---
title: Latte
canonical_url: "https://www.modelscope.cn/models/AI-ModelScope/Latte"
md_url: "https://www.modelscope.cn/models/AI-ModelScope/Latte.md"
repository: AI-ModelScope/Latte
last_updated: 2024-02-22
license: mit
pipeline_tag: text-to-video-synthesis
tasks:
  - text-to-video-synthesis
parameters: 4.8B
tensor_type:
  - F32
library_name:
  - safetensors
  - diffusers
  - pytorch
frameworks:
  - Pytorch
downloads: 293
stars: 0
---

# Latte

> Latte - AI-ModelScope 在 ModelScope 开源的模型。Latte: Latent Diffusion Transformer for Video Generation (Latte)

AI-ModelScope/Latte 是 ModelScope 魔搭社区上的 4.8B 参数text-to-video-synthesis模型，采用 mit 许可。

- **Repository**: AI-ModelScope/Latte
- **License**: mit
- **Tasks**: text-to-video-synthesis
- **Parameters**: 4.8B
- **Downloads**: 293
- **Stars**: 0
- **Last updated**: 2024-02-22

Source: https://www.modelscope.cn/models/AI-ModelScope/Latte

---

Latte: Latent Diffusion Transformer for Video Generation (Latte)

Please refer to the project page of [Latte](https://maxin-cn.github.io/latte_project/).