Instructions to use intel-ai/Muse-Spark-Open-816B-A42B-INT4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use intel-ai/Muse-Spark-Open-816B-A42B-INT4 with Transformers:

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

pipe = pipeline("text-generation", model="intel-ai/Muse-Spark-Open-816B-A42B-INT4")
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("intel-ai/Muse-Spark-Open-816B-A42B-INT4")
model = AutoModelForMultimodalLM.from_pretrained("intel-ai/Muse-Spark-Open-816B-A42B-INT4", 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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use intel-ai/Muse-Spark-Open-816B-A42B-INT4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "intel-ai/Muse-Spark-Open-816B-A42B-INT4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "intel-ai/Muse-Spark-Open-816B-A42B-INT4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/intel-ai/Muse-Spark-Open-816B-A42B-INT4

SGLang

How to use intel-ai/Muse-Spark-Open-816B-A42B-INT4 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 "intel-ai/Muse-Spark-Open-816B-A42B-INT4" \
    --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": "intel-ai/Muse-Spark-Open-816B-A42B-INT4",
		"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 "intel-ai/Muse-Spark-Open-816B-A42B-INT4" \
        --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": "intel-ai/Muse-Spark-Open-816B-A42B-INT4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use intel-ai/Muse-Spark-Open-816B-A42B-INT4 with Docker Model Runner:
```
docker model run hf.co/intel-ai/Muse-Spark-Open-816B-A42B-INT4
```

Model Details

This model is a W4A16 (int4, group_size 128, symmetric) model of meta-models/Muse-Spark-Open-816B-A42B generated by intel/auto-round using RTN. Please follow the license of the original model.

Quantized (W4A16): routed experts
Kept in BF16: attention, router, pooler, unpooler, embeddings, lm_head, vision tower, audio tower

How to Use

Environment

pip install vllm==0.29.0

vLLM Usage

vllm serve intel-ai/Muse-Spark-Open-816B-A42B-INT4 \
    --reasoning-parser musespark1x

Evaluation

Benchmark	Base Model	W4A16	Recovery
GSM8K	92.87	96.36	103.76%
MMLU	87.66	89.38	101.96%
PIQA	71.10	84.77	119.23%
HellaSwag	84.22	72.77	86.40%
Average	83.96	85.82	102.22%

Generate the Model

auto-round \
    --model_name meta-models/Muse-Spark-Open-816B-A42B \
    --scheme W4A16 \
    --model_free \
    --disable_torch_compile \
    --ignore_layers "self_attn,pre_expert_proj,post_expert_proj,mlp.gate,vision_tower,vision_adapter,vision_projection,audio_tower,embed_tokens,lm_head" \
    --format auto_round \
    --output_dir Muse-Spark-Open-816B-A42B-INT4

Ethical Considerations and Limitations

The model can produce factually incorrect output, and should not be relied on to produce factually accurate information. Because of the limitations of the pretrained model and the finetuning datasets, it is possible that this model could generate lewd, biased or otherwise offensive outputs. Therefore, before deploying any applications of the model, developers should perform safety testing.

Caveats and Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Here are a couple of useful links to learn more about Intel's AI software:

Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please consult an attorney before using this model for commercial purposes.

Cite

@article{cheng2023optimize,
  title={Optimize weight rounding via signed gradient descent for the quantization of llms},
  author={Cheng, Wenhua and Zhang, Weiwei and Shen, Haihao and Cai, Yiyang and He, Xin and Lv, Kaokao and Liu, Yi},
  journal={arXiv preprint arXiv:2309.05516},
  year={2023}
}

arxiv github

Signed-off-by: Lv, Kaokao kaokao.lv@intel.com

Downloads last month: 306

Safetensors

Model size

152B params

Tensor type

I32

BF16

F16

F32

Model tree for intel-ai/Muse-Spark-Open-816B-A42B-INT4

Base model

meta-models/Muse-Spark-Open-816B-A42B

Quantized

(24)

this model

Paper for intel-ai/Muse-Spark-Open-816B-A42B-INT4

Optimize Weight Rounding via Signed Gradient Descent for the Quantization of LLMs

Paper • 2309.05516 • Published Sep 11, 2023 • 16