Wan2.1-T2V-14B-SFP-DMD-4Step-LoRA-r32-iter600

Wan2.1 T2V 14B SFP DMD 4-Step LoRA - Rank 32, Iter 600

This repository contains the iteration-600 checkpoint from an experimental Wan2.1-T2V-14B Self-Forcing-Plus DMD run. It includes a portable generator LoRA for inference and the complete generator/critic/optimizer checkpoint for resuming the original 64-rank training job.

This is an intermediate checkpoint from a run configured for 3,000 iterations. It is published for controlled rank comparisons and should not be treated as the final model.

Checkpoint identity

Each model copy trains LoRA weights for every Linear module inside the 40 Wan attention blocks. The generator and critic use separate adapters. The published inference adapter contains only the generator weights.

Training configuration

The training path keeps the original Self-Forcing-Plus rollout, DMD loss, sampler, update order, and prompt-only data flow. At step 600, the validated Adam counters are 120 generator updates and 600 critic updates.

Files

• generator_lora_lightx2v.safetensors: generator-only BF16 LoRA for four-step LightX2V inference.
• inference_config.json: the 4-step, 81-frame, 832x480 LightX2V settings.
• training_config.yaml: source SFP DMD training configuration.
• training_checkpoint/model.pt: FP32 generator and critic LoRAs plus adapter metadata at step 600.
• training_checkpoint/optimizer_rank_00000.pt through optimizer_rank_00063.pt: rank-local AdamW and RNG states.
• training_checkpoint/_SUCCESS: complete-checkpoint marker.
• export_generator_lora.py: exporter used to create the inference adapter.
• provenance.json: revisions, checksums, and validation facts.

The files under training_checkpoint/ are PyTorch pickle checkpoints. Load them only in a trusted environment. The inference adapter uses Safetensors.

LightX2V inference

Download this repository and change lora_configs[0].path in inference_config.json to the absolute path of generator_lora_lightx2v.safetensors. Then run from a LightX2V checkout:

python -m torch.distributed.run --nproc_per_node=8 -m lightx2v.infer \
  --model_cls wan2.1_distill \
  --task t2v \
  --model_path /path/to/Wan2.1-T2V-14B \
  --config_json /path/to/inference_config.json \
  --prompt "your prompt" \
  --negative_prompt "your negative prompt" \
  --save_result_path output.mp4 \
  --seed 42

The validated setup uses LoRA strength 1.0, four denoising steps, 81 output frames, 832x480 resolution, and no external CFG pass.

Resume training

Resume with the original modified Self-Forcing-Plus code and point resume_ckpt at the downloaded training_checkpoint directory. Restoring the supplied optimizer shards requires a world size of 64. The checkpoint was validated for all 64 shards before publication.

Integrity