Instructions to use slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit with Transformers:

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

pipe = pipeline("text-generation", model="slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit", 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("slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit", trust_remote_code=True)
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 slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit

SGLang

How to use slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit 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 "slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit" \
    --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": "slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit",
		"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 "slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit" \
        --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": "slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit with Docker Model Runner:
```
docker model run hf.co/slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit
```

⚠️ Note:

These model weights are for personal testing purposes only. The goal is to find a quantization method that achieves high compression while preserving as much of the model's original performance as possible. The current compression scheme may not be optimal, so please use these weights with caution.

Creation

This model was created by applying the bitsandbytes and transformers as presented in the code snipet below.

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig

model_name = "moonshotai/Moonlight-16B-A3B-Instruct"
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",  # NF4 for weight
    bnb_4bit_use_double_quant=True,
    bnb_4bit_compute_dtype=torch.bfloat16  # bnb supports bfloat16
)

bnb_model = AutoModelForCausalLM.from_pretrained(
    model_name,
    trust_remote_code=True,
    torch_dtype=torch.bfloat16,
    quantization_config=bnb_config
)

tokenizer = AutoTokenizer.from_pretrained( model_name, trust_remote_code=True)

bnb_model.push_to_hub("slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit/Moonlight-16B-A3B-Instruct-bnb-4bit")
tokenizer.push_to_hub("slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit/Moonlight-16B-A3B-Instruct-bnb-4bit")

Sources

https://huggingface.co/moonshotai/Moonlight-16B-A3B-Instruct

Downloads last month: 9

Model tree for slowfastai/Moonlight-16B-A3B-Instruct-bnb-4bit

Base model

moonshotai/Moonlight-16B-A3B-Instruct

Quantized

(7)

this model