Initial commit

.github/ISSUE_TEMPLATE/bug_report.md

+---
+name: Bug Report
+about: Use this template to report bugs in the project.
+title: "[Bug] "
+labels: bug
+assignees: ''
+
+---
+
+### Description
+Briefly describe the bug you encountered.
+
+### Steps to Reproduce
+1. First step of the operation.
+2. Second step of the operation.
+3. ...
+
+### Expected Result
+Describe the normal behavior you expected.
+
+### Actual Result
+Describe the abnormal situation that actually occurred.
+
+### Environment Information
+- Operating System: [e.g., Ubuntu 22.04]
+- Commit ID: [Version of the project]
+
+### Log Information
+Please provide relevant error logs or debugging information.
+
+### Additional Information
+If there is any other information that can help solve the problem, please add it here.

.github/workflows/lint.yml

+name: lint
+
+on:
+  pull_request:
+  push:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Python 3.11
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.11'
+
+      - name: Cache Python dependencies
+        uses: actions/cache@v3
+        with:
+          path: ~/.cache/pip
+          key: ${{ runner.os }}-pip-${{ hashFiles('requirements.txt') }}
+          restore-keys: |
+            ${{ runner.os }}-pip-
+
+      - name: Install pre-commit hook
+        run: |
+          pip install pre-commit ruff
+
+      - name: Check pre-commit config file
+        run: |
+          if [ ! -f ".pre-commit-config.yaml" ]; then
+            echo "Error: .pre-commit-config.yaml not found."
+            exit 1
+          fi
+
+      - name: Linting
+        run: |
+          echo "Running pre-commit on all files..."
+          pre-commit run --all-files || {
+            echo "Linting failed. Please check the above output for details."
+            exit 1
+          }

.gitignore

+*.pth
+*.pt
+*.onnx
+*.pk
+*.model
+*.zip
+*.tar
+*.pyc
+*.log
+*.o
+*.so
+*.a
+*.exe
+*.out
+.idea
+**.DS_Store**
+**/__pycache__/**
+**.swp
+.vscode/
+.env
+.log
+*.pid
+*.ipynb*
+*.mp4
+build/
+dist/
+.cache/
+server_cache/
+app/.gradio/
+*.pkl
+save_results/*
+*.egg-info/

.pre-commit-config.yaml

+# Follow https://verdantfox.com/blog/how-to-use-git-pre-commit-hooks-the-hard-way-and-the-easy-way
+repos:
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.11.0
+    hooks:
+      - id: ruff
+        args: [--fix, --respect-gitignore, --config=pyproject.toml]
+      - id: ruff-format
+        args: [--config=pyproject.toml]
+
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.5.0
+    hooks:
+      - id: trailing-whitespace
+      - id: end-of-file-fixer
+      - id: check-yaml
+      - id: check-toml
+      - id: check-added-large-files
+        args: ['--maxkb=3000']  # Allow files up to 3MB
+      - id: check-case-conflict
+      - id: check-merge-conflict
+      - id: debug-statements

