Unverified Commit f21da849 authored by Yang Yong (雍洋)'s avatar Yang Yong (雍洋) Committed by GitHub
Browse files
parent 3efc43f5
...@@ -8,7 +8,7 @@ ...@@ -8,7 +8,7 @@
[![Doc](https://img.shields.io/badge/docs-English-99cc2)](https://lightx2v-en.readthedocs.io/en/latest) [![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) [![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) [![Papers](https://img.shields.io/badge/论文集-中文-99cc2)](https://lightx2v-papers-zhcn.readthedocs.io/zh-cn/latest)
[![Docker](https://badgen.net/badge/icon/docker?icon=docker&label)](https://hub.docker.com/r/lightx2v/lightx2v/tags) [![Docker](https://img.shields.io/badge/Docker-2496ED?style=flat&logo=docker&logoColor=white)](https://hub.docker.com/r/lightx2v/lightx2v/tags)
**\[ English | [中文](README_zh.md) \]** **\[ English | [中文](README_zh.md) \]**
...@@ -18,6 +18,10 @@ ...@@ -18,6 +18,10 @@
**LightX2V** is an advanced lightweight video generation inference framework engineered to deliver efficient, high-performance video synthesis solutions. This unified platform integrates multiple state-of-the-art video generation techniques, supporting diverse generation tasks including text-to-video (T2V) and image-to-video (I2V). **X2V represents the transformation of different input modalities (X, such as text or images) into video output (V)**. **LightX2V** is an advanced lightweight video generation inference framework engineered to deliver efficient, high-performance video synthesis solutions. This unified platform integrates multiple state-of-the-art video generation techniques, supporting diverse generation tasks including text-to-video (T2V) and image-to-video (I2V). **X2V represents the transformation of different input modalities (X, such as text or images) into video output (V)**.
## :fire: Latest News
- **November 21, 2025:** 🚀 We support the HunyuanVideo1.5 video generation model since Day 0. With the same number of GPUs, LightX2V can deliver a speed improvement of over 2 times and supports deployment on GPUs with lower memory (such as the 24GB RTX 4090). It also supports CFG/Ulysses parallelism, efficient offloading, TeaCache/MagCache technologies, and more. We will soon update our models on our [HuggingFace page](https://huggingface.co/lightx2v), including quantization, step distillation, VAE distillation, and other related models.
## 💡 Quick Start ## 💡 Quick Start
For comprehensive usage instructions, please refer to our documentation: **[English Docs](https://lightx2v-en.readthedocs.io/en/latest/) | [中文文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/)** For comprehensive usage instructions, please refer to our documentation: **[English Docs](https://lightx2v-en.readthedocs.io/en/latest/) | [中文文档](https://lightx2v-zhcn.readthedocs.io/zh-cn/latest/)**
......
...@@ -8,7 +8,7 @@ ...@@ -8,7 +8,7 @@
[![Doc](https://img.shields.io/badge/docs-English-99cc2)](https://lightx2v-en.readthedocs.io/en/latest) [![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) [![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) [![Papers](https://img.shields.io/badge/论文集-中文-99cc2)](https://lightx2v-papers-zhcn.readthedocs.io/zh-cn/latest)
[![Docker](https://badgen.net/badge/icon/docker?icon=docker&label)](https://hub.docker.com/r/lightx2v/lightx2v/tags) [![Docker](https://img.shields.io/badge/Docker-2496ED?style=flat&logo=docker&logoColor=white)](https://hub.docker.com/r/lightx2v/lightx2v/tags)
**\[ [English](README.md) | 中文 \]** **\[ [English](README.md) | 中文 \]**
...@@ -18,6 +18,11 @@ ...@@ -18,6 +18,11 @@
**LightX2V** 是一个先进的轻量级视频生成推理框架,专为提供高效、高性能的视频合成解决方案而设计。该统一平台集成了多种前沿的视频生成技术,支持文本生成视频(T2V)和图像生成视频(I2V)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)转换为视频输出(V)** **LightX2V** 是一个先进的轻量级视频生成推理框架,专为提供高效、高性能的视频合成解决方案而设计。该统一平台集成了多种前沿的视频生成技术,支持文本生成视频(T2V)和图像生成视频(I2V)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)转换为视频输出(V)**
## :fire: 最新动态
- **2025年11月21日:** 🚀 我们Day0支持了HunyuanVideo1.5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)更新量化,步数蒸馏,VAE蒸馏等相关模型。
## 💡 快速开始 ## 💡 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/)** 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/)**
......
