Instructions to use kueizen/Marco-Mini-Instruct-REAP70 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kueizen/Marco-Mini-Instruct-REAP70 with Transformers:

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

pipe = pipeline("text-generation", model="kueizen/Marco-Mini-Instruct-REAP70")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("kueizen/Marco-Mini-Instruct-REAP70")
model = AutoModelForCausalLM.from_pretrained("kueizen/Marco-Mini-Instruct-REAP70", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use kueizen/Marco-Mini-Instruct-REAP70 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kueizen/Marco-Mini-Instruct-REAP70"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kueizen/Marco-Mini-Instruct-REAP70",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/kueizen/Marco-Mini-Instruct-REAP70

SGLang

How to use kueizen/Marco-Mini-Instruct-REAP70 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 "kueizen/Marco-Mini-Instruct-REAP70" \
    --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": "kueizen/Marco-Mini-Instruct-REAP70",
		"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 "kueizen/Marco-Mini-Instruct-REAP70" \
        --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": "kueizen/Marco-Mini-Instruct-REAP70",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use kueizen/Marco-Mini-Instruct-REAP70 with Docker Model Runner:
```
docker model run hf.co/kueizen/Marco-Mini-Instruct-REAP70
```

Marco-Mini-Instruct-REAP70

Marco-Mini-Instruct (17.3B total, ~0.86B active, 256 experts, 8 active per token) with 70% of its experts removed by REAP: 77 of 256 experts kept. These are the full-precision (bf16) safetensors weights, for fine-tuning, re-quantising or running with transformers.

For ready-to-run quantised files at every pruning ratio, see kueizen/Marco-Mini-Instruct-REAP-GGUF.

Load it

from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "kueizen/Marco-Mini-Instruct-REAP70"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, torch_dtype="auto", device_map="auto")

How this was made

Pruning: REAP (Router-weighted Expert Activation Pruning, Cerebras Research), using the reference implementation. REAP scores each expert by its router weight times the size of its output over a calibration set, then removes the lowest-scoring experts whole. It is one-shot: no retraining. Seed 42.
Calibration set: theblackcat102/evol-codealpaca-v1 (train split, shuffled, seed 42), 64 samples per category, batch size 1, max sequence length 2048 tokens.
Experts removed: 70% (77 of 256 kept).

We have not run task benchmarks on these checkpoints (MMLU, coding, multilingual). Test them on your own workload before relying on them.

License and credits

Derived from ATH-MaaS/Marco-Mini-Instruct and released under the same Apache 2.0 licence. What we changed: removed experts with REAP. Nothing else was modified or retrained. REAP is by Cerebras Research: paper, code. All credit for the base model goes to its authors.

Published by Kueizen.

Downloads last month: 196

Safetensors

Model size

5B params

Tensor type

BF16

Model tree for kueizen/Marco-Mini-Instruct-REAP70

Base model

ATH-MaaS/Marco-Mini-Instruct

Finetuned

(7)

this model

Paper for kueizen/Marco-Mini-Instruct-REAP70

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

Paper • 2510.13999 • Published Oct 15, 2025 • 23