LICENSE

+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

README_lightx2v_zh.md

+<div align="center" style="font-family: charter;">
+  <h1>⚡️ LightX2V:<br> 轻量级视频生成推理框架</h1>
+
+<img alt="logo" src="assets/img_lightx2v.png" width=75%></img>
+
+[![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Ask DeepWiki](https://deepwiki.com/badge.svg)](https://deepwiki.com/ModelTC/lightx2v)
+[![Doc](https://img.shields.io/badge/docs-English-99cc2)](https://lightx2v-en.readthedocs.io/en/latest)
+[![Doc](https://img.shields.io/badge/文档-中文-99cc2)](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest)
+[![Papers](https://img.shields.io/badge/论文集-中文-99cc2)](https://lightx2v-papers-zhcn.readthedocs.io/zh-cn/latest)
+[![Docker](https://img.shields.io/badge/Docker-2496ED?style=flat&logo=docker&logoColor=white)](https://hub.docker.com/r/lightx2v/lightx2v/tags)
+
+**\[ [English](README.md) | 中文 \]**
+
+</div>
+
+--------------------------------------------------------------------------------
+
+**LightX2V** 是一个先进的轻量级图像视频生成推理框架，专为提供高效、高性能的图像视频生成解决方案而设计。该统一平台集成了多种前沿的图像视频生成技术，支持文本生成视频(T2V)和图像生成视频(I2V)，文本生图片(T2I)，图像编辑(I2I)等多样化生成任务。**X2V 表示将不同的输入模态(X，如文本或图像)转换为视觉输出(Vision)**。
+
+> 🌐 **立即在线体验！** 无需安装即可体验 LightX2V：**[LightX2V 在线服务](https://x2v.light-ai.top/login)** - 免费、轻量、快速的AI数字人视频生成平台。
+
+> 👋 **加入微信交流群，LightX2V加群机器人微信号: random42seed**
+
+## 🧾 社区代码贡献指南
+
+在提交之前，请确保代码格式符合项目规范。可以使用如下执行命令，确保项目代码格式的一致性。
+
+```bash
+pip install ruff pre-commit
+pre-commit run --all-files
+```
+
+除了LightX2V团队的贡献，我们也收到一些社区开发者的贡献，包括但不限于：
+
+- [triple-Mu](https://github.com/triple-Mu)
+- [vivienfanghuagood](https://github.com/vivienfanghuagood)
+- [yeahdongcn](https://github.com/yeahdongcn)
+- [kikidouloveme79](https://github.com/kikidouloveme79)
+
+## :fire: 最新动态
+
+- **2026年1月20日:** 🚀 我们支持了[LTX-2](https://huggingface.co/Lightricks/LTX-2)音频-视频生成模型，包含CFG并行、block级别offload、FP8 per-tensor量化等先进特性。使用示例可参考[examples/ltx2](https://github.com/ModelTC/LightX2V/tree/main/examples/ltx2)和[scripts/ltx2](https://github.com/ModelTC/LightX2V/tree/main/scripts/ltx2)。
+
+- **2026年1月6日:** 🚀 我们更新了[Qwen-Image-2512](https://huggingface.co/Qwen/Qwen-Image-2512)和[Qwen/Qwen-Image-Edit-2511](https://huggingface.co/Qwen/Qwen-Image-Edit-2511)的8步的CFG/步数蒸馏模型。可以在[Qwen-Image-Edit-2511-Lightning](https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning)和[Qwen-Image-2512-Lightning](https://huggingface.co/lightx2v/Qwen-Image-2512-Lightning)下载对应的权重进行使用。使用教程参考[这里](https://github.com/ModelTC/LightX2V/tree/main/examples/qwen_image)。
+
+- **2026年1月6日:** 🚀 支持燧原 Enflame S60 (GCU) 的部署。
+
+- **2025年12月31日:** 🚀 我们Day0支持了[Qwen-Image-2512](https://huggingface.co/Qwen/Qwen-Image-2512) 文生图模型. 我们的[HuggingFace](https://huggingface.co/lightx2v/Qwen-Image-2512-Lightning) 已经更新了CFG/步数蒸馏lora权重。使用方式可以参考[这里](https://github.com/ModelTC/LightX2V/tree/main/examples/qwen_image)。
+
+- **2025年12月27日:** 🚀 支持摩尔线程 MUSA 的部署。
+
+- **2025年12月25日:** 🚀 支持 AMD ROCm 和 Ascend 910B 的部署。
+
+- **2025年12月23日:** 🚀 我们Day0支持了[Qwen/Qwen-Image-Edit-2511](https://huggingface.co/Qwen/Qwen-Image-Edit-2511)的图像编辑模型，H100单卡，LightX2V可带来约1.4倍的速度提升，支持CFG并行/Ulysses并行，高效Offload等技术。我们的[HuggingFace](https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning)已经更新了CFG/步数蒸馏lora和FP8权重。使用方式可以参考[这里](https://github.com/ModelTC/LightX2V/tree/main/examples/qwen_image)。结合LightX2V，4步CFG/步数蒸馏，FP8模型，最高可以加速约42倍。可以在[LightX2V 在线服务](https://x2v.light-ai.top/login)的图生图的Qwen-Image-Edit-2511进行体验。
+
+- **2025年12月22日:** 🚀 新增 **Wan2.1 NVFP4 量化感知 4 步蒸馏模型** 支持；模型与权重已发布在 HuggingFace： [Wan-NVFP4](https://huggingface.co/lightx2v/Wan-NVFP4)。
+
+- **2025年12月15日:** 🚀 支持 海光DCU 硬件上的部署。
+
+- **2025年12月4日:** 🚀 支持 GGUF 格式模型推理，以及在寒武纪 MLU590、MetaX C500 硬件上的部署。
+
+- **2025年11月24日:** 🚀 我们发布了HunyuanVideo-1.5的4步蒸馏模型！这些模型支持**超快速4步推理**，无需CFG配置，相比标准50步推理可实现约**25倍加速**。现已提供基础版本和FP8量化版本：[Hy1.5-Distill-Models](https://huggingface.co/lightx2v/Hy1.5-Distill-Models)。
+
+- **2025年11月21日:** 🚀 我们Day0支持了[HunyuanVideo-1.5](https://huggingface.co/tencent/HunyuanVideo-1.5)的视频生成模型，同样GPU数量，LightX2V可带来约2倍以上的速度提升，并支持更低显存GPU部署(如24G RTX4090)。支持CFG并行/Ulysses并行，高效Offload，TeaCache/MagCache等技术。同时支持沐曦，寒武纪等国产芯片部署。我们很快将在我们的[HuggingFace主页](https://huggingface.co/lightx2v)更新更多模型，包括步数蒸馏，VAE蒸馏等相关模型。量化模型和轻量VAE模型现已可用：[Hy1.5-Quantized-Models](https://huggingface.co/lightx2v/Hy1.5-Quantized-Models)用于量化推理，[HunyuanVideo-1.5轻量TAE](https://huggingface.co/lightx2v/Autoencoders/blob/main/lighttaehy1_5.safetensors)用于快速VAE解码。使用教程参考[这里](https://github.com/ModelTC/LightX2V/tree/main/scripts/hunyuan_video_15)，或查看[示例目录](https://github.com/ModelTC/LightX2V/tree/main/examples)获取代码示例。
+
+
+## 🏆 性能测试数据 (更新于 2025.12.01)
+
+### 📊 推理框架之间性能对比 (H100)
+
+| Framework | GPUs | Step Time | Speedup |
+|-----------|---------|---------|---------|
+| Diffusers | 1 | 9.77s/it | 1x |
+| xDiT | 1 | 8.93s/it | 1.1x |
+| FastVideo | 1 | 7.35s/it | 1.3x |
+| SGL-Diffusion | 1 | 6.13s/it | 1.6x |
+| **LightX2V** | 1 | **5.18s/it** | **1.9x** 🚀 |
+| FastVideo | 8 | 2.94s/it | 1x |
+| xDiT | 8 | 2.70s/it | 1.1x |
+| SGL-Diffusion | 8 | 1.19s/it | 2.5x |
+| **LightX2V** | 8 | **0.75s/it** | **3.9x** 🚀 |
+
+### 📊 推理框架之间性能对比 (RTX 4090D)