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn2",
"cpu_offload": true,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 7.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn2",
"parallel": {
"seq_p_attn_type": "ulysses-4090",
"seq_p_size": 8
}
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn2",
"cpu_offload": true,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false,
"dit_quantized_ckpt": "/path/to/480p_t2v_fp8.safetensors",
"dit_quantized": true,
"dit_quant_scheme": "fp8-q8f",
"qwen25vl_quantized_ckpt": "/path/to/qwen25vl_fp8.safetensors",
"qwen25vl_quantized": true,
"qwen25vl_quant_scheme": "fp8-q8f"
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn3",
"cpu_offload": true,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn3",
"cpu_offload": true,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false,
"parallel": {
"seq_p_attn_type": "ulysses",
"seq_p_size": 8
}
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn3",
"qwen25vl_cpu_offload": true,
"dit_quantized_ckpt": "/path/to/480p_t2v_fp8.safetensors",
"dit_quantized": true,
"dit_quant_scheme": "fp8-sgl",
"qwen25vl_quantized_ckpt": "/path/to/qwen25vl_fp8.safetensors",
"qwen25vl_quantized": true,
"qwen25vl_quant_scheme": "fp8-sgl"
}
{
"infer_steps": 50,
"transformer_model_name": "480p_i2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn2",
"feature_caching": "Mag",
"magcache_calibration": false,
"magcache_K": 6,
"magcache_thresh": 0.24,
"magcache_retention_ratio": 0.2,
"magcache_ratios": [[1.0, 1.01562, 1.00781, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.99609, 1.0, 0.99609, 1.0, 0.99609, 1.0, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99219, 0.99219, 0.99219, 0.98828, 0.98828, 0.98828, 0.98828, 0.98438, 0.98438, 0.98047, 0.98047, 0.97656, 0.97266, 0.96484, 0.95703, 0.94922, 0.92969, 0.91016, 0.88672], [1.0, 1.02344, 1.00781, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.99609, 1.0, 0.99609, 1.0, 0.99609, 1.0, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99609, 0.99219, 0.99219, 0.99219, 0.99219, 0.98828, 0.98828, 0.98828, 0.98438, 0.98438, 0.98047, 0.98047, 0.97656, 0.97266, 0.96484, 0.95703, 0.94922, 0.93359, 0.91016, 0.88672]]
}
{
"infer_steps": 50,
"transformer_model_name": "480p_i2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "sage_attn2",
"feature_caching": "Mag",
"magcache_calibration": true
}
{
"infer_steps": 50,
"transformer_model_name": "480p_i2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3",
"feature_caching": "Tea",
"coefficients": [8.08528429e+03 ,-2.44607178e+03, 2.49489589e+02, -9.10697865e+00, 1.20261379e-01],
"teacache_thresh": 0.15,
"cpu_offload": false,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false
}
{
"infer_steps": 50,
"transformer_model_name": "480p_i2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 7.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3",
"feature_caching": "Tea",
"coefficients": [3.84300014e+03, -1.39247433e+03, 1.69167679e+02, -7.07679232e+00, 1.02419011e-01],
"teacache_thresh": 0.15,
"cpu_offload": false,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3",
"feature_caching": "Tea",
"coefficients": [-2.97190924e+04, 2.22834983e+04, -4.37418360e+03, 3.39340251e+02, -1.01365855e+01, 1.29101768e-01],
"teacache_thresh": 0.15,
"cpu_offload": false,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false
}
{
"infer_steps": 50,
"transformer_model_name": "729p_t2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3",
"feature_caching": "Tea",
"coefficients": [-3.08907507e+04, 1.67786188e+04, -3.19178643e+03, 2.60740519e+02, -8.19205881e+00, 1.07913775e-01],
"teacache_thresh": 0.15,
"cpu_offload": false,
"offload_granularity": "block",
"vae_cpu_offload": false,
"byt5_cpu_offload": false,
"qwen25vl_cpu_offload": true,
"siglip_cpu_offload": false
}
{
"infer_steps": 50,
"transformer_model_name": "480p_i2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3"
}
{
"infer_steps": 50,
"transformer_model_name": "720p_i2v",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 7.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3"
}
{
"infer_steps": 50,
"transformer_model_name": "720p_i2v_distilled",
"fps": 24,
"target_video_length": 121,
"vae_stride": [4, 16, 16],
"sample_shift": 9.0,
"sample_guide_scale": 6.0,
"enable_cfg": false,
"attn_type": "flash_attn3"
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 5.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3"
}
{
"infer_steps": 50,
"transformer_model_name": "720p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 9.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3"
}
{
"infer_steps": 50,
"transformer_model_name": "480p_t2v",
"fps": 24,
"target_video_length": 121,
"aspect_ratio": "16:9",
"vae_stride": [4, 16, 16],
"sample_shift": 7.0,
"sample_guide_scale": 6.0,
"enable_cfg": true,
"attn_type": "flash_attn3",
"use_tae": true,
"tae_path": "/path/to/lighttae"
}
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment