Video · text-to-video · legacy-unet-diffusion

AnimateDiff

guoyww

Verified against the official repo default RealisticVision scripts.animate demo; file SHA256, decoded pixel SHA256, and artifact size match exactly.

Runner verified统一环境animatediff

构建运行命令

选择仓库中已记录的 variant，然后复制对应的准备、检查或运行命令。

worldfoundry.pipeline

Variant

任务text-to-video

环境worldfoundry-unified-cu128

设备CUDA 12.8

worldfoundry-eval evaluate \
  --mode model \
  --model-id animatediff \
  --model-runner worldfoundry:pipeline \
  --model-manifest-dir worldfoundry/data/models/catalog \
  --requests-path tmp/requests.jsonl \
  --output-dir tmp/model_eval/animatediff \
  --metric artifact_count \
  --json

兼容性与版本

由 Manifest 生成

集成状态: integrated
Runner 证据: verified
环境: worldfoundry-unified-cu128
Python: 3.11
CUDA: CUDA 12.8
PyTorch: torch
源码 revision: —
Checkpoint revision: —

Runtime profile: animatediff
Pipeline binding: animatediff
Runner: worldfoundry.pipeline
Pipeline target: worldfoundry.pipelines.component_pipelines:AnimateDiffPipeline
Backend stage: in_tree_runtime
Runtime 状态: official_repo_animation_pipeline_ready
Driver 状态: compatible
环境类型: 统一环境

安装环境

环境解析器会读取已记录 profile，并选择此处显示的统一或独立环境。

bash scripts/setup/model_env_install.sh --model animatediff

依赖版本约束 20

torch
torchvision
diffusers
transformers
huggingface_hub
imageio
imageio-ffmpeg
numpy
decord
omegaconf
gradio
gradio_client
fastapi
pydantic
socksio
setuptools
safetensors
einops
wandb
accelerate

Conda 依赖 3

python
pip
ffmpeg

Checkpoint 与资产

分配算力前先做本地检查；gated、private 与 license 字段直接来自 checkpoint manifest。

worldfoundry-eval zoo model-download --model-id animatediff --check-local --json

guoyww/animatediff

Revision: —
License: apache-2.0
Gated: —
Private: —

guoyww/animatediff_t2i_backups

Revision: —
License: —
Gated: —
Private: —

stable-diffusion-v1-5/stable-diffusion-v1-5

Revision: —
License: —
Gated: —
Private: —

openai/clip-vit-large-patch14

Revision: —
License: —
Gated: —
Private: —

运行与输出

生成的命令通过共享 evaluation 边界运行，并持久保存结果 manifest 与 artifact。

worldfoundry-eval evaluate \
  --mode model \
  --model-id animatediff \
  --model-runner worldfoundry:pipeline \
  --model-manifest-dir worldfoundry/data/models/catalog \
  --requests-path tmp/requests.jsonl \
  --output-dir tmp/model_eval/animatediff \
  --metric artifact_count \
  --json

输入契约

字段	已记录契约
`image`	Optional
`prompt`	Required

Artifact 契约

Artifact 类型	文件名 / 路径
`generated_video`	`animatediff.gif`

证据与来源

Catalog 集成、原生 demo parity 与 runner parity 是三条独立记录。

集成状态: integrated
Runner 证据: verified
原生 Demo 证据: 未记录
验证 Imports: torch, diffusers, transformers, imageio, gradio

配方溯源

worldfoundry/data/models/catalog/video/animatediff.yamlGitHubhttps://github.com/guoyww/AnimateDiff

构建运行命令

输入契约

Artifact 契约

相关配方