+
+| Framework | GPUs | Step Time | Speedup |
+|-----------|---------|---------|---------|
+| Diffusers | 1 | 30.50s/it | 1x |
+| FastVideo | 1 | 22.66s/it | 1.3x |
+| xDiT | 1 | OOM | OOM |
+| SGL-Diffusion | 1 | OOM | OOM |
+| **LightX2V** | 1 | **20.26s/it** | **1.5x** 🚀 |
+| FastVideo | 8 | 15.48s/it | 1x |
+| xDiT | 8 | OOM | OOM |
+| SGL-Diffusion | 8 | OOM | OOM |
+| **LightX2V** | 8 | **4.75s/it** | **3.3x** 🚀 |
+
+### 📊 LightX2V不同配置之间性能对比
+
+| Framework | GPU | Configuration | Step Time | Speedup |
+|-----------|-----|---------------|-----------|---------------|
+| **LightX2V** | H100 | 8 GPUs + cfg | 0.75s/it | 1x |
+| **LightX2V** | H100 | 8 GPUs + no cfg | 0.39s/it | 1.9x |
+| **LightX2V** | H100 | **8 GPUs + no cfg + fp8** | **0.35s/it** | **2.1x** 🚀 |
+| **LightX2V** | 4090D | 8 GPUs + cfg | 4.75s/it | 1x |
+| **LightX2V** | 4090D | 8 GPUs + no cfg | 3.13s/it | 1.5x |
+| **LightX2V** | 4090D | **8 GPUs + no cfg + fp8** | **2.35s/it** | **2.0x** 🚀 |
+
+**注意**: 所有以上性能数据均在 Wan2.1-I2V-14B-480P(40 steps, 81 frames) 上测试。此外，我们[HuggingFace 主页](https://huggingface.co/lightx2v)还提供了4步蒸馏模型。
+
+
+## 💡 快速开始
+
+
+详细使用说明请参考我们的文档：**[英文文档](https://lightx2v-en.readthedocs.io/en/latest/) | [中文文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/)**
+
+**我们强烈推荐使用 Docker 环境，这是最简单快捷的环境安装方式。具体参考：文档中的快速入门章节。**
+
+### 从 Git 安装
+```bash
+pip install -v git+https://github.com/ModelTC/LightX2V.git
+```
+
+### 从源码构建
+```bash
+git clone https://github.com/ModelTC/LightX2V.git
+cd LightX2V
+uv pip install -v . # pip install -v .
+```
+
+### （可选）安装注意力/量化算子
+注意力算子安装说明请参考我们的文档：**[英文文档](https://lightx2v-en.readthedocs.io/en/latest/getting_started/quickstart.html#step-4-install-attention-operators) | [中文文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/getting_started/quickstart.html#id9)**
+
+### 使用示例
+```python
+# examples/wan/wan_i2v.py
+"""
+Wan2.2 image-to-video generation example.
+This example demonstrates how to use LightX2V with Wan2.2 model for I2V generation.
+"""
+
+from lightx2v import LightX2VPipeline
+
+# Initialize pipeline for Wan2.2 I2V task
+# For wan2.1, use model_cls="wan2.1"
+pipe = LightX2VPipeline(
+    model_path="/path/to/Wan2.2-I2V-A14B",
+    model_cls="wan2.2_moe",
+    task="i2v",
+)
+
+# Alternative: create generator from config JSON file
+# pipe.create_generator(
+#     config_json="configs/wan22/wan_moe_i2v.json"
+# )
+
+# Enable offloading to significantly reduce VRAM usage with minimal speed impact
+# Suitable for RTX 30/40/50 consumer GPUs
+pipe.enable_offload(
+    cpu_offload=True,
+    offload_granularity="block",  # For Wan models, supports both "block" and "phase"
+    text_encoder_offload=True,
+    image_encoder_offload=False,
+    vae_offload=False,
+)
+
+# Create generator manually with specified parameters
+pipe.create_generator(
+    attn_mode="sage_attn2",
+    infer_steps=40,
+    height=480,  # Can be set to 720 for higher resolution
+    width=832,  # Can be set to 1280 for higher resolution
+    num_frames=81,
+    guidance_scale=[3.5, 3.5],  # For wan2.1, guidance_scale is a scalar (e.g., 5.0)
+    sample_shift=5.0,
+)
+
+# Generation parameters
+seed = 42
+prompt = "Summer beach vacation style, a white cat wearing sunglasses sits on a surfboard. The fluffy-furred feline gazes directly at the camera with a relaxed expression. Blurred beach scenery forms the background featuring crystal-clear waters, distant green hills, and a blue sky dotted with white clouds. The cat assumes a naturally relaxed posture, as if savoring the sea breeze and warm sunlight. A close-up shot highlights the feline's intricate details and the refreshing atmosphere of the seaside."
+negative_prompt = "镜头晃动，色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走"
+image_path="/path/to/img_0.jpg"
+save_result_path = "/path/to/save_results/output.mp4"
+
+# Generate video
+pipe.generate(
+    seed=seed,
+    image_path=image_path,
+    prompt=prompt,
+    negative_prompt=negative_prompt,
+    save_result_path=save_result_path,
+)
+
+```
+
+**NVFP4（量化感知 4 步）资源**
+- 推理示例：`examples/wan/wan_i2v_nvfp4.py`（I2V），`examples/wan/wan_t2v_nvfp4.py`（T2V）。
+- NVFP4 算子编译/安装指南：参见 `lightx2v_kernel/README.md`。
+
+> 💡 **更多示例**: 更多使用案例，包括量化、卸载、缓存等进阶配置，请参考 [examples 目录](https://github.com/ModelTC/LightX2V/tree/main/examples)。
+
+## 🤖 支持的模型生态
+
+### 官方开源模型
+- ✅ [LTX-2](https://huggingface.co/Lightricks/LTX-2)
+- ✅ [HunyuanVideo-1.5](https://huggingface.co/tencent/HunyuanVideo-1.5)
+- ✅ [Wan2.1 & Wan2.2](https://huggingface.co/Wan-AI/)
+- ✅ [Qwen-Image](https://huggingface.co/Qwen/Qwen-Image)
+- ✅ [Qwen-Image-Edit](https://huggingface.co/spaces/Qwen/Qwen-Image-Edit)
+- ✅ [Qwen-Image-Edit-2509](https://huggingface.co/Qwen/Qwen-Image-Edit-2509)
+- ✅ [Qwen-Image-Edit-2511](https://huggingface.co/Qwen/Qwen-Image-Edit-2511)
+
+### 量化模型和蒸馏模型/Lora (**🚀 推荐：4步推理**)
+- ✅ [Wan2.1-Distill-Models](https://huggingface.co/lightx2v/Wan2.1-Distill-Models)
+- ✅ [Wan2.2-Distill-Models](https://huggingface.co/lightx2v/Wan2.2-Distill-Models)
+- ✅ [Wan2.1-Distill-Loras](https://huggingface.co/lightx2v/Wan2.1-Distill-Loras)
+- ✅ [Wan2.2-Distill-Loras](https://huggingface.co/lightx2v/Wan2.2-Distill-Loras)
+- ✅ [Wan2.1-Distill-NVFP4](https://huggingface.co/lightx2v/Wan-NVFP4)
+- ✅ [Qwen-Image-Edit-2511-Lightning](https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning)
+
+### 轻量级自编码器模型(**🚀 推荐：推理快速 + 内存占用低**)
+- ✅ [Autoencoders](https://huggingface.co/lightx2v/Autoencoders)
+
+### 自回归模型
+- ✅ [Wan2.1-T2V-CausVid](https://huggingface.co/lightx2v/Wan2.1-T2V-14B-CausVid)
+- ✅ [Self-Forcing](https://github.com/guandeh17/Self-Forcing)
+- ✅ [Matrix-Game-2.0](https://huggingface.co/Skywork/Matrix-Game-2.0)
+
+🔔 可以关注我们的[HuggingFace主页](https://huggingface.co/lightx2v)，及时获取我们团队的模型。
+
+💡 参考[模型结构文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/getting_started/model_structure.html)快速上手 LightX2V
+
+## 🚀 前端展示
+
+我们提供了多种前端界面部署方式：
+
+- **🎨 Gradio界面**: 简洁易用的Web界面，适合快速体验和原型开发
+  - 📖 [Gradio部署文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_gradio.html)
+- **🎯 ComfyUI界面**: 强大的节点式工作流界面，支持复杂的视频生成任务
+  - 📖 [ComfyUI部署文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_comfyui.html)
+- **🚀 Windows一键部署**: 专为Windows用户设计的便捷部署方案，支持自动环境配置和智能参数优化
+  - 📖 [Windows一键部署文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_local_windows.html)
+
+**💡 推荐方案**:
+- **首次使用**: 建议选择Windows一键部署方案
+- **高级用户**: 推荐使用ComfyUI界面获得更多自定义选项
+- **快速体验**: Gradio界面提供最直观的操作体验
+
+## 🚀 核心特性
+
+### 🎯 **极致性能优化**
+- **🔥 SOTA推理速度**: 通过步数蒸馏和系统优化实现**20倍**极速加速(单GPU)
+- **⚡️ 革命性4步蒸馏**: 将原始40-50步推理压缩至仅需4步，且无需CFG配置
+- **🛠️ 先进算子支持**: 集成顶尖算子，包括[Sage Attention](https://github.com/thu-ml/SageAttention)、[Flash Attention](https://github.com/Dao-AILab/flash-attention)、[Radial Attention](https://github.com/mit-han-lab/radial-attention)、[q8-kernel](https://github.com/KONAKONA666/q8_kernels)、[sgl-kernel](https://github.com/sgl-project/sglang/tree/main/sgl-kernel)、[vllm](https://github.com/vllm-project/vllm)
+
+### 💾 **资源高效部署**
+- **💡 突破硬件限制**: **仅需8GB显存 + 16GB内存**即可运行14B模型生成480P/720P视频
+- **🔧 智能参数卸载**: 先进的磁盘-CPU-GPU三级卸载架构，支持阶段/块级别的精细化管理
+- **⚙️ 全面量化支持**: 支持`w8a8-int8`、`w8a8-fp8`、`w4a4-nvfp4`等多种量化策略
+
+### 🎨 **丰富功能生态**
+- **📈 智能特征缓存**: 智能缓存机制，消除冗余计算，提升效率
+- **🔄 并行推理加速**: 多GPU并行处理，显著提升性能表现
+- **📱 灵活部署选择**: 支持Gradio、服务化部署、ComfyUI等多种部署方式
+- **🎛️ 动态分辨率推理**: 自适应分辨率调整，优化生成质量
+- **🎞️ 视频帧插值**: 基于RIFE的帧插值技术，实现流畅的帧率提升
+
+
+## 📚 技术文档
+
+### 📖 **方法教程**
+- [模型量化](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/quantization.html) - 量化策略全面指南
+- [特征缓存](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/cache.html) - 智能缓存机制详解
+- [注意力机制](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/attention.html) - 前沿注意力算子
+- [参数卸载](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/offload.html) - 三级存储架构
+- [并行推理](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/parallel.html) - 多GPU加速策略
+- [变分辨率推理](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/changing_resolution.html) - U型分辨率策略
+- [步数蒸馏](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/step_distill.html) - 4步推理技术
+- [视频帧插值](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/method_tutorials/video_frame_interpolation.html) - 基于RIFE的帧插值技术
+
+### 🛠️ **部署指南**
+- [低资源场景部署](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/for_low_resource.html) - 优化的8GB显存解决方案
+- [低延迟场景部署](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/for_low_latency.html) - 极速推理优化
+- [Gradio部署](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_gradio.html) - Web界面搭建
+- [服务化部署](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_service.html) - 生产级API服务部署
+- [Lora模型部署](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/lora_deploy.html) - Lora灵活部署
+
+
+## 🤝 致谢
+
+我们向所有启发和促进LightX2V开发的模型仓库和研究社区表示诚挚的感谢。此框架基于开源社区的集体努力而构建。包括但不限于：
+
+- [Tencent-Hunyuan](https://github.com/Tencent-Hunyuan)
+- [Wan-Video](https://github.com/Wan-Video)
+- [Qwen-Image](https://github.com/QwenLM/Qwen-Image)
+- [LightLLM](https://github.com/ModelTC/LightLLM)
+- [sglang](https://github.com/sgl-project/sglang)
+- [vllm](https://github.com/vllm-project/vllm)
+- [flash-attention](https://github.com/Dao-AILab/flash-attention)
+- [SageAttention](https://github.com/thu-ml/SageAttention)
+- [flashinfer](https://github.com/flashinfer-ai/flashinfer)
+- [MagiAttention](https://github.com/SandAI-org/MagiAttention)
+- [radial-attention](https://github.com/mit-han-lab/radial-attention)
+- [xDiT](https://github.com/xdit-project/xDiT)
+- [FastVideo](https://github.com/hao-ai-lab/FastVideo)
+
+## 🌟 Star 历史
+
+[![Star History Chart](https://api.star-history.com/svg?repos=ModelTC/lightx2v&type=Timeline)](https://star-history.com/#ModelTC/lightx2v&Timeline)
+
+## ✏️ 引用
+
+如果您发现LightX2V对您的研究有用，请考虑引用我们的工作：
+
+```bibtex
+@misc{lightx2v,
+ author = {LightX2V Contributors},
+ title = {LightX2V: Light Video Generation Inference Framework},
+ year = {2025},
+ publisher = {GitHub},
+ journal = {GitHub repository},
+ howpublished = {\url{https://github.com/ModelTC/lightx2v}},
+}
+```
+
+## 📞 联系与支持
+
+如有任何问题、建议或需要支持，欢迎通过以下方式联系我们：
+- 🐛 [GitHub Issues](https://github.com/ModelTC/lightx2v/issues) - 错误报告和功能请求
+
+---
+
+<div align="center">
+由 LightX2V 团队用 ❤️ 构建
+</div>

