Skip to content

misc

Modules

fastvideo.train.entrypoint.misc.wan_ode_init_conversion

Convert Self-Forcing ode_init.pt to HuggingFace diffusers format.

The official ode_init.pt from https://huggingface.co/gdhe17/Self-Forcing/resolve/main/checkpoints/ode_init.pt stores weights under {"generator": {<original_wan_keys>}}.

This script converts those keys to diffusers WanTransformer3DModel format, verifies them against a reference model, and saves a complete diffusers-compatible model directory (transformer + scheduler + vae + text_encoder + tokenizer).

Usage

python -m fastvideo.train.entrypoint.misc.wan_ode_init_conversion --input /path/to/ode_init.pt --output /path/to/WanOdeInit --base-model Wan-AI/Wan2.1-T2V-1.3B-Diffusers

Functions

fastvideo.train.entrypoint.misc.wan_ode_init_conversion.convert_state_dict 
convert_state_dict(orig_sd: dict[str, Tensor]) -> dict[str, Tensor]

Convert an entire original-Wan state dict.

Source code in fastvideo/train/entrypoint/misc/wan_ode_init_conversion.py 
85
86
87
def convert_state_dict(orig_sd: dict[str, torch.Tensor], ) -> dict[str, torch.Tensor]:
    """Convert an entire original-Wan state dict."""
    return {_convert_key(k): v for k, v in orig_sd.items()}