README.md

metadata

pipeline_tag: text-to-video
tags:
  - omnimem
  - wan
  - video-generation
  - self-forcing

self_forcing_from_cd

OmniMem ordinary self-forcing checkpoint initialized from CD-trained generator weights.

This release contains the generator EMA weights at self-forcing training step 600. It is the initialization checkpoint used by long_rolling_plain_from_cd.

Checkpoint lineage

CD step-4500 EMA → ordinary self-forcing step-600 EMA → plain rolling training.

Contents

• config.json
• Two safetensors weight shards and their index
• SHA256SUMS for payload integrity checking

The weights retain their original FP32 storage. The model configuration uses NSA top-12 block selection.

Usage

Use compatible OmniMem code with CausalWanNSAModel and CausalWanT2VPipeline. The Wan2.1 VAE, UMT5 text encoder, and tokenizer are separate required components.

from omnimem.models.transformers.causal_wan_nsa_model import CausalWanNSAModel

model = CausalWanNSAModel.from_pretrained(
    "omnimem/self_forcing_from_cd"
)

The original checkpoint configuration is preserved unchanged. Its saved class name records the training wrapper, and _name_or_path records the upstream CD initialization source. The released payload is the ordinary self-forcing step-600 EMA checkpoint.