app/README.md

+# Gradio Demo
+
+Please refer our gradio deployment doc:
+
+[English doc: Gradio Deployment](https://lightx2v-en.readthedocs.io/en/latest/deploy_guides/deploy_gradio.html)
+
+[中文文档: Gradio 部署](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_gradio.html)
+
+## 🚀 Quick Start (快速开始)
+
+For Windows users, we provide a convenient one-click deployment solution with automatic environment configuration and intelligent parameter optimization. Please refer to the [One-Click Gradio Launch](https://lightx2v-en.readthedocs.io/en/latest/deploy_guides/deploy_local_windows.html) section for detailed instructions.
+
+对于Windows用户，我们提供了便捷的一键部署方式，支持自动环境配置和智能参数优化。详细操作请参考[一键启动Gradio](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/deploy_guides/deploy_local_windows.html)章节。

app/gradio_demo.py

+"""
+重构后的 Gradio Demo 主入口文件
+整合了所有模块，支持中英文切换
+"""
+
+import argparse
+import gc
+import json
+import logging
+import os
+import warnings
+
+import torch
+from loguru import logger
+from utils.i18n import DEFAULT_LANG, set_language
+from utils.model_utils import cleanup_memory, extract_op_name, get_model_configs
+from utils.ui_builder import build_ui, generate_unique_filename, get_auto_config_dict
+
+from lightx2v.utils.input_info import init_empty_input_info, update_input_info_from_dict
+from lightx2v.utils.set_config import get_default_config
+
+warnings.filterwarnings("ignore", category=UserWarning, module="huggingface_hub")
+warnings.filterwarnings("ignore", category=UserWarning, module="huggingface_hub.utils")
+logging.getLogger("httpx").setLevel(logging.ERROR)
+logging.getLogger("httpcore").setLevel(logging.ERROR)
+logging.getLogger("urllib3").setLevel(logging.ERROR)
+
+os.environ["PROFILING_DEBUG_LEVEL"] = "2"
+os.environ["PYTORCH_CUDA_ALLOC_CONF"] = "expandable_segments:True"
+os.environ["DTYPE"] = "BF16"
+
+
+logger.add(
+    "inference_logs.log",
+    rotation="100 MB",
+    encoding="utf-8",
+    enqueue=True,
+    backtrace=True,
+    diagnose=True,
+)
+
+
+global_runner = None
+current_config = None
+cur_dit_path = None
+
+
+def run_inference(
+    prompt="",
+    negative_prompt="",
+    save_result_path="",
+    infer_steps=4,
+    num_frames=81,
+    resolution="480p",
+    seed=42,
+    sample_shift=5,
+    cfg_scale=1,
+    fps=16,
+    model_path_input=None,
+    model_type_input="wan2.1",
+    task_type_input="i2v",
+    dit_path_input=None,
+    high_noise_path_input=None,
+    low_noise_path_input=None,
+    t5_path_input=None,
+    clip_path_input="",
+    image_path=None,
+    vae_path=None,
+    qwen_image_dit_path_input=None,
+    qwen_image_vae_path_input=None,
+    qwen_image_scheduler_path_input=None,
+    qwen25vl_encoder_path_input=None,
+    z_image_dit_path_input=None,
+    z_image_vae_path_input=None,
+    z_image_scheduler_path_input=None,
+    qwen3_encoder_path_input=None,
+    aspect_ratio="1:1",
+    use_lora=None,
+    lora_path=None,
+    lora_strength=None,
+    high_noise_lora_path=None,
+    low_noise_lora_path=None,
+    high_noise_lora_strength=None,
+    low_noise_lora_strength=None,
+):
+    cleanup_memory()
+
+    auto_config = get_auto_config_dict(model_type=model_type_input, resolution=resolution, num_frames=num_frames, task_type=task_type_input)
+
+    # 从 auto_config 中获取 offload 和 rope 相关配置
+    rope_chunk = auto_config["rope_chunk_val"]
+    rope_chunk_size = auto_config["rope_chunk_size_val"]
+    cpu_offload = auto_config["cpu_offload_val"]
+    offload_granularity = auto_config["offload_granularity_val"]
+    lazy_load = auto_config["lazy_load_val"]
+    t5_cpu_offload = auto_config["t5_cpu_offload_val"]
+    clip_cpu_offload = auto_config["clip_cpu_offload_val"]
+    vae_cpu_offload = auto_config["vae_cpu_offload_val"]
+    unload_modules = auto_config["unload_modules_val"]
+    attention_type = auto_config["attention_type_val"]
+    quant_op = auto_config["quant_op_val"]
+    use_tiling_vae = auto_config["use_tiling_vae_val"]
+    clean_cuda_cache = auto_config["clean_cuda_cache_val"]
+    quant_op = extract_op_name(quant_op)
+    attention_type = extract_op_name(attention_type)
+    task = task_type_input
+
+    is_image_output = task in ["i2i", "t2i"]
+    save_result_path = generate_unique_filename(output_dir, is_image=is_image_output)
+
+    if cfg_scale == 1:
+        enable_cfg = False
+    else:
+        enable_cfg = True
+
+    model_config = get_model_configs(
+        model_type_input,
+        model_path_input,
+        dit_path_input,
+        high_noise_path_input,
+        low_noise_path_input,
+        t5_path_input,
+        clip_path_input,
+        vae_path,
+        qwen_image_dit_path_input,
+        qwen_image_vae_path_input,
+        qwen_image_scheduler_path_input,
+        qwen25vl_encoder_path_input,
+        z_image_dit_path_input,
+        z_image_vae_path_input,
+        z_image_scheduler_path_input,
+        qwen3_encoder_path_input,
+        quant_op,
+        use_lora=use_lora,
+        lora_path=lora_path,
+        lora_strength=lora_strength,
+        high_noise_lora_path=high_noise_lora_path,
+        low_noise_lora_path=low_noise_lora_path,
+        high_noise_lora_strength=high_noise_lora_strength,
+        low_noise_lora_strength=low_noise_lora_strength,
+    )
+    model_cls = model_config["model_cls"]
+    model_path = model_config["model_path"]
+
+    global global_runner, current_config, cur_dit_path
+
+    logger.info(f"Auto-determined model_cls: {model_cls} (model type: {model_type_input})")
+
+    if model_cls.startswith("wan2.2"):
+        current_dit_path = f"{high_noise_path_input}|{low_noise_path_input}" if high_noise_path_input and low_noise_path_input else None
+    else:
+        current_dit_path = dit_path_input
+
+    needs_reinit = lazy_load or unload_modules or global_runner is None or cur_dit_path != current_dit_path
+
+    config_graio = {
+        "infer_steps": infer_steps,
+        "target_video_length": num_frames,
+        "resolution": resolution,
+        "resize_mode": "adaptive",
+        "self_attn_1_type": attention_type,
+        "cross_attn_1_type": attention_type,
+        "cross_attn_2_type": attention_type,
+        "attn_type": attention_type,
+        "enable_cfg": enable_cfg,
+        "sample_guide_scale": cfg_scale,
+        "sample_shift": sample_shift,
+        "fps": fps,
+        "feature_caching": "NoCaching",
+        "do_mm_calib": False,
+        "parallel_attn_type": None,
+        "parallel_vae": False,
+        "max_area": False,
+        "vae_stride": (4, 8, 8),
+        "patch_size": (1, 2, 2),
+        "lora_path": None,
+        "strength_model": 1.0,
+        "use_prompt_enhancer": False,
+        "text_len": 512,
+        "denoising_step_list": [1000, 750, 500, 250],
+        "cpu_offload": True if "wan2.2" in model_cls else cpu_offload,
+        "offload_granularity": ("phase" if "wan2.2" in model_cls else offload_granularity),
+        "t5_cpu_offload": t5_cpu_offload,
+        "clip_cpu_offload": clip_cpu_offload,
+        "vae_cpu_offload": vae_cpu_offload,
+        "use_tiling_vae": use_tiling_vae,
+        "lazy_load": lazy_load,
+        "rope_chunk": rope_chunk,
+        "rope_chunk_size": rope_chunk_size,
+        "clean_cuda_cache": clean_cuda_cache,
+        "unload_modules": unload_modules,
+        "seq_parallel": False,
+        "warm_up_cpu_buffers": False,
+        "boundary_step_index": 2,
+        "boundary": 0.900,
+        "use_image_encoder": False if "wan2.2" in model_cls else True,
+        "rope_type": "torch",
+        "t5_lazy_load": lazy_load,
+        "bucket_shape": {
+            "0.667": [[480, 832], [544, 960], [720, 960]],
+            "1.500": [[832, 480], [960, 544], [960, 720]],
+            "1.000": [[480, 480], [576, 576], [720, 720]],
+        },
+        "aspect_ratio": aspect_ratio,
+    }
+
+    args = argparse.Namespace(
+        model_cls=model_cls,
+        seed=seed,
+        task=task,
+        model_path=model_path,
+        prompt_enhancer=None,
+        prompt=prompt,
+        negative_prompt=negative_prompt,
+        image_path=image_path,
+        save_result_path=save_result_path,
+        return_result_tensor=False,
+        aspect_ratio=aspect_ratio,
+        target_shape=[],
+    )
+    input_info = init_empty_input_info(args.task)
+    config = get_default_config()
+    config.update({k: v for k, v in vars(args).items()})
+    config.update(config_graio)
+    config.update(model_config)
+
+    # 如果 model_config 中包含 lora_configs，设置 lora_dynamic_apply
+    if config.get("lora_configs"):
+        config["lora_dynamic_apply"] = True
+
+    logger.info(f"Using model: {model_path}")
+    logger.info(f"Inference config:\n{json.dumps(config, indent=4, ensure_ascii=False)}")
+
+    # 初始化或重用 runner
+    runner = global_runner
+    if needs_reinit:
+        if runner is not None:
+            del runner
+            torch.cuda.empty_cache()
+            gc.collect()
+
+        from lightx2v.infer import init_runner
+
+        runner = init_runner(config)
+
+        data = args.__dict__
+        update_input_info_from_dict(input_info, data)
+
+        current_config = config
+        cur_dit_path = current_dit_path
+
+        if not lazy_load:
+            global_runner = runner
+    else:
+        runner.config = config
+        data = args.__dict__
+        update_input_info_from_dict(input_info, data)
+
+    runner.run_pipeline(input_info)
+    cleanup_memory()
+
+    return save_result_path
+
+
+def main(lang=DEFAULT_LANG):
+    """主函数"""
+    set_language(lang)
+    demo = build_ui(
+        model_path=model_path,
+        output_dir=output_dir,
+        run_inference=run_inference,
+        lang=lang,
+    )
+
+    # 启动 Gradio 应用
+    demo.launch(
+        share=True,
+        server_port=args.server_port,
+        server_name=args.server_name,
+        inbrowser=True,
+        allowed_paths=[output_dir],
+        max_file_size="1gb",
+    )
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="轻量级视频生成")
+    parser.add_argument("--model_path", type=str, required=True, help="模型文件夹路径")
+    parser.add_argument("--server_port", type=int, default=7862, help="服务器端口")
+    parser.add_argument("--server_name", type=str, default="0.0.0.0", help="服务器IP")
+    parser.add_argument("--output_dir", type=str, default="./outputs", help="输出视频保存目录")
+    parser.add_argument("--lang", type=str, default="zh", choices=["zh", "en"], help="界面语言")
+    args = parser.parse_args()
+
+    global model_path, model_cls, output_dir
+    model_path = args.model_path
+    model_cls = "wan2.1"
+    output_dir = args.output_dir
+
+    main(lang=args.lang)

