Qwen2.5-VL-72B-Instruct-quantized-FP8-Dynamic

Model Overview

• Model Architecture: Qwen2.5-VL-72B-Instruct
  • Input: Vision-Text
  • Output: Text
• Model Optimizations:
  • Weight quantization: FP8
  • Activation quantization: FP8
• Release Date: 2/24/2025
• Version: 1.0
• Model Developers: Neural Magic

Quantized version of Qwen/Qwen2.5-VL-72B-Instruct.

Model Optimizations

This model was obtained by quantizing the weights of Qwen/Qwen2.5-VL-72B-Instruct to FP8 data type, ready for inference with vLLM >= 0.5.2.

Deployment

Use with vLLM

This model can be deployed efficiently using the vLLM backend, as shown in the example below.

from vllm.assets.image import ImageAsset
from vllm import LLM, SamplingParams

# prepare model
llm = LLM(
 model="neuralmagic/Qwen2.5-VL-72B-Instruct-FP8-Dynamic",
 trust_remote_code=True,
 max_model_len=4096,
 max_num_seqs=2,
)

# prepare inputs
question = "What is the content of this image?"
inputs = {
 "prompt": f"<|user|>\n<|image_1|>\n{question}<|end|>\n<|assistant|>\n",
 "multi_modal_data": {
 "image": ImageAsset("cherry_blossom").pil_image.convert("RGB")
 },
}

# generate response
print("========== SAMPLE GENERATION ==============")
outputs = llm.generate(inputs, SamplingParams(temperature=0.2, max_tokens=64))
print(f"PROMPT : {outputs[0].prompt}")
print(f"RESPONSE: {outputs[0].outputs[0].text}")
print("==========================================")

vLLM also supports OpenAI-compatible serving. See the documentation for more details.

Creation

This model was created with llm-compressor by running the code snippet below as part a multimodal announcement blog.

Evaluation

The model was evaluated using mistral-evals for vision-related tasks and using lm_evaluation_harness for select text-based benchmarks. The evaluations were conducted using the following commands:

Accuracy

Category	Metric	Qwen/Qwen2.5-VL-72B-Instruct	neuralmagic/Qwen2.5-VL-72B-Instruct-FP8-Dynamic	Recovery (%)
Vision	MMMU (val, CoT) explicit_prompt_relaxed_correctness	64.33	66.88	103.96%
	VQAv2 (val) vqa_match	81.94	81.94	100.00%
	DocVQA (val) anls	94.71	94.64	99.93%
	ChartQA (test, CoT) anywhere_in_answer_relaxed_correctness	88.96	89.04	100.09%
	Mathvista (testmini, CoT) explicit_prompt_relaxed_correctness	78.18	77.78	99.49%
	Average Score	81.62	81.86	100.29%
Text	MGSM (CoT)	75.45	75.29	99.79%
	MMLU (5-shot)	86.16	86.12	99.95%

Inference Performance

This model achieves up to 1.79x speedup in single-stream deployment and up to 1.84x speedup in multi-stream asynchronous deployment, depending on hardware and use-case scenario. The following performance benchmarks were conducted with vLLM version 0.7.2, and GuideLLM.

Single-stream performance (measured with vLLM version 0.7.2)

				Document Visual Question Answering 1680W x 2240H 64/128		Visual Reasoning 640W x 480H 128/128		Image Captioning 480W x 360H 0/128
Hardware	Number of GPUs	Model	Average Cost Reduction	Latency (s)	Queries Per Dollar	Latency (s)th>	Queries Per Dollar	Latency (s)	Queries Per Dollar
A100	4	Qwen/Qwen2.5-VL-72B-Instruct	6.4	78	4.5	111	4.4	113
	2	neuralmagic/Qwen2.5-VL-72B-Instruct-quantized.w8a8	1.85	7.0	143	4.9	205	4.8	211
	1	neuralmagic/Qwen2.5-VL-72B-Instruct-quantized.w4a16	3.33	9.4	213	5.1	396	4.8	420
H100	4	Qwen/Qwen2.5-VL-72B-Instruct	4.3	68	3.0	97	2.9	100
	2	neuralmagic/Qwen2.5-VL-72B-Instruct-FP8-Dynamic	1.79	4.6	122	3.3	173	3.2	177
	1	neuralmagic/Qwen2.5-VL-72B-Instruct-quantized.w4a16	5.66	4.3	252	4.4	251	4.2	259

**Use case profiles: Image Size (WxH) / prompt tokens / generation tokens

**QPD: Queries per dollar, based on on-demand cost at Lambda Labs (observed on 2/18/2025).

Multi-stream asynchronous performance (measured with vLLM version 0.7.2)

			Document Visual Question Answering 1680W x 2240H 64/128		Visual Reasoning 640W x 480H 128/128		Image Captioning 480W x 360H 0/128
Hardware	Model	Average Cost Reduction	Maximum throughput (QPS)	Queries Per Dollar	Maximum throughput (QPS)	Queries Per Dollar	Maximum throughput (QPS)	Queries Per Dollar
A100x4	Qwen/Qwen2.5-VL-72B-Instruct	0.4	180	1.1	539	1.2	595
	neuralmagic/Qwen2.5-VL-72B-Instruct-quantized.w8a8	1.80	0.6	289	2.0	1020	2.3	1133
	neuralmagic/Qwen2.5-VL-72B-Instruct-quantized.w4a16	2.75	0.7	341	3.2	1588	4.1	2037
H100x4	Qwen/Qwen2.5-VL-72B-Instruct	0.5	134	1.2	357	1.3	379
	neuralmagic/Qwen2.5-VL-72B-Instruct-FP8-Dynamic	1.73	0.9	247	2.2	621	2.4	669
	neuralmagic/Qwen2.5-VL-72B-Instruct-quantized.w4a16	8.27	3.3	913	3.3	898	3.6	991

**Use case profiles: Image Size (WxH) / prompt tokens / generation tokens

**QPS: Queries per second.

**QPD: Queries per dollar, based on on-demand cost at Lambda Labs (observed on 2/18/2025).

Base model

Qwen/Qwen2.5-VL-72B-Instruct

(32)

Quantized variants of Qwen 2.5 Instruct and Qwen VL models • 10 items • Updated Apr 30 • 2