diff --git a/README.md b/README.md index 34ed3f9..b45d8bc 100644 --- a/README.md +++ b/README.md @@ -1,3 +1,9 @@ + +> [!NOTE] +> 本文档由 WeHub 基于上游 README 翻译整理，属于社区翻译，非官方中文文档。 +> [English](./README.en.md) · [原始项目](https://github.com/bytedance/Lance) · [上游 README](https://github.com/bytedance/Lance/blob/HEAD/README.md) +> 原作者、版权与许可证归属以原始项目及本仓库 LICENSE 文件为准。 +
Lance logo @@ -21,7 +27,7 @@
ByteDance
- * Equal contribution    ✉ Corresponding authors    § Project lead + * 共同一作    ✉ 通讯作者    § 项目负责人

Homepage @@ -29,45 +35,45 @@ Model Demo
- English | 简体中文 + English | 简体中文

-> **Note:** Lance is a research project rather than a polished product model. The released checkpoint was trained with up to 128 A100 GPUs, with training conducted up to 768x768 image generation and 480p, 12 FPS video generation. Our goal is to share a research artifact for studying unified image/video understanding, generation, and editing under a relatively small model and limited compute budget. Output quality may vary across prompts, resolutions, duration, motion complexity, and editing scenarios, and we see further opportunities to improve the post-training recipe. We appreciate constructive feedback from the community as we continue improving the project. +> **注意：** Lance 是一个研究项目，而不是经过充分产品化打磨的模型。当前开源 checkpoint 使用不超过 128 张 A100 GPU 训练，训练阶段覆盖到 768x768 图像生成和 480p、12 FPS 视频生成。我们希望将 Lance 作为一个研究参考，分享在较小模型规模和相对有限算力下统一图像/视频理解、生成和编辑的建模思路、训练流程和推理代码。模型效果可能会随 prompt、分辨率、时长、运动复杂度和编辑场景而波动，post-training recipe 仍有进一步改进空间。我们欢迎社区提供建设性反馈，帮助项目持续改进。 -## 🔥 Updates +## 🔥 更新 -- **`2026/06/17`**: 🛠️ Released the fine-tuning code for Lance. See the training guide in [TRAIN](train.md). -- **`2026/06/03`**: 🚀 Lance is now supported in [vLLM-Omni](https://github.com/vllm-project/vllm-omni). See the [recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/ByteDance/Lance.md)! -- **`2026/05/29`**: 💪 Added support for Image-to-Video generation. [More to see](assets/docs/changelog/2026-05-29.md)! -- **`2026/05/26`**: 🎨 The Gradio interface now supports image and video generation, editing, and understanding. [Try it out](assets/docs/changelog/2026-05-26.md)! -- **`2026/05/25`**: ✨ The [Hugging Face Space](https://huggingface.co/spaces/bytedance-research/Lance) is now live, thanks to the HF team! -- **`2026/05/19`**: 🤗 The technical report is now available