app/run_gradio.sh

+#!/bin/bash
+
+# Lightx2v Gradio Demo Startup Script
+# Supports both Image-to-Video (i2v) and Text-to-Video (t2v) modes
+
+# ==================== Configuration Area ====================
+# ⚠️  Important: Please modify the following paths according to your actual environment
+
+# 🚨 Storage Performance Tips 🚨
+# 💾 Strongly recommend storing model files on SSD solid-state drives!
+# 📈 SSD can significantly improve model loading speed and inference performance
+# 🐌 Using mechanical hard drives (HDD) may cause slow model loading and affect overall experience
+
+
+# Lightx2v project root directory path
+# Example: /home/user/lightx2v or /data/video_gen/lightx2v
+lightx2v_path=/data/video_gen/lightx2v_debug/LightX2V
+
+# Model path configuration
+# Example: /path/to/Wan2.1-I2V-14B-720P-Lightx2v
+model_path=/models/
+
+# Server configuration
+server_name="0.0.0.0"
+server_port=8036
+
+# Output directory configuration
+output_dir="./outputs"
+
+# GPU configuration
+gpu_id=0
+
+# ==================== Environment Variables Setup ====================
+export CUDA_VISIBLE_DEVICES=$gpu_id
+export CUDA_LAUNCH_BLOCKING=1
+export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
+export PROFILING_DEBUG_LEVEL=2
+export PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True
+
+# ==================== Parameter Parsing ====================
+# Default interface language
+lang="zh"
+
+# 解析命令行参数
+while [[ $# -gt 0 ]]; do
+    case $1 in
+        --lang)
+            lang="$2"
+            shift 2
+            ;;
+        --port)
+            server_port="$2"
+            shift 2
+            ;;
+        --gpu)
+            gpu_id="$2"
+            export CUDA_VISIBLE_DEVICES=$gpu_id
+            shift 2
+            ;;
+        --output_dir)
+            output_dir="$2"
+            shift 2
+            ;;
+        --model_path)
+            model_path="$2"
+            shift 2
+            ;;
+        --help)
+            echo "🎬 Lightx2v Gradio Demo Startup Script"
+            echo "=========================================="
+            echo "Usage: $0 [options]"
+            echo ""
+            echo "📋 Available options:"
+            echo "  --lang zh|en      Interface language (default: zh)"
+            echo "                     zh: Chinese interface"
+            echo "                     en: English interface"
+            echo "  --port PORT       Server port (default: 8032)"
+            echo "  --gpu GPU_ID      GPU device ID (default: 0)"
+            echo "  --model_path PATH Model path (default: configured in script)"
+            echo "  --output_dir DIR  Output video save directory (default: ./outputs)"
+            echo "  --help            Show this help message"
+            echo ""
+            echo "📝 Notes:"
+            echo "  - Task type (i2v/t2v) and model type are selected in the web UI"
+            echo "  - Model class is auto-detected based on selected diffusion model"
+            echo "  - Edit script to configure model paths before first use"
+            echo "  - Ensure required Python dependencies are installed"
+            echo "  - Recommended to use GPU with 8GB+ VRAM"
+            echo "  - 🚨 Strongly recommend storing models on SSD for better performance"
+            exit 0
+            ;;
+        *)
+            echo "Unknown parameter: $1"
+            echo "Use --help to see help information"
+            exit 1
+            ;;
+    esac
+done
+
+# ==================== Parameter Validation ====================
+if [[ "$lang" != "zh" && "$lang" != "en" ]]; then
+    echo "Error: Language must be 'zh' or 'en'"
+    exit 1
+fi
+
+# Check if model path exists
+if [[ ! -d "$model_path" ]]; then
+    echo "❌ Error: Model path does not exist"
+    echo "📁 Path: $model_path"
+    echo "🔧 Solutions:"
+    echo "  1. Check model path configuration in script"
+    echo "  2. Ensure model files are properly downloaded"
+    echo "  3. Verify path permissions are correct"
+    echo "  4. 💾 Recommend storing models on SSD for faster loading"
+    exit 1
+fi
+
+# 使用新的统一入口文件
+demo_file="gradio_demo.py"
+echo "🌏 Using unified interface (language: $lang)"
+
+# Check if demo file exists
+if [[ ! -f "$demo_file" ]]; then
+    echo "❌ Error: Demo file does not exist"
+    echo "📄 File: $demo_file"
+    echo "🔧 Solutions:"
+    echo "  1. Ensure script is run in the correct directory"
+    echo "  2. Check if file has been renamed or moved"
+    echo "  3. Re-clone or download project files"
+    exit 1
+fi
+
+# ==================== System Information Display ====================
+echo "=========================================="
+echo "🚀 Lightx2v Gradio Demo Starting..."
+echo "=========================================="
+echo "📁 Project path: $lightx2v_path"
+echo "🤖 Model path: $model_path"
+echo "🌏 Interface language: $lang"
+echo "🖥️  GPU device: $gpu_id"
+echo "🌐 Server address: $server_name:$server_port"
+echo "📁 Output directory: $output_dir"
+echo "📝 Note: Task type and model class are selected in web UI"
+echo "=========================================="
+
+# Display system resource information
+echo "💻 System resource information:"
+free -h | grep -E "Mem|Swap"
+echo ""
+
+# Display GPU information
+if command -v nvidia-smi &> /dev/null; then
+    echo "🎮 GPU information:"
+    nvidia-smi --query-gpu=name,memory.total,memory.free --format=csv,noheader,nounits | head -1
+    echo ""
+fi
+
+# ==================== Start Demo ====================
+echo "🎬 Starting Gradio demo..."
+echo "📱 Please access in browser: http://$server_name:$server_port"
+echo "⏹️  Press Ctrl+C to stop service"
+echo "🔄 First startup may take several minutes to load resources..."
+echo "=========================================="
+
+# Start Python demo
+if [[ "$demo_file" == "gradio_demo.py" ]]; then
+    python $demo_file \
+        --model_path "$model_path" \
+        --server_name "$server_name" \
+        --server_port "$server_port" \
+        --output_dir "$output_dir" \
+        --lang "$lang"
+else
+python $demo_file \
+    --model_path "$model_path" \
+    --server_name "$server_name" \
+    --server_port "$server_port" \
+    --output_dir "$output_dir"
+fi
+
+# Display final system resource usage
+echo ""
+echo "=========================================="
+echo "📊 Final system resource usage:"
+free -h | grep -E "Mem|Swap"

