Why the inference speed is slower than the README file?

[sunshinewhy]

I use single A800 GPU to run the command line:
cd HunyuanVideo

python3 sample_video.py
--video-size 720 1280
--video-length 129
--infer-steps 50
--prompt "A cat walks on the grass, realistic style."
--flow-reverse
--use-cpu-offload
--save-path ./results
it takes nearly 3000s to generate a video, but the README file records 1900s. Why does the difference occur?

[sunshinewhy]

BTW, can you release the latest image file include gradio？

[feifeibear]

Because the readme report latency on H100.