---
license: apache-2.0
pipeline_tag: image-text-to-text
library_name: transformers
base_model: Qwen/Qwen3-VL-8B-Instruct
tags:
- vision-language
- image-text-to-text
- qwen3-vl
- sdvc
---

# SDVC_perception_VL-2B

This repository packages a vision-language checkpoint for SDVC perception experiments.

The checkpoint files and architecture configuration are intentionally kept compatible with the original Qwen3-VL release. Only the repository-facing model card and loading examples have been changed for this packaging task.

## Usage

```python
from transformers import AutoProcessor, Qwen3VLForConditionalGeneration

repo_id = "YeMoKoo/SDVC_preception_VL-2B"

model = Qwen3VLForConditionalGeneration.from_pretrained(
    repo_id,
    dtype="auto",
    device_map="auto",
)
processor = AutoProcessor.from_pretrained(repo_id)
```

## Example

```python
messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
            },
            {"type": "text", "text": "Describe this image."},
        ],
    }
]

inputs = processor.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device)

generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
    out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
    generated_ids_trimmed,
    skip_special_tokens=True,
    clean_up_tokenization_spaces=False,
)
print(output_text)
```

## Notes

- Target Hub repository: `YeMoKoo/SDVC_preception_VL-2B`
- Display name used in this card: `SDVC_perception_VL-2B`
- The safetensors checkpoint files are unchanged from the local source checkpoint.
- Structural config values are unchanged so the model remains load-compatible with Transformers.

## Citation

```bibtex
@misc{qwen3technicalreport,
      title={Qwen3 Technical Report},
      author={Qwen Team},
      year={2025},
      eprint={2505.09388},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2505.09388},
}
```