Physis-AI/physisserve-minimaxh3_33b-same_step-ls-dev

PhysisServe — MiniMax H3 33B

A development release of MiniMax H3 33B with autoregressive video weights trained by Physis-AI for same-step noisy-KV inference. The ls-dev suffix identifies Physis-AI-trained weights.

Code · Technical blog · Model collection

These weights are intended for PhysisServe's four-step autoregressive inference. The Diffusers component layout does not make a stock Diffusers pipeline equivalent to this inference procedure: same-step KV caching, the attention sink, and the denoising schedule are implemented by PhysisServe.

Inference profile

The following defaults are defined by the PhysisServe model profile.

Setting	Default
Resolution (width × height)	1344 × 768
Output FPS	24
Denoising steps per chunk	4
First chunk	1 latent frame
Each subsequent chunk	1 latent frame
Chunks per generation, including the first	48
Attention sink	Near-clean bootstrap, σ = 0.001
KV history	Up to 18 recent chunks, plus the first-chunk sink without duplication

At denoising step s, a chunk reads noisy KV entries recorded at the same step s by preceding chunks. The first chunk is generated before the rest and receives an additional forward pass to build the persistent sink. A latent frame is not a decoded video frame; the VAE determines the temporal expansion. FPS is fixed by the model profile.

Quick start

Use Python 3.10+ with a CUDA-enabled PyTorch installation, then install PhysisServe:

pip install "git+https://github.com/Physis-AI/PhysisServe.git"

Save the following as generate.py and run python generate.py. Adjust pipeline_parallel_size to your available GPUs and memory before running; 1 below is an API example, not a single-GPU memory guarantee.

import imageio.v3 as iio
from physisserve import PhysisServe, SamplingParams


def main():
    model = PhysisServe(
        "Physis-AI/physisserve-minimaxh3_33b-same_step-ls-dev",
        pipeline_parallel_size=1,
    )
    sampling_params = SamplingParams(mode="offline_decode", seed=0)
    prompts = ["A corgi running along the beach at sunset."]
    outputs = model.generate(prompts, sampling_params)

    for index, output in enumerate(outputs):
        iio.imwrite(
            f"video-{index}.mp4",
            output["video"].numpy(),
            fps=output["fps"],
            codec="libx264",
            pixelformat="yuv420p",
            macro_block_size=None,
            output_params=["-movflags", "+faststart"],
        )


if __name__ == "__main__":
    main()

offline_decode saves latents under a separate directory in ./latents for each call, waits for all streams to finish denoising, and then decodes them. Set SamplingParams(latent_dir="your-directory", mode="offline_decode") to change the save location. The example writes playable MP4 files; generate() itself does not write MP4s.

A single prompt string or a list of prompts uses the same API. Multiple prompts create separate streams and increase memory demand.

Acknowledgments

We thank MiniMax for the base model and the Diffusers contributors for the model components.

License

See the MiniMax H3 Community License Agreement included in this repository, including its use restrictions. The PhysisServe engine's code license is separate and does not replace the terms for these weights or their bundled components.