About the reproduction of figs in the paper

EnVision-Research / LucidDreamer

Official implementation of "LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching"

MIT License

749 stars 32 forks source link

About the reproduction of figs in the paper #33

Closed chenxinl closed 6 months ago

chenxinl commented 6 months ago

Hi,

Thanks for your great work.

I tried to reproduce the figs such as Fig. 1 in the paper following the training scripts in ./configs with some modifications. But the results are less than satisfactory.

Could you give more training configs of Fig. 1 in the paper?

Thanks.

AbnerVictor commented 6 months ago

Hi, a similar issue has been answered in #22 , please feel free to ask if you have any further questions.