on [arXiv](http://arxiv.org/abs/2605.18678). -- **`2026/05/18`**: 🔥 We launched the [project homepage](https://lance-project.github.io/) and released the initial inference code and model weights on [GitHub](https://github.com/bytedance/Lance/) and [Hugging Face](https://huggingface.co/bytedance-research/Lance). +- **`2026/06/17`**: 🛠️ 发布 Lance 微调代码。查看 [训练指南](train_zh.md) 了解细节。 +- **`2026/06/03`**: 🚀 Lance 现已被 [vLLM-Omni](https://github.com/vllm-project/vllm-omni) 支持。查看 [recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/ByteDance/Lance.md)！ +- **`2026/05/29`**: 💪 增加 Image-to-Video generation 支持。[查看更多](assets/docs/changelog/2026-05-29.md)！ +- **`2026/05/26`**: 🎨 Gradio 界面现已支持图像和视频生成、编辑与理解任务。[欢迎体验](assets/docs/changelog/2026-05-26.md)！ +- **`2026/05/25`**: ✨ [Hugging Face Space](https://huggingface.co/spaces/bytedance-research/Lance) 已上线，感谢 HF 团队的支持！ +- **`2026/05/19`**: 🤗 技术报告现已发布于 [arXiv](http://arxiv.org/abs/2605.18678)。 +- **`2026/05/18`**: 🔥 我们发布了 [项目主页](https://lance-project.github.io/)，并在 [GitHub](https://github.com/bytedance/Lance/) 和 [Hugging Face](https://huggingface.co/bytedance-research/Lance) 上开源了初版推理代码和模型权重。 -## 🌟 Highlights +## 🌟 亮点 -**Lance** is a 3B native unified multimodal model that supports **image and video understanding, generation, and editing** within a single framework. +**Lance** 是一个 3B 参数、原生统一的多模态模型，在单一框架下同时支持 **图像与视频的理解、生成和编辑**。 -- **Efficient at 3B scale.** With only **3B active parameters**, Lance achieves competitive performance across image generation, image editing, and video generation benchmarks. -- **Training from scratch.** Lance is trained from scratch with a staged multi-task recipe and within a budget of **up to 128 A100 GPUs**. +- **3B 规模高效。** 仅使用 **3B 激活参数**，Lance 即可在图像生成、图像编辑和视频生成等基准上取得有竞争力的表现。 +- **从零训练。** Lance 采用分阶段多任务训练配方从零训练，并在 **不超过 128 张 A100 GPU** 的预算内完成训练。 -We are actively updating and improving this repository. If you find any bugs or have suggestions, please feel free to open an issue or submit a pull request (PR) 💖. +我们正在持续更新和改进本仓库。如果你发现任何问题或有改进建议，欢迎提出 issue 或提交 pull request（PR）💖。
Lance benchmark overview across image generation, image editing, video generation, and video understanding
-## 🎨 Demo +## 🎨 演示
-Show demo results +展开查看演示结果
- 🔥 We recommend visiting our homepage for more visual results. 🔥 + 🔥 建议浏览我们的 主页 查看更多效果。🔥
-

