Instructions to use 0xSero/Qwen3.6-28B-REAP with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 0xSero/Qwen3.6-28B-REAP with Transformers:

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

pipe = pipeline("text-generation", model="0xSero/Qwen3.6-28B-REAP")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("0xSero/Qwen3.6-28B-REAP")
model = AutoModelForCausalLM.from_pretrained("0xSero/Qwen3.6-28B-REAP")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use 0xSero/Qwen3.6-28B-REAP with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "0xSero/Qwen3.6-28B-REAP"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0xSero/Qwen3.6-28B-REAP",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/0xSero/Qwen3.6-28B-REAP

SGLang

How to use 0xSero/Qwen3.6-28B-REAP 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 "0xSero/Qwen3.6-28B-REAP" \
    --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": "0xSero/Qwen3.6-28B-REAP",
		"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 "0xSero/Qwen3.6-28B-REAP" \
        --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": "0xSero/Qwen3.6-28B-REAP",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use 0xSero/Qwen3.6-28B-REAP with Docker Model Runner:
```
docker model run hf.co/0xSero/Qwen3.6-28B-REAP
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Please support my work: https://donate.sybilsolutions.ai

Qwen3.6-28B-REAP20-Opus-A3B

A 20%-expert-pruned + Opus-trace fine-tuned variant of Qwen/Qwen3.6-35B-A3B, produced via Cerebras REAP (Router-weighted Expert Activation Pruning, arXiv:2510.13999) followed by LoRA SFT on public Claude Opus reasoning traces.

Headline numbers

Metric	Base Qwen3.6-35B-A3B	This model (20% REAP + Opus SFT)	Δ
MMLU (200-sample lite)	`{{MMLU_BASE}}`	`{{MMLU_THIS}}`	`{{MMLU_DELTA}}`
GSM8K (100-sample lite)	`{{GSM_BASE}}`	`{{GSM_THIS}}`	`{{GSM_DELTA}}`
HumanEval (50 parse-rate)	`{{HE_BASE}}`	`{{HE_THIS}}`	`{{HE_DELTA}}`
Structured JSON parse (20)	`{{JSON_BASE}}`	`{{JSON_THIS}}`	`{{JSON_DELTA}}`
Mermaid render (10)	`{{MERM_BASE}}`	`{{MERM_THIS}}`	`{{MERM_DELTA}}`
AdvBench refusal (32)	`{{REFUSE_BASE}}`	`{{REFUSE_THIS}}`	`{{REFUSE_DELTA}}`

Architecture

Base: Qwen3.6-35B-A3B (40 layers, 256 experts/layer, 8 routed + 1 shared active, qwen3_5_moe)
After 20% REAP: 205 experts/layer kept, 51 experts/layer pruned → ~~28B total params, still **~~3B active**
Fine-tune: LoRA rank 32, α 64 on q,k,v,o,gate,up,down projections. bf16 weights after merge.

Pipeline

Calibration merge — 5,000 stratified samples from:
- /Users/sero/.../reap-expert-swap/dataset/calibration-20k.jsonl (general, coding, reasoning, etc.)
- 0xSero/structured-outputs-calibration-v1 (JSON / Mermaid / schema)
REAP observation (this fork's Qwen3_5Moe-aware observer, multi-GPU layerwise on 8× A100-40GB): {{OBS_DURATION}}
REAP prune @ 20% using reap saliency metric, renormalized router weights, seed 42.
Opus-trace SFT via LLaMA-Factory + DeepSpeed ZeRO-3 (8× A100). LoRA 2 epochs on nohurry/Opus-4.6-Reasoning-3000x-filtered (2,326 reasoning trajectories with explicit <think>…</think>\nanswer structure).
GGUF — bf16, Q8_0, Q6_K, Q5_K_M, Q4_K_M with imatrix from merged calibration.

Sidecar observations

REAP observation artifacts live in the separate dataset repo 0xSero/qwen3.6-35b-a3b-reap-observations.

Known limitations

Refusal behavior follows the base model plus Opus SFT; no explicit abliteration was applied in this release. The model will refuse straight adversarial probes at roughly base-model rates.
Reasoning quality on GSM8K-style problems depends on the <think> chain-of-thought; short max-tokens limits hurt accuracy.
Structured-output calibration is oversampled vs. base mix (JSON/Mermaid experts preferentially retained).

License

Apache 2.0, inherited from base model. This checkpoint is a derivative work; please preserve attribution.

Citation

@misc{lasby2025reap,
  title   = {REAP: Router-weighted Expert Activation Pruning for Mixture-of-Experts},
  author  = {Lasby, Mike and others},
  year    = {2025},
  eprint  = {2510.13999},
  archivePrefix = {arXiv},
}