# Vchitect/Latte

[TMLR 2025] Latte: Latent Diffusion Transformer for Video Generation.

Repository: https://github.com/Vchitect/Latte
Canonical: https://ross.abutalabs.com/products/latte
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-10T04:52:44+00:00

## Health v2 (maintenance only)
Score: 71/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 97, release rhythm 35, longevity 74
- inputs: {"age_days": 1040, "days_push": 23, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1948, forks 192 (observed 2026-08-28T04:05:57.973703+00:00)

## What it is
Official PyTorch implementation of Latte, a latent diffusion transformer for video generation. It includes model definitions, pre-trained checkpoints, and training/sampling/evaluation code for text-to-video synthesis.

## Use cases
- generate videos from text prompts
- train a diffusion transformer on video data
- sample videos from pre-trained checkpoints
- research video generation models
- evaluate video generation quality

## When to choose
- you need a research-grade text-to-video diffusion model with open weights
- you want to experiment with latent diffusion transformer architectures for video

## When to avoid
- you need production-ready, optimized video generation inference
- you lack GPU resources for large diffusion models

## Facets
- artifact type: library
- maturity: active
- function: deep-learning, video-processing, machine-learning
- domain: deep-learning
- platform: python
- tags: diffusion-transformer, text-to-video, video-generation, pytorch, research-code, video, gpu

## Member repositories
- Vchitect/Latte (main) score 71

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:57.973703+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:06:58.758060+00:00, confidence not recorded.
  - readme: https://github.com/Vchitect/Latte (fetched 2026-08-28T04:05:57.973703+00:00, sha d6273834bf53)
- Data as of 2026-08-30T08:39:29.467469+00:00.