Text-to-Video

+

文生视频

@@ -84,7 +90,7 @@ We are actively updating and improving this repository. If you find any bugs or
-

Video Editing

+

视频编辑

@@ -101,7 +107,7 @@ We are actively updating and improving this repository. If you find any bugs or
-

Multi-turn Consistency Editing

+

多轮一致性编辑

@@ -109,7 +115,7 @@ We are actively updating and improving this repository. If you find any bugs or
-

Intelligent Video Generation

+

智能视频生成

@@ -122,30 +128,31 @@ We are actively updating and improving this repository. If you find any bugs or -## 🚀 Installation +## 🚀 安装 -### Recommended Environment +### 推荐环境 -- **Software:** Python 3.10+, CUDA 12.4+ (required) -- **Hardware:** A GPU with at least 40GB VRAM is required for inference +- **软件环境：** Python 3.10+，CUDA 12.4+（必需） +- **硬件环境：** 推理至少需要一张显存不低于 40GB 的 GPU -We have tested the following dependency combinations on NVIDIA A100: +我们在 NVIDIA A100 上测试通过了以下依赖组合： - PyTorch 2.8.0 + cu126 + flash-attn 2.8.3 - PyTorch 2.5.1 + cu124 + flash-attn 2.6.3 -The default installation commands use the PyTorch 2.8.0 + cu126 setup. For other GPU models, please choose and validate a PyTorch build and a matching `flash-attn` version according to your driver, CUDA runtime, Python version, and GPU architecture. +默认安装命令使用 PyTorch 2.8.0 + cu126 环境。对于其他 GPU 型号，请根据驱动版本、CUDA runtime、Python 版本和 GPU 架构自行选择并验证匹配的 PyTorch 与 `flash-attn` 版本组合。 -### Installation Steps -First, clone the repository: +### 安装步骤 + +首先，克隆代码仓库： ```bash git clone https://github.com/bytedance/Lance.git cd Lance ``` -Then, set up the environment: +然后，配置环境： ```bash conda create -n Lance python=3.11 -y @@ -155,15 +162,14 @@ pip install -r requirements.txt pip install flash-attn==2.8.3 --no-build-isolation ``` -> **Note:** If installing `flash-attn` from source fails, you can install a prebuilt wheel instead. The wheelhouse below is from a third-party repository and is provided for **reference only**; please verify that any wheel you install matches your Python, PyTorch and CUDA versions. - +> **注意：** 如果从源码安装 `flash-attn` 失败，可以改为安装预编译 wheel。下面的 wheelhouse 来自第三方仓库，仅作为**参考提供**；请在安装前确认 wheel 与当前 Python、PyTorch 和 CUDA 版本匹配： +> > ```bash > pip install --no-cache-dir --no-deps --force-reinstall \ > "https://huggingface.co/strangertoolshf/flash_attention_2_wheelhouse/resolve/main/wheelhouse-flash_attn-2.8.3/linux_x86_64/torch2.8/cu12/abiTRUE/cp311/flash_attn-2.8.3+cu12torch2.8cxx11abiTRUE-cp311-cp311-linux_x86_64.whl" > ``` - -Then, download the model weights from [Lance-3B on Hugging Face](https://huggingface.co/bytedance-research/Lance) and place them in the `downloads/` directory: +然后，从 [Hugging Face 上的 Lance-3B](https://huggingface.co/bytedance-research/Lance) 下载所需的全部模型权重，并放置到 `downloads/` 目录下： ```bash from huggingface_hub import snapshot_download @@ -181,24 +187,24 @@ snapshot_download(cache_dir=cache_dir, ) ``` +## 📚 使用方法 -## 📚 Usage +### 推理 -### Inference - -#### Basic Usage +#### 基本用法 ```bash bash inference_lance.sh ``` -- Before running, please configure the inference parameters at the top of `inference_lance.sh`. -- **Supported tasks:** `t2i`, `t2v`, `i2v`, `image_edit`, `video_edit`, `x2t_image`, and `x2t_video`. You can modify `TASK_DEFAULT_CONFIGS` in `inference_lance.py` to customize the default data samples for each task. -- **Note:** For all tasks, we recommend following the `prompt` format used in the provided examples when writing input prompts, as this typically leads to better generation quality. +- 运行前，请先在 `inference_lance.sh` 顶部配置推理参数。 +- **支持任务：** `t2i`、`t2v`、`i2v`、`image_edit`、`video_edit`、`x2t_image` 和 `x2t_video`。你也可以在 `inference_lance.py` 中修改 `TASK_DEFAULT_CONFIGS`，自定义每个任务默认使用的数据样例。 +- **注意：** 对于所有任务，建议在编写输入 prompt 时参考提供示例中的 `prompt` 格式，这通常有助于获得更好的生成效果。 -#### Task Examples -##### Text-to-Video +#### 任务示例 + +##### 文生视频 ```bash bash inference_lance.sh \ @@ -211,7 +217,7 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/t2v ``` -##### Image-to-Video +##### 图生视频 ```bash bash inference_lance.sh \ @@ -224,11 +230,11 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/i2v ``` -Optional parameters for video generation task examples: +生成任务可选参数： -- `--ENHANCE_PROMPT true`: enable prompt rewrite for T2V/I2V. Prompt enhancement generally improves generation quality. This option requires `openai==2.26.0`, which is already listed in `requirements.txt`; if you did not install from `requirements.txt`, run `pip install openai==2.26.0` first. Before enabling it, set `API_KEY`, `MODEL_NAME`, and `BASE_URL` in `common/utils/caption_rewrite.py`. If no valid rewrite config is provided there, prompt rewrite is skipped; in that case, we recommend **writing prompts in the style of the provided examples**. +- `--ENHANCE_PROMPT true`：启用 T2V/I2V prompt rewrite。T2V 使用纯文本 rewrite，I2V 使用文本加输入图像 rewrite。prompt rewrite 通常能提升生成效果。该选项需要 `openai==2.26.0`，已写入 `requirements.txt`；如果没有通过 `requirements.txt` 安装依赖，请先执行 `pip install openai==2.26.0`。启用前请先在 `common/utils/caption_rewrite.py` 中配置 `API_KEY`、`MODEL_NAME` 和 `BASE_URL`；如果没有配置有效 rewrite 参数，会自动跳过 prompt rewrite，此时建议尽量参考提供示例中的 prompt 风格手写输入。 -##### Text-to-Image +##### 文生图 ```bash bash inference_lance.sh \ @@ -240,7 +246,7 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/t2i ``` -##### Video Editing +##### 视频编辑 ```bash bash inference_lance.sh \ @@ -250,7 +256,7 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/video_edit ``` -##### Image Editing +##### 图像编辑 ```bash bash inference_lance.sh \ @@ -260,7 +266,7 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/image_edit ``` -##### Video Understanding +##### 视频理解 ```bash bash inference_lance.sh \ @@ -271,7 +277,7 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/x2t_video ``` -##### Image Understanding +##### 图像理解 ```bash bash inference_lance.sh \ @@ -281,69 +287,68 @@ bash inference_lance.sh \ --SAVE_PATH_GEN results/x2t_image ``` -Optional parameters for all task examples: - -- `--CONFIG_PATH path/to/config.json`: use a custom validation JSON/JSONL file instead of the task default example config. +所有任务示例可选参数： +- `--CONFIG_PATH path/to/config.json`：使用自定义验证 JSON/JSONL 文件，而不是任务默认示例配置。
-Show task and parameter reference +展开任务和参数参考 -#### Available Tasks +#### 可用任务 -| Task Name | Description | Example JSON | +| 任务名 | 说明 | 示例 JSON | |------------------------|--------------------------------------------------|----------------------------------------------| -| `t2v` | Text-to-Video generation | `config/examples/t2v_example.json` | -| `t2i` | Text-to-Image generation | `config/examples/t2i_example.json` | -| `i2v` | Image-to-Video generation | `config/examples/i2v_example.json` | -| `image_edit` | Image editing | `config/examples/image_edit_example.json` | -| `video_edit` | Video editing | `config/examples/video_edit_example.json` | -| `x2t_image` | Image understanding | `config/examples/x2t_image_example.json` | -| `x2t_video` | Video understanding | `config/examples/x2t_video_example.json` | +| `t2v` | 文生视频 | `config/examples/t2v_example.json` | +| `t2i` | 文生图 | `config/examples/t2i_example.json` | +| `i2v` | 图生视频 | `config/examples/i2v_example.json` | +| `image_edit` | 图像编辑 | `config/examples/image_edit_example.json` | +| `video_edit` | 视频编辑 | `config/examples/video_edit_example.json` | +| `x2t_image` | 图像理解 | `config/examples/x2t_image_example.json` | +| `x2t_video` | 视频理解 | `config/examples/x2t_video_example.json` | -For understanding examples: +关于理解任务的示例文件： -- `config/examples/x2t_image_example.json`: image understanding examples for visual question answering, reasoning and image captioning. -- `config/examples/x2t_video_example.json`: video understanding examples for video question answering and video captioning. +- `config/examples/x2t_image_example.json`：用于图像理解示例，包括视觉问答、基于图像的推理和图像描述。 +- `config/examples/x2t_video_example.json`：用于视频理解示例，包括视频问答和视频描述。 -#### Parameters +#### 参数说明 -You can configure the following hyperparameters at the top of the `inference_lance.sh` script: +你可以在 `inference_lance.sh` 顶部配置以下超参数： -| Parameter | Default Value | Description | +| 参数 | 默认值 | 说明 | | --- | --- | --- | -| `MODEL_PATH` | `"downloads/Lance_3B"` | Path to the downloaded Lance model weights (`Lance_3B` or `Lance_3B_Video`). | -| `NUM_GPUS` | `1` | Number of GPUs to use for inference. | -| `VALIDATION_NUM_TIMESTEPS` | `30` | Number of denoising steps (e.g., 30 or 50). | -| `VALIDATION_TIMESTEP_SHIFT` | `3.5` | Timestep shift parameter for flow matching scheduling. | -| `CFG_TEXT_SCALE` | `4.0` | Classifier-Free Guidance (CFG) scale for text conditioning. | -| `VALIDATION_DATA_SEED` | `42` | Random seed for generation reproducibility. | -| `NUM_FRAMES` | `50` | Number of frames for video generation (Max: 121). *Unused for image tasks.* | -| `VIDEO_HEIGHT` / `VIDEO_WIDTH`| `768` | Spatial resolution. *Unused for editing tasks (determined by input image/video).* | -| `RESOLUTION` | `"video_480p"` | Base resolution preset (`image_768res` or `video_480p`). | -| `CONFIG_PATH` | `""` | Optional path to a custom validation JSON/JSONL file. When empty, the task default example config is used. | -| `ENHANCE_PROMPT` | `false` | Optional T2V/I2V prompt rewrite switch. T2V uses text-only rewrite; I2V uses text plus the input image. Prompt enhancement generally improves generation quality. This option requires `openai==2.26.0`; it is included in `requirements.txt`, or install it manually with `pip install openai==2.26.0`. Configure `API_KEY`, `MODEL_NAME`, and `BASE_URL` in `common/utils/caption_rewrite.py` before setting this to `true`; without a valid rewrite config, we recommend writing prompts in the style of the provided examples. | +| `MODEL_PATH` | `"downloads/Lance_3B"` | 下载后的 Lance 模型权重路径（如 `Lance_3B` 或 `Lance_3B_Video`）。 | +| `NUM_GPUS` | `1` | 用于推理的 GPU 数量。 | +| `VALIDATION_NUM_TIMESTEPS` | `30` | 去噪步数（例如 30 或 50）。 | +| `VALIDATION_TIMESTEP_SHIFT` | `3.5` | Flow matching 调度中的 timestep shift 参数。 | +| `CFG_TEXT_SCALE` | `4.0` | 文本条件的 CFG（Classifier-Free Guidance）系数。 | +| `VALIDATION_DATA_SEED` | `42` | 用于复现实验的随机种子。 | +| `NUM_FRAMES` | `50` | 视频生成帧数（最大 121）。*图像任务不使用该参数。* | +| `VIDEO_HEIGHT` / `VIDEO_WIDTH`| `768` | 空间分辨率。*编辑任务不使用该参数（由输入图像/视频决定）。* | +| `RESOLUTION` | `"video_480p"` | 基础分辨率预设（如 `image_768res` 或 `video_480p`）。 | +| `CONFIG_PATH` | `""` | 可选的自定义验证 JSON/JSONL 文件路径。为空时使用任务默认示例配置。 | +| `ENHANCE_PROMPT` | `false` | 可选的 T2V/I2V prompt rewrite 开关。T2V 使用纯文本 rewrite，I2V 使用文本加输入图像 rewrite。prompt rewrite 通常能提升生成效果。该选项需要 `openai==2.26.0`，已写入 `requirements.txt`；也可以手动执行 `pip install openai==2.26.0`。启用前请先在 `common/utils/caption_rewrite.py` 中配置 `API_KEY`、`MODEL_NAME` 和 `BASE_URL`；如果没有有效 rewrite 参数，建议尽量参考提供示例中的 prompt 风格手写输入。 |
### 🖥️ Gradio -You can launch the local Gradio demo for video/image generation, editing, and understanding: +你可以启动本地 Gradio demo，体验视频/图像生成、编辑和理解： ```bash python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 ``` -### Benchmarks +### 基准评测
-DPG-Bench Evaluation +DPG-Bench 评测
- + @@ -355,7 +360,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -373,7 +378,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -403,18 +408,18 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860
Models模型 # Params. Global Entity
Generation-only Models仅生成模型
SDXL3.5B83.2782.4380.9186.7680.4174.65Qwen-Image20B91.3291.5692.0294.3192.7388.32
Unified Models统一模型
Janus-Pro-7B7B86.9088.9089.4089.3289.4884.19
-

