File size: 4,363 Bytes
93fca56 999ab47 93fca56 5cc4626 93fca56 efe9263 93fca56 999ab47 26a49b4 93fca56 999ab47 93fca56 999ab47 93fca56 5cc4626 93fca56 999ab47 1108a17 93fca56 999ab47 93fca56 999ab47 93fca56 5cc4626 999ab47 26a49b4 93fca56 5cc4626 efe9263 93fca56 5cc4626 93fca56 5cc4626 93fca56 5cc4626 93fca56 5cc4626 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 |
---
license: mit
library_name: transformers
pipeline_tag: image-text-to-text
---
# Skywork-R1V3-38B-AWQ
<div align="center">
<img src="skywork-logo.png" alt="Introduction Image" width="500" height="400">
</div>
## 📖 [R1V3 Report](https://arxiv.org/abs/2507.06167) | 💻 [GitHub](https://github.com/SkyworkAI/Skywork-R1V) | 🌐 [ModelScope](https://modelscope.cn/models/Skywork/Skywork-R1V3-38B)
<div align="center">
[](https://github.com/SkyworkAI/Skywork-R1V/stargazers)[](https://github.com/SkyworkAI/Skywork-R1V/fork)
</div>
## Evaluation
<div align="center">
<b>Comprehensive performance comparison across text and multimodal reasoning benchmarks.</b>
</div>
<table align="center" border="1" style="border-collapse: collapse; width: 100%;">
<thead>
<tr>
<th>Model</th>
<th align="center">MMMU</th>
<th align="center">MathVista</th>
</tr>
</thead>
<tbody>
<tr>
<td colspan="3" align="center"><i
>Proprietary Models</i></td>
</tr>
<tr>
<td>Claude-3.7-Sonnet</td>
<td align="center">75.0</td>
<td
align="center">66.8</td>
</tr>
<tr>
<td>OpenAI-4o</td>
<td align="center">70.7</td>
<td align="center">62.9</td>
</tr>
<tr>
<td colspan="3" align="center"><i>Open-Source Models</i></td>
</tr>
<tr>
<td>InternVL3-78B</td>
<td align="center">72.2</td>
<td align="center">72.2</td>
</tr>
<tr>
<td>Qwen2.5-VL-72B</td>
<td align="center">70.3</td>
<td align="center">74.8</td>
</tr>
<tr>
<td>QvQ-Preview-72B</td>
<td align="center">70.3</td>
<td align="center">71.4</td>
</tr>
<tr>
<td>Skywork-R1V3</td>
<td align="center"><b>76.0</b></td>
<td align="center"><b>77.1</b></td>
</tr>
<tr>
<td>Skywork-R1V3-AWQ</td>
<td align="center">66.7</td>
<td align="center">70.5</td>
</tr>
</tbody>
</table>
## Usage
You can use the quantized model with different inference frameworks:
### Using VLLM
#### Python API
```python
import os
from vllm import LLM, SamplingParams
from vllm.entrypoints.chat_utils import load_chat_template
model_name = "Skywork/Skywork-R1V3-38B-AWQ" # or local path
llm = LLM(model_name,
dtype='float16',
quantization="awq",
gpu_memory_utilization=0.9,
max_model_len=4096,
trust_remote_code=True,
)
# Add your inference code here
```
#### OpenAI-compatible API Server
```bash
MODEL_ID="Skywork/Skywork-R1V3-38B-AWQ" # or local path
CUDA_VISIBLE_DEVICES=0 \
python -m vllm.entrypoints.openai.api_server \
--model $MODEL_ID \
--dtype float16 \
--quantization awq \
--port 23334 \
--max-model-len 12000 \
--gpu-memory-utilization 0.9 \
--trust-remote-code
```
### Using LMDeploy
```python
import os
from lmdeploy import pipeline, TurbomindEngineConfig, ChatTemplateConfig
from lmdeploy.vl import load_image
model_path = "Skywork/Skywork-R1V3-38B-AWQ" # or local path
engine_config = TurbomindEngineConfig(cache_max_entry_count=0.75)
chat_template_config = ChatTemplateConfig(model_name=model_path)
pipe = pipeline(model_path,
backend_config=engine_config,
chat_template_config=chat_template_config,
)
# Example: Multimodal inference
image = load_image('table.jpg')
response = pipe(('Describe this image?', image))
print(response.text)
```
## Hardware Requirements
The AWQ quantization reduces the memory footprint compared to the original FP16 model. We recommend:
- At least one GPU with 30GB+ VRAM for inference
- For optimal performance with longer contexts, 40GB+ VRAM is recommended
## Citation
If you use this model in your research, please cite:
```bibtex
@misc{shen2025skyworkr1v3technicalreport,
title={Skywork-R1V3 Technical Report},
author={Wei Shen and Jiangbo Pei and Yi Peng and Xuchen Song and Yang Liu and Jian Peng and Haofeng Sun and Yunzhuo Hao and Peiyu Wang and Jianhao Zhang and Yahui Zhou},
year={2025},
eprint={2507.06167},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2507.06167},
}
```
|