache (~90K tokens), without losing quality.

What this is

28% of experts removed using REAP (Routing-Expert Activation Pruning) with a saliency × activation-count ordering, then quantized to NVFP4 using llm-compressor. Final size: ~164GB.

Pruning is heterogeneous — each layer retains a different number of experts based on global importance ranking. Early layers (which carry more redundancy) are pruned more aggressively; late layers are barely touched. This maximizes quality retention for a given size budget.

Benchmark results (non-thinking, lm-eval-harness)

Benchmark	This model	Nvidia NVFP4 (full, ~240GB)
IFEval	92.55 (avg)	91.20
MMLU Redux (generative)	90.94	91.24
GSM8K CoT (Llama)	96.74	96.80

28% fewer experts, 30%+ smaller on disk, and benchmark scores within noise of the full model.

Requirements

vLLM ≥ 0.16.1 (nightly dev builds from the cu130 index work)
Transformers ≥ 5.3

vLLM patches required

This model uses variable expert counts per layer (not a fixed number), which stock vLLM doesn't support yet. Two files need patching — see the patches/ directory for detailed instructions:

qwen3_next.py — read per-layer expert count instead of assuming a single global value
qwen3_5.py — infer expert count from tensor shape during weight loading

Patches were tested on vLLM 0.16.1rc1.dev188 (nightly cu130).

Details

Base model: Qwen3.5-397B-A17B
Pruning: REAP with 1,188 curated calibration samples, saliency × count global ordering
Quantization: NVFP4 (FP4 E2M1, group size 16, duo scaling)
Target hardware: 2× RTX PRO 6000

Downloads last month: 772

Model tree for rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4

Base model

Qwen/Qwen3.5-397B-A17B

Quantized

(69)

this model

Paper for rene98c/Qwen3.5-397B-A17B-REAP-28-NVFP4

REAP the Experts: Why Pruning Prevails for One-Shot MoE compression

Paper • 2510.13999 • Published Oct 15, 2025 • 20