† indicates methods that use LLM rewriters for prompt rewriting before generation.

+

† 表示该方法在生成前使用 LLM rewriter 进行提示词改写。

-GenEval Evaluation +GenEval 评测
- + @@ -427,7 +432,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -445,7 +450,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -478,18 +483,18 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860
Models模型 # Params. 1-Obj. 2-Obj.
Generation-only Models仅生成模型
SDXL3.5B0.980.740.390.850.150.230.55Qwen-Image20B0.990.920.890.880.760.770.87
Unified Models统一模型
Janus-Pro-7B7B0.990.890.590.900.790.660.80
-

† indicates methods that use LLM rewriters for prompt rewriting before generation.

+

† 表示该方法在生成前使用 LLM rewriter 进行提示词改写。

-GEdit-Bench Evaluation +GEdit-Bench 评测
- + @@ -507,7 +512,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -519,7 +524,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -546,21 +551,21 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860
-VBench Evaluation (Video Generation) +VBench 评测（视频生成）
Models模型 # Params. BC CA
Generation-only Models仅生成模型
Gemini 2.0------------6.32Qwen-Image-Edit20B8.238.307.338.057.496.748.578.098.298.488.508.01
Unified Models统一模型
Lumina-DiMOO8B3.434.273.082.774.745.194.443.804.382.684.203.91
- - + + - + @@ -597,7 +602,7 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 - + @@ -621,29 +626,29 @@ python lance_gradio.py --server-name 0.0.0.0 --server-port 7860 -#### Running Benchmarks +#### 运行基准评测 -Ready-to-run benchmark scripts are provided under `benchmarks/`: +`benchmarks/` 目录下提供了可直接运行的基准评测脚本： -| Benchmark | Modality | Script | +| 基准 | 模态 | 脚本 | |------------------------|----------|---------------------------------------------------------------| -| GenEVAL (image gen) | Image | `benchmarks/image_gen/GenEVAL/sample_GenEVAL.sh` | -| DPG (image gen) | Image | `benchmarks/image_gen/DPG/sample_DPG.sh` | -| GEdit (image edit) | Image | `benchmarks/image_gen/GEdit/sample_GEdit.sh` | -| VBench (video gen) | Video | `benchmarks/video_gen/Vbench/sample_vbench.sh` | +| GenEVAL（图像生成） | 图像 | `benchmarks/image_gen/GenEVAL/sample_GenEVAL.sh` | +| DPG（图像生成） | 图像 | `benchmarks/image_gen/DPG/sample_DPG.sh` | +| GEdit（图像编辑） | 图像 | `benchmarks/image_gen/GEdit/sample_GEdit.sh` | +| VBench（视频生成） | 视频 | `benchmarks/video_gen/Vbench/sample_vbench.sh` | -## 📄 License +## 📄 许可证 Copyright 2025 ByteDance Ltd. and/or its affiliates. -## 🙏 Acknowledgements +## 🙏 致谢 -We would like to thank the contributors of [BAGEL](https://github.com/ByteDance-Seed/bagel), [Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct), and [Wan2.2](https://github.com/Wan-Video/Wan2.2) for their open research and contributions. +我们感谢 [BAGEL](https://github.com/ByteDance-Seed/bagel)、[Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct) 和 [Wan2.2](https://github.com/Wan-Video/Wan2.2) 的贡献者，感谢他们开放的研究与社区贡献。 -## 💖 Citation +## 💖 引用 -If you find **Lance** useful for your project or research, welcome to 🌟 this repo and cite our work using the following BibTeX: +如果 **Lance** 对您的项目或研究有帮助，欢迎 🌟 本仓库，并使用以下 BibTeX 引用我们的工作： ```bibtex @misc{fu2026lanceunifiedmultimodalmodeling, @@ -657,6 +662,6 @@ If you find **Lance** useful for your project or research, welcome to 🌟 this } ``` -## 📞 Contact +## 📞 联系方式 -For questions, issues, or collaborations, please contact [Mengqi Huang](https://corleone-huang.github.io/) and [Jianzhu Guo](https://guojianzhu.com/). +如有问题、反馈或合作需求，请联系 [Mengqi Huang](https://corleone-huang.github.io/) 和 [Jianzhu Guo](https://guojianzhu.com/)。
TypeModel类型模型 # Params. Total Score ↑
Gen. Only仅生成 ModelScope1.7B75.75
Wan2.1-T2V14B83.69
Unified统一模型 HaproOmni7B78.10