app/run_gradio_win.bat

+@echo off
+chcp 65001 >nul
+echo 🎬 LightX2V Gradio Windows Startup Script
+echo ==========================================
+
+REM ==================== Configuration Area ====================
+REM ⚠️  Important: Please modify the following paths according to your actual environment
+
+REM 🚨 Storage Performance Tips 🚨
+REM 💾 Strongly recommend storing model files on SSD solid-state drives!
+REM 📈 SSD can significantly improve model loading speed and inference performance
+REM 🐌 Using mechanical hard drives (HDD) may cause slow model loading and affect overall experience
+
+REM LightX2V project root directory path
+REM Example: D:\LightX2V
+set lightx2v_path=/path/to/LightX2V
+
+REM Model path configuration
+REM Model root directory path
+REM Example: D:\models\LightX2V
+set model_path=/path/to/LightX2V
+
+REM Server configuration
+set server_name=127.0.0.1
+set server_port=8032
+
+REM Output directory configuration
+set output_dir=./outputs
+
+REM GPU configuration
+set gpu_id=0
+
+REM ==================== Environment Variables Setup ====================
+set CUDA_VISIBLE_DEVICES=%gpu_id%
+set PYTHONPATH=%lightx2v_path%;%PYTHONPATH%
+set PROFILING_DEBUG_LEVEL=2
+set PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True
+
+REM ==================== Parameter Parsing ====================
+REM Default interface language
+set lang=zh
+
+REM Parse command line arguments
+:parse_args
+if "%1"=="" goto :end_parse
+if "%1"=="--lang" (
+    set lang=%2
+    shift
+    shift
+    goto :parse_args
+)
+if "%1"=="--port" (
+    set server_port=%2
+    shift
+    shift
+    goto :parse_args
+)
+if "%1"=="--gpu" (
+    set gpu_id=%2
+    set CUDA_VISIBLE_DEVICES=%gpu_id%
+    shift
+    shift
+    goto :parse_args
+)
+if "%1"=="--output_dir" (
+    set output_dir=%2
+    shift
+    shift
+    goto :parse_args
+)
+if "%1"=="--help" (
+    echo 🎬 LightX2V Gradio Windows Startup Script
+    echo ==========================================
+    echo Usage: %0 [options]
+    echo.
+    echo 📋 Available options:
+    echo   --lang zh^|en      Interface language (default: zh)
+    echo                      zh: Chinese interface
+    echo                      en: English interface
+    echo   --port PORT        Server port (default: 8032)
+    echo   --gpu GPU_ID       GPU device ID (default: 0)
+    echo   --output_dir OUTPUT_DIR
+    echo                      Output video save directory (default: ./outputs)
+    echo   --help             Show this help message
+    echo.
+    echo 🚀 Usage examples:
+    echo   %0                                    # Default startup
+    echo   %0 --lang zh --port 8032              # Start with specified parameters
+    echo   %0 --lang en --port 7860              # English interface
+    echo   %0 --gpu 1 --port 8032                # Use GPU 1
+    echo   %0 --output_dir ./custom_output       # Use custom output directory
+    echo.
+    echo 📝 Notes:
+    echo   - Edit script to configure model path before first use
+    echo   - Ensure required Python dependencies are installed
+    echo   - Recommended to use GPU with 8GB+ VRAM
+    echo   - 🚨 Strongly recommend storing models on SSD for better performance
+    pause
+    exit /b 0
+)
+echo Unknown parameter: %1
+echo Use --help to see help information
+pause
+exit /b 1
+
+:end_parse
+
+REM ==================== Parameter Validation ====================
+if "%lang%"=="zh" goto :valid_lang
+if "%lang%"=="en" goto :valid_lang
+echo Error: Language must be 'zh' or 'en'
+pause
+exit /b 1
+
+:valid_lang
+
+REM Check if model path exists
+if not exist "%model_path%" (
+    echo ❌ Error: Model path does not exist
+    echo 📁 Path: %model_path%
+    echo 🔧 Solutions:
+    echo   1. Check model path configuration in script
+    echo   2. Ensure model files are properly downloaded
+    echo   3. Verify path permissions are correct
+    echo   4. 💾 Recommend storing models on SSD for faster loading
+    pause
+    exit /b 1
+)
+
+REM Select demo file based on language
+if "%lang%"=="zh" (
+    set demo_file=gradio_demo_zh.py
+    echo 🌏 Using Chinese interface
+) else (
+    set demo_file=gradio_demo.py
+    echo 🌏 Using English interface
+)
+
+REM Check if demo file exists
+if not exist "%demo_file%" (
+    echo ❌ Error: Demo file does not exist
+    echo 📄 File: %demo_file%
+    echo 🔧 Solutions:
+    echo   1. Ensure script is run in the correct directory
+    echo   2. Check if file has been renamed or moved
+    echo   3. Re-clone or download project files
+    pause
+    exit /b 1
+)
+
+REM ==================== System Information Display ====================
+echo ==========================================
+echo 🚀 LightX2V Gradio Starting...
+echo ==========================================
+echo 📁 Project path: %lightx2v_path%
+echo 🤖 Model path: %model_path%
+echo 🌏 Interface language: %lang%
+echo 🖥️  GPU device: %gpu_id%
+echo 🌐 Server address: %server_name%:%server_port%
+echo 📁 Output directory: %output_dir%
+echo ==========================================
+
+REM Display system resource information
+echo 💻 System resource information:
+wmic OS get TotalVisibleMemorySize,FreePhysicalMemory /format:table
+
+REM Display GPU information
+nvidia-smi --query-gpu=name,memory.total,memory.free --format=csv,noheader,nounits 2>nul
+if errorlevel 1 (
+    echo 🎮 GPU information: Unable to get GPU info
+) else (
+    echo 🎮 GPU information:
+    nvidia-smi --query-gpu=name,memory.total,memory.free --format=csv,noheader,nounits
+)
+
+REM ==================== Start Demo ====================
+echo 🎬 Starting Gradio demo...
+echo 📱 Please access in browser: http://%server_name%:%server_port%
+echo ⏹️  Press Ctrl+C to stop service
+echo 🔄 First startup may take several minutes to load resources...
+echo ==========================================
+
+REM Start Python demo
+python %demo_file% ^
+    --model_path "%model_path%" ^
+    --server_name %server_name% ^
+    --server_port %server_port% ^
+    --output_dir "%output_dir%"
+
+REM Display final system resource usage
+echo.
+echo ==========================================
+echo 📊 Final system resource usage:
+wmic OS get TotalVisibleMemorySize,FreePhysicalMemory /format:table
+
+pause

