Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
xuwx1
LightX2V
Commits
0e08595c
Unverified
Commit
0e08595c
authored
Dec 05, 2025
by
Gu Shiqiao
Committed by
GitHub
Dec 05, 2025
Browse files
Update README_zh.md (#572)
parent
5cc6356e
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
10 additions
and
9 deletions
+10
-9
README_zh.md
README_zh.md
+10
-9
No files found.
README_zh.md
View file @
0e08595c
...
...
@@ -20,7 +20,7 @@
## :fire: 最新动态
-
**2025年12月4日:**
🚀
我们
支持
了
GGUF格式模型推理
。我们支持了
寒武纪MLU590
和沐曦C500
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
)
。
...
...
@@ -31,7 +31,7 @@
### 📊 推理框架之间性能对比 (H100)
| Framework |
Card
s | Step Time | Speedup |
| Framework |
GPU
s | Step Time | Speedup |
|-----------|---------|---------|---------|
| Diffusers | 1 | 9.77s/it | 1x |
| xDiT | 1 | 8.93s/it | 1.1x |
...
...
@@ -45,7 +45,7 @@
### 📊 推理框架之间性能对比 (RTX 4090D)
| Framework |
Card
s | Step Time | Speedup |
| Framework |
GPU
s | Step Time | Speedup |
|-----------|---------|---------|---------|
| Diffusers | 1 | 30.50s/it | 1x |
| FastVideo | 1 | 22.66s/it | 1.3x |
...
...
@@ -61,12 +61,12 @@
| Framework | GPU | Configuration | Step Time | Speedup |
|-----------|-----|---------------|-----------|---------------|
|
**LightX2V**
| H100 | 8
card
s + cfg | 0.75s/it | 1x |
|
**LightX2V**
| H100 | 8
card
s + no cfg | 0.39s/it | 1.9x |
|
**LightX2V**
| H100 |
**8
card
s + no cfg + fp8**
|
**0.35s/it**
|
**2.1x**
🚀 |
|
**LightX2V**
| 4090D | 8
card
s + cfg | 4.75s/it | 1x |
|
**LightX2V**
| 4090D | 8
card
s + no cfg | 3.13s/it | 1.5x |
|
**LightX2V**
| 4090D |
**8
card
s + no cfg + fp8**
|
**2.35s/it**
|
**2.0x**
🚀 |
|
**LightX2V**
| H100 | 8
GPU
s + cfg | 0.75s/it | 1x |
|
**LightX2V**
| H100 | 8
GPU
s + no cfg | 0.39s/it | 1.9x |
|
**LightX2V**
| H100 |
**8
GPU
s + no cfg + fp8**
|
**0.35s/it**
|
**2.1x**
🚀 |
|
**LightX2V**
| 4090D | 8
GPU
s + cfg | 4.75s/it | 1x |
|
**LightX2V**
| 4090D | 8
GPU
s + no cfg | 3.13s/it | 1.5x |
|
**LightX2V**
| 4090D |
**8
GPU
s + 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。
...
...
@@ -292,3 +292,4 @@ pipe.generate(
<div
align=
"center"
>
由 LightX2V 团队用 ❤️ 构建
</div>
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment