Downloads · 30 days
0
GuoShuaiGO/percentile-hif4-w4a4-wan22
percentile-hif4-w4a4-wan22 is a image-to-video model from GuoShuaiGO. Use it for the image-to-video task on the model card, and read the license before you ship it in a product. It is set up for wan2.2. The card lists the license as apache-2.0.
Official quantized checkpoint release for our award-winning solution to the ICME 2026 Low-Bit-width Large-Model Quantization Challenge:
Downloads · 30 days
0
Access
Public
Updated May 19, 2026
Repo size
9.9 MB
Likes
0
Public
Click a slice to open those files.
.pth9.9 MB · 100%
From the Hugging Face model README
Official quantized checkpoint release for our award-winning solution to the ICME 2026 Low-Bit-width Large-Model Quantization Challenge:
Percentile-Calibrated HiFloat4 W4A4 Post-Training Quantization for Wan2.2 Text-to-Video Generation
This repository contains the official minimal PTQ checkpoint for the public release of the challenge solution above.
Code repository: https://github.com/GuoShuaiGO/percentile-hif4-wan22-icme2026
Checkpoint repository: https://huggingface.co/GuoShuaiGO/percentile-hif4-w4a4-wan22
Base model: Wan-AI/Wan2.2-I2V-A14B-Diffusers
Quantized artifact: quant_ckpt.pth
Release config: hif4_w4a4.yaml
The checkpoint stores only the inference-time PTQ state needed to reconstruct the released quantized model on top of the public Wan2.2 base model. It does not duplicate the full Wan base weights.
quant_ckpt.pth: minimal inference-time PTQ checkpointhif4_w4a4.yaml: quantization configuration used for the released settingUse this checkpoint with the companion official code repository for this release.
The code repository provides:
Companion code repository: https://github.com/GuoShuaiGO/percentile-hif4-wan22-icme2026
The competition task is text-to-video generation, while the public base checkpoint used in this release is Wan2.2-I2V-A14B-Diffusers, whose public pipeline requires an image input. The companion code reproduces prompt-only public inference by feeding an explicit blank placeholder image and logging that behavior during generation.
After downloading the public Wan2.2 base model and cloning the companion code repository:
mkdir -p checkpoints
huggingface-cli download GuoShuaiGO/percentile-hif4-w4a4-wan22 quant_ckpt.pth --local-dir ./checkpoints
huggingface-cli download GuoShuaiGO/percentile-hif4-w4a4-wan22 hif4_w4a4.yaml --local-dir ./checkpoints
python examples/wan2.2_hif4/quant_inference.py \
--model-path ./models/Wan2.2-I2V-A14B-Diffusers \
--quant-ckpt ./checkpoints/quant_ckpt.pth \
--dataset-json ./assets/demo_prompts.json \
--prompt-key cap \
--num-samples -1 \
--height 720 \
--width 1280 \
--num-frames 61 \
--num-inference-steps 40 \
--output-dir ./logs/wan2.2_hif4/generated_videos
This checkpoint release is distributed under the Apache-2.0 License.