WorldFoundry
English中文

Video · image-to-video · text-to-video · video-extension

Pusa VidGen

RaphaelLiu

WorldFoundry uses the in-tree Pusa V1 Wan2.2 runner and local DiffSynth runtime; no external source checkout is required at inference time.

IntegratedUnified environmentpusa-vidgen

Build your run

Choose a recorded variant, then copy the exact setup or inspection command.

worldfoundry.pipeline
Taskimage-to-video
Environmentworldfoundry-unified-cu128
DeviceCUDA 12.8
worldfoundry-eval evaluate \
  --mode model \
  --model-id pusa-vidgen \
  --model-runner worldfoundry:pipeline \
  --model-manifest-dir worldfoundry/data/models/catalog \
  --requests-path tmp/requests.jsonl \
  --output-dir tmp/model_eval/pusa-vidgen \
  --metric artifact_count \
  --json
01

Compatibility & versions

Manifest-backed recipe

Integration
integrated
Runner evidence
configured
Environment
worldfoundry-unified-cu128
Python
3.11
CUDA
CUDA 12.8
PyTorch
torch
Source revision
—
Checkpoint revision
—
Runtime profile
pusa-vidgen
Pipeline binding
pusa-vidgen
Runner
worldfoundry.pipeline
Pipeline target
worldfoundry.pipelines.pusa_vidgen.pipeline_pusa_vidgen:PusaVidGenPipeline
Backend stage
in_tree_runtime
Runtime status
integrated
Driver status
compatible_unified_default
Environment kind
Unified environment
02

Install environment

The environment resolver reads the recorded profile and chooses the unified or dedicated environment shown here.

bash scripts/setup/model_env_install.sh --model pusa-vidgen
Package constraints 4
  • torch
  • torchvision
  • torchaudio
  • xfuser
Conda packages 5
  • python=3.11
  • pip
  • ffmpeg
  • setuptools
  • wheel
03

Checkpoints & assets

Run the local check before allocating compute. Gated, private, and license fields below come directly from the checkpoint manifest.

worldfoundry-eval zoo model-download --model-id pusa-vidgen --check-local --json
RaphaelLiu/Pusa-Wan2.2-V1
Revision
—
License
—
Gated
—
Private
—
Wan-AI/Wan2.2-T2V-A14B
Revision
—
License
—
Gated
—
Private
—
lightx2v/Wan2.2-Lightning
Revision
—
License
—
Gated
—
Private
—
04

Launch & outputs

The generated command uses the shared evaluation boundary and writes durable result manifests and artifacts.

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

Input contract

FieldRecorded contract
imageRequired
promptRequired
videoRequired

Artifact contract

Artifact kindFilename / path
generated_video—
generated_videopusa_vidgen_public_runner.mp4
05

Evidence & sources

Catalog integration, native-demo parity, and runner parity are independent records.

Integration
integrated
Runner evidence
configured
Native demo evidence
Not recorded
Validation imports
torch, torchvision, numpy, transformers, diffusers, xfuser, worldfoundry
Recipe provenance
worldfoundry/data/models/catalog/video/pusa-vidgen.yamlProjecthttps://pusav1.org/GitHubhttps://github.com/Yaofang-Liu/Pusa-VidGenHugging Facehttps://huggingface.co/RaphaelLiu/Pusa-Wan2.2-V1