Instructions to use FlagRelease/Moonlight-16B-A3B-ascend-FlagOS with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FlagRelease/Moonlight-16B-A3B-ascend-FlagOS with Transformers:

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

pipe = pipeline("text-generation", model="FlagRelease/Moonlight-16B-A3B-ascend-FlagOS", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("FlagRelease/Moonlight-16B-A3B-ascend-FlagOS", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("FlagRelease/Moonlight-16B-A3B-ascend-FlagOS", trust_remote_code=True, 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use FlagRelease/Moonlight-16B-A3B-ascend-FlagOS with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FlagRelease/Moonlight-16B-A3B-ascend-FlagOS"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FlagRelease/Moonlight-16B-A3B-ascend-FlagOS",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/FlagRelease/Moonlight-16B-A3B-ascend-FlagOS

SGLang

How to use FlagRelease/Moonlight-16B-A3B-ascend-FlagOS 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 "FlagRelease/Moonlight-16B-A3B-ascend-FlagOS" \
    --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": "FlagRelease/Moonlight-16B-A3B-ascend-FlagOS",
		"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 "FlagRelease/Moonlight-16B-A3B-ascend-FlagOS" \
        --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": "FlagRelease/Moonlight-16B-A3B-ascend-FlagOS",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use FlagRelease/Moonlight-16B-A3B-ascend-FlagOS with Docker Model Runner:
```
docker model run hf.co/FlagRelease/Moonlight-16B-A3B-ascend-FlagOS
```

Introduction

Moonlight-16B-A3B is a high-performance large language model optimized for the ascend GPU platform. Built on a Mixture of Experts (MoE) architecture, the model features a total of 16 billion parameters with approximately 3 billion active parameters per inference, striking an optimal balance between high performance and efficient inference throughput. Deeply optimized for ascend GPU hardware, Moonlight-16B-A3B supports the vLLM inference framework, making it well-suited for large-scale deployment scenarios.

Integrated Deployment

Out-of-the-box inference scripts with pre-configured hardware and software parameters
Released FlagOS-ascend container image supporting deployment within minutes

Consistency Validation

Rigorously evaluated through benchmark testing: Performance and results from the FlagOS software stack are compared against native stacks on multiple public benchmarks.

Evaluation Results

Benchmark Result

Metrics	Moonlight-16B-A3B-Nvidia-Origin	Moonlight-16B-A3B-ascend-FlagOS
GPQA_Diamond	0.1384	0.1091
LiveBench New	0.0475	0.0382
musr	0.0172	0.0820
mmlu_pro	0.1986	0.2657
aime	0.0000	0.0000

User Guide

Environment Setup

Item	Version
Docker Version	Docker version 20.10.8, build 3967b7d
Operating System	openEuler 22.03 LTS-SP4 Kernel: 5.10.0 Arch: aarch64

Operation Steps

Download FlagOS Image

docker pull harbor.baai.ac.cn/external-cooperation/moonlight-16b-a3b-ascend-tree_0.5.0_ascend3.2-gems_5.0.2-vllm_0.13.0_empty-plugin_0.1.1-cx_none-python_3.11.14-torch_2.8.0-pcp_none-npu_ascend910c-arc_aarch64-driver_25.5.0:2607301800

Download Open-source Model Weights

pip install modelscope

modelscope download \
  --model FlagRelease/Moonlight-16B-A3B-ascend-FlagOS \
  --local_dir /data/Moonlight-16B-A3B-ascend-FlagOS

Start the Container

docker run -itd \
  --name flagos \
  --privileged \
  --ipc=host \
  --network host \
  --shm-size=32g \
  -v /usr/local/Ascend/driver:/usr/local/Ascend/driver \
  -v /usr/local/sbin/npu-smi:/usr/local/sbin/npu-smi \
  -v /usr/local/dcmi:/usr/local/dcmi \
  -v /usr/local/sbin:/usr/local/sbin \
  -v /etc/ascend_install.info:/etc/ascend_install.info \
  -v /data/Moonlight-16B-A3B-ascend-FlagOS:/data/Moonlight-16B-A3B-ascend-FlagOS \
  --workdir /workspace \
harbor.baai.ac.cn/external-cooperation/moonlight-16b-a3b-ascend-tree_0.5.0_ascend3.2-gems_5.0.2-vllm_0.13.0_empty-plugin_0.1.1-cx_none-python_3.11.14-torch_2.8.0-pcp_none-npu_ascend910c-arc_aarch64-driver_25.5.0:2607301800

Enter the Container

docker exec -it flagos /bin/bash

Start the Server

Inside the container:

export ASCEND_RT_VISIBLE_DEVICES=0,1
export VLLM_FL_FLAGOS_WHITELIST=cos,sin,lt_scalar,lt,arange,repeat_interleave_self_int,all,embedding
export VLLM_PLUGINS=fl
export USE_FLAGGEMS=1
nohup vllm serve \
    --model /data/Moonlight-16B-A3B-ascend-FlagOS \
    --served-model-name moonlight-16b-a3b-flagos \
    --port 8003 \
    --trust-remote-code \
    --tensor-parallel-size 2 \
    --enforce-eager \
    > /workspace/flagos_server.log 2>&1 &

Service Invocation

Invocation Script

curl http://localhost:8003/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "moonlight-16b-a3b-flagos",
    "messages": [{"role": "user", "content": "hello!"}]
  }'

AnythingLLM Integration Guide

1. Download & Install

Visit the official site: https://anythingllm.com/
Choose the appropriate version for your OS (Windows/macOS/Linux)
Follow the installation wizard to complete the setup

2. Configuration

Launch AnythingLLM
Open settings (bottom left, fourth tab)
Configure core LLM parameters:
- API URL: http://localhost:8003/v1
- Model: moonlight-16b-a3b-flagos
Click "Save Settings" to apply changes

3. Model Interaction

After model loading is complete:
Click "New Conversation"
Enter your question (e.g., "Explain the basics of quantum computing")
Click the send button to get a response

Technical Overview

FlagOS is a fully open-source system software stack designed to unify the "model–system–chip" layers and foster an open, collaborative ecosystem. It enables a "develop once, run anywhere" workflow across diverse AI accelerators, unlocking hardware performance, eliminating fragmentation among vendor-specific software stacks, and substantially lowering the cost of porting and maintaining AI workloads.

With core technologies such as FlagScale, together with vllm-plugin-fl, distributed training/inference framework, FlagGems universal operator library, FlagCX communication library, and FlagTree unified compiler, the FlagRelease platform leverages the FlagOS stack to automatically produce and release various combinations of <chip + open-source model>.

This enables efficient and automated model migration across diverse chips, opening a new chapter for large model deployment and application.

FlagGems

FlagGems is a high-performance, generic operator library implemented in Triton language. It is built on a collection of backend-neutral kernels that aims to accelerate LLM training and inference across diverse hardware platforms.

FlagTree

FlagTree is an open-source unified compiler for multiple AI chips. It provides unified compilation capabilities across multiple backends and rapidly implements single-repository multi-backend support.

FlagScale and vllm-plugin-fl

FlagScale is a comprehensive toolkit designed to support the entire lifecycle of large models. It integrates capabilities from Megatron-LM and vLLM to provide an end-to-end solution for training and inference.

vllm-plugin-fl is a vLLM plugin built on the FlagOS unified multi-chip backend.

FlagCX

FlagCX is a scalable and adaptive cross-chip communication library for distributed AI workloads.

FlagEval Evaluation Framework

FlagEval is a comprehensive evaluation system and open platform for large models. It supports large-scale benchmark evaluation across NLP, CV, Audio, and Multimodal tasks.

Contributing

We warmly welcome global developers to join us:

Submit Issues to report problems
Create Pull Requests to contribute code
Improve technical documentation
Expand hardware adaptation support

License

The model weights are derived from moonshotai/Moonlight-16B-A3B and are open-sourced under the Apache License 2.0: https://www.apache.org/licenses/LICENSE-2.0.txt

Downloads last month: 1

Safetensors

Model size

16B params

Tensor type

BF16