app/utils/i18n.py

+"""国际化支持模块"""
+
+import os
+
+# 默认语言
+DEFAULT_LANG = os.getenv("GRADIO_LANG", "zh")
+
+# 翻译字典
+TRANSLATIONS = {
+    "zh": {
+        "title": "🎬 LightX2V 图片/视频生成器",
+        "model_config": "🗂️ 模型配置",
+        "model_config_hint": "💡 **提示**：请确保以下每个模型选项至少有一个已下载✅的模型可用，否则可能无法正常生成视频。",
+        "fp8_not_supported": "⚠️ **您的设备不支持fp8推理**，已自动隐藏包含fp8的模型选项。",
+        "model_type": "模型类型",
+        "model_type_info": "Wan2.2 需要分别指定高噪模型和低噪模型; Qwen-Image-Edit-2511 用于图片编辑(i2i); Qwen-Image-2512 用于文本生成图片(t2i); Z-Image-Turbo 用于文本生成图片(t2i)",
+        "qwen3_encoder": "📝 Qwen3 编码器",
+        "scheduler": "⏱️ 调度器",
+        "qwen25vl_encoder": "📝 Qwen25-VL 编码器",
+        "task_type": "任务类型",
+        "task_type_info": "I2V: 图生视频, T2V: 文生视频, T2I: 文生图, I2I: 图片编辑",
+        "download_source": "📥 下载源",
+        "download_source_info": "选择模型下载源",
+        "diffusion_model": "🎨 Diffusion模型",
+        "high_noise_model": "🔊 高噪模型",
+        "low_noise_model": "🔇 低噪模型",
+        "text_encoder": "📝 文本编码器",
+        "text_encoder_tokenizer": "📝 文本编码器 Tokenizer",
+        "image_encoder": "🖼️ 图像编码器",
+        "image_encoder_tokenizer": "🖼️ 图像编码器 Tokenizer",
+        "vae": "🎞️ VAE编码/解码器",
+        "attention_operator": "⚡ 注意力算子",
+        "attention_operator_info": "使用适当的注意力算子加速推理",
+        "quant_operator": "⚡矩阵乘法算子",
+        "quant_operator_info": "选择低精度矩阵乘法算子以加速推理",
+        "input_params": "📥 输入参数",
+        "input_image": "输入图像（可拖入多张图片）",
+        "image_preview": "已上传的图片预览",
+        "image_path": "图片路径",
+        "prompt": "提示词",
+        "prompt_placeholder": "描述视频/图片内容...",
+        "negative_prompt": "负向提示词",
+        "negative_prompt_placeholder": "不希望出现在视频/图片中的内容...",
+        "max_resolution": "最大分辨率",
+        "max_resolution_info": "如果显存不足，可调低分辨率",
+        "random_seed": "随机种子",
+        "infer_steps": "推理步数",
+        "infer_steps_distill": "蒸馏模型推理步数默认为4。",
+        "infer_steps_info": "视频生成的推理步数。增加步数可能提高质量但降低速度。",
+        "sample_shift": "分布偏移",
+        "sample_shift_info": "控制样本分布偏移的程度。值越大表示偏移越明显。",
+        "cfg_scale": "CFG缩放因子",
+        "cfg_scale_info": "控制提示词的影响强度。值越高，提示词的影响越大。当值为1时，自动禁用CFG。",
+        "enable_cfg": "启用无分类器引导",
+        "fps": "每秒帧数(FPS)",
+        "fps_info": "视频的每秒帧数。较高的FPS会产生更流畅的视频。",
+        "num_frames": "总帧数",
+        "num_frames_info": "视频中的总帧数。更多帧数会产生更长的视频。",
+        "video_duration": "视频时长（秒）",
+        "video_duration_info": "视频的时长（秒）。实际帧数 = 时长 × FPS。",
+        "output_path": "输出视频路径",
+        "output_path_info": "必须包含.mp4扩展名。如果留空或使用默认值，将自动生成唯一文件名。",
+        "output_image_path": "输出图片路径",
+        "output_image_path_info": "必须包含.png扩展名。如果留空或使用默认值，将自动生成唯一文件名。",
+        "output_result": "📤 生成的结果",
+        "output_image": "输出图片",
+        "generate_video": "🎬 生成视频",
+        "generate_image": "🖼️ 生成图片",
+        "infer_steps_image_info": "图片编辑的推理步数，默认为8。",
+        "aspect_ratio": "宽高比",
+        "aspect_ratio_info": "选择生成图片的宽高比",
+        "model_config_hint_image": "💡 **提示**：请确保以下每个模型选项至少有一个已下载✅的模型可用，否则可能无法正常生成图片。",
+        "download": "📥 下载",
+        "downloaded": "✅ 已下载",
+        "not_downloaded": "❌ 未下载",
+        "download_complete": "✅ {model_name} 下载完成",
+        "download_start": "开始从 {source} 下载 {model_name}...",
+        "please_select_model": "请先选择模型",
+        "loading_models": "正在加载 Hugging Face 模型列表缓存...",
+        "models_loaded": "模型列表缓存加载完成",
+        "use_lora": "使用 LoRA",
+        "lora": "🎨 LoRA",
+        "lora_info": "选择要使用的 LoRA 模型",
+        "lora_strength": "LoRA 强度",
+        "lora_strength_info": "控制 LoRA 的影响强度，范围 0-10",
+        "high_noise_lora": "🔊 高噪模型 LoRA",
+        "high_noise_lora_info": "选择高噪模型使用的 LoRA",
+        "high_noise_lora_strength": "高噪模型 LoRA 强度",
+        "high_noise_lora_strength_info": "控制高噪模型 LoRA 的影响强度，范围 0-10",
+        "low_noise_lora": "🔇 低噪模型 LoRA",
+        "low_noise_lora_info": "选择低噪模型使用的 LoRA",
+        "low_noise_lora_strength": "低噪模型 LoRA 强度",
+        "low_noise_lora_strength_info": "控制低噪模型 LoRA 的影响强度，范围 0-10",
+    },
+    "en": {
+        "title": "🎬 LightX2V Image/Video Generator",
+        "model_config": "🗂️ Model Configuration",
+        "model_config_hint": "💡 **Tip**: Please ensure at least one downloaded ✅ model is available for each model option below, otherwise video generation may fail.",
+        "fp8_not_supported": "⚠️ **Your device does not support fp8 inference**, fp8 model options have been automatically hidden.",
+        "model_type": "Model Type",
+        "model_type_info": "Wan2.2 requires separate high-noise and low-noise models; Qwen-Image-Edit-2511 is for image editing (i2i); Qwen-Image-2512 is for text-to-image (t2i); Z-Image-Turbo is for text-to-image (t2i)",
+        "qwen3_encoder": "📝 Qwen3 Encoder",
+        "scheduler": "⏱️ Scheduler",
+        "qwen25vl_encoder": "📝 Qwen25-VL Encoder",
+        "task_type": "Task Type",
+        "task_type_info": "I2V: Image-to-Video, T2V: Text-to-Video, T2I: Text-to-Image, I2I: Image Editing",
+        "download_source": "📥 Download Source",
+        "download_source_info": "Select model download source",
+        "diffusion_model": "🎨 Diffusion Model",
+        "high_noise_model": "🔊 High Noise Model",
+        "low_noise_model": "🔇 Low Noise Model",
+        "text_encoder": "📝 Text Encoder",
+        "text_encoder_tokenizer": "📝 Text Encoder Tokenizer",
+        "image_encoder": "🖼️ Image Encoder",
+        "image_encoder_tokenizer": "🖼️ Image Encoder Tokenizer",
+        "vae": "🎞️ VAE Encoder/Decoder",
+        "attention_operator": "⚡ Attention Operator",
+        "attention_operator_info": "Use appropriate attention operator to accelerate inference",
+        "quant_operator": "⚡ Matrix Multiplication Operator",
+        "quant_operator_info": "Select low-precision matrix multiplication operator to accelerate inference",
+        "input_params": "📥 Input Parameters",
+        "input_image": "Input Image (drag multiple images)",
+        "image_preview": "Uploaded Image Preview",
+        "image_path": "Image Path",
+        "prompt": "Prompt",
+        "prompt_placeholder": "Describe video/image content...",
+        "negative_prompt": "Negative Prompt",
+        "negative_prompt_placeholder": "Content you don't want in the video/image...",
+        "max_resolution": "Max Resolution",
+        "max_resolution_info": "Reduce resolution if VRAM is insufficient",
+        "random_seed": "Random Seed",
+        "infer_steps": "Inference Steps",
+        "infer_steps_distill": "Distill model inference steps default to 4.",
+        "infer_steps_info": "Number of inference steps for video generation. More steps may improve quality but reduce speed.",
+        "sample_shift": "Sample Shift",
+        "sample_shift_info": "Control the degree of sample distribution shift. Higher values indicate more obvious shift.",
+        "cfg_scale": "CFG Scale",
+        "cfg_scale_info": "Control the influence strength of prompts. Higher values mean stronger prompt influence. When value is 1, CFG is automatically disabled.",
+        "enable_cfg": "Enable Classifier-Free Guidance",
+        "fps": "Frames Per Second (FPS)",
+        "fps_info": "Frames per second of the video. Higher FPS produces smoother videos.",
+        "num_frames": "Total Frames",
+        "num_frames_info": "Total number of frames in the video. More frames produce longer videos.",
+        "video_duration": "Video Duration (seconds)",
+        "video_duration_info": "Duration of the video in seconds. Actual frames = duration × FPS.",
+        "output_path": "Output Video Path",
+        "output_path_info": "Must include .mp4 extension. If left empty or using default value, a unique filename will be automatically generated.",
+        "output_image_path": "Output Image Path",
+        "output_image_path_info": "Must include .png extension. If left empty or using default value, a unique filename will be automatically generated.",
+        "output_result": "📤 Generated Result",
+        "output_image": "Output Image",
+        "generate_video": "🎬 Generate Video",
+        "generate_image": "🖼️ Generate Image",
+        "infer_steps_image_info": "Number of inference steps for image editing, default is 8.",
+        "aspect_ratio": "Aspect Ratio",
+        "aspect_ratio_info": "Select the aspect ratio for generated images",
+        "model_config_hint_image": "💡 **Tip**: Please ensure at least one downloaded ✅ model is available for each model option below, otherwise image generation may fail.",
+        "download": "📥 Download",
+        "downloaded": "✅ Downloaded",
+        "not_downloaded": "❌ Not Downloaded",
+        "download_complete": "✅ {model_name} download complete",
+        "download_start": "Starting to download {model_name} from {source}...",
+        "please_select_model": "Please select a model first",
+        "loading_models": "Loading Hugging Face model list cache...",
+        "models_loaded": "Model list cache loaded",
+        "use_lora": "Use LoRA",
+        "lora": "🎨 LoRA",
+        "lora_info": "Select LoRA model to use",
+        "lora_strength": "LoRA Strength",
+        "lora_strength_info": "Control LoRA influence strength, range 0-10",
+        "high_noise_lora": "🔊 High Noise Model LoRA",
+        "high_noise_lora_info": "Select high noise model LoRA to use",
+        "high_noise_lora_strength": "High Noise Model LoRA Strength",
+        "high_noise_lora_strength_info": "Control high noise model LoRA influence strength, range 0-10",
+        "low_noise_lora": "🔇 Low Noise Model LoRA",
+        "low_noise_lora_info": "Select low noise model LoRA to use",
+        "low_noise_lora_strength": "Low Noise Model LoRA Strength",
+        "low_noise_lora_strength_info": "Control low noise model LoRA influence strength, range 0-10",
+    },
+}
+
+
+def t(key: str, lang: str = None) -> str:
+    """获取翻译文本"""
+    if lang is None:
+        lang = DEFAULT_LANG
+
+    if lang not in TRANSLATIONS:
+        lang = "zh"
+
+    return TRANSLATIONS[lang].get(key, key)
+
+
+def set_language(lang: str):
+    """设置语言"""
+    global DEFAULT_LANG
+    if lang in TRANSLATIONS:
+        DEFAULT_LANG = lang
+        os.environ["GRADIO_LANG"] = lang