---
title: DefakerOne2
canonical_url: "https://www.modelscope.cn/models/DanceText/DefakerOne2"
md_url: "https://www.modelscope.cn/models/DanceText/DefakerOne2.md"
repository: DanceText/DefakerOne2
last_updated: 2026-10-29
parameters: 4.4B
tensor_type:
  - BF16
library_name:
  - safetensors
downloads: 28
stars: 0
---

# DefakerOne2

> DefakerOne2 - DanceText 在 ModelScope 开源的模型。Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage

- **Repository**: DanceText/DefakerOne2
- **Parameters**: 4.4B
- **Downloads**: 28
- **Stars**: 0
- **Last updated**: 2026-10-29

Source: https://www.modelscope.cn/models/DanceText/DefakerOne2

---

<h2 align="center">
  Venus-DefakerOne2: A Vision-Forensics Foundation Model<br>
  with Broader Coverage
</h2>

<p align="center">
  <b>GuangJian Team, Ant Group</b><br>
  <b>Technical Committee on Digital Media Forensics and Security, CSIG</b>
</p>

<p align="center">
  <a href="https://github.com/inclusionAI/DefakerOne2"><img src="https://img.shields.io/badge/GitHub-Code-181717?logo=github" alt="GitHub"></a>
  <a href="https://www.modelscope.cn/models/DanceText/DefakerOne2"><img src="https://img.shields.io/badge/ModelScope-DefakerOne2-624AFF" alt="ModelScope Model"></a>
  <a href="https://www.modelscope.cn/models/DanceText/RRVBench"><img src="https://img.shields.io/badge/ModelScope-RRVBench-624AFF" alt="ModelScope RRVBench"></a>
  <a href="https://venus-guangjian.github.io/Venus-DeFakerOne2/"><img src="https://img.shields.io/badge/Project-Page-success" alt="Project Page"></a>
</p>

### 1. Environment Requirements

Python **3.12.3**

```bash
pip install torch==2.9.0 torchvision==0.24.0 ms-swift=4.1.0 numpy==1.26.0 scipy==1.11.3 scikit-learn==1.3.2 torchao==0.16.0
```

### 2. Model Download

```bash
modelscope download --model DanceText/DefakerOne2 --local_dir [target_save_path]
```

Two checkpoints are provided under the download directory:

| Directory | Description |
| --- | --- |
| `preview/` | **DefakerOne2-preview** — can be used for reproducing the scores reported in the paper. |
| `release/` | **DefakerOne2-release** — trained with more data; expected to perform better in most cases. |

### 3. Inference Data Format

Inference inputs are **JSONL** files, one sample per line.

**Image**

```json
{"messages": [{"role": "user", "content": "<image>Is this image real or fake? Answer only Real or Fake"}], "images": ["your_image_path.jpg"]}
```

**Video**

```json
{"messages": [{"role": "user", "content": "<video>Is this video real or fake? Answer only Real or Fake"}], "images": ["your_video_path.mp4"]}
```

The prompt is fixed and direct, so the model commits to its conclusion in a **single output token** (`Real` / `Fake`).

### 4. Inference Command

```bash
CUDA_VISIBLE_DEVICES=0 MIN_PIXELS=262144 MAX_PIXELS=2408448 --model [model_weight_dir] --val_dataset [val_dataset_jsonl_path] --result_path [prediction_jsonl_save_path] --max_new_token 16 --enable_thinking False --batch_size 1 --num_workers 8 --logprobs True --top_logprobs 20
```
```
Git下载
```
#Git模型下载
git clone https://www.modelscope.cn/DanceText/DefakerOne2.git
```

<p style="color: lightgrey;">如果您是本模型的贡献者，我们邀请您根据<a href="https://modelscope.cn/docs/ModelScope%E6%A8%A1%E5%9E%8B%E6%8E%A5%E5%85%A5%E6%B5%81%E7%A8%8B%E6%A6%82%E8%A7%88" style="color: lightgrey; text-decoration: underline;">模型贡献文档</a>，及时完善模型卡片内容。</p>