Ruizhi Shao

Hi @Cli98 , Could you provide more detailed information about this error?

Hi @ffli , Yes, you can directly build them from git and delete them in requirements.txt.

Hi @nkundtz , Thank you for your advice, I have updated [requirements.txt](https://github.com/DSaurus/MVDream/blob/main/requirements.txt) (removing xformers ) and added [instructions](https://github.com/DSaurus/threestudio-mvdream/tree/main?tab=readme-ov-file#installation) in installation.

Hi, @junaidwahid You could consider changing the `batch_size` to `[12,8,2]` or use [threestudio-stable-nerf-renderer extension](https://github.com/DSaurus/threestudio-stable-nerf-renderer). ``` random_camera: # threestudio/data/uncond.py -> RandomCameraDataModuleConfig height: [64, 128, 256] width: [64, 128, 256] batch_size: [12,...

Hi @echosum , You can refer to [DreamCraft3D](https://github.com/DSaurus/threestudio-dreamcraft3D).

Hi @cseander , I think it is a CPU bottleneck. You can first test the stable-zero123 diffusion model by [threestudio-mvimg-gen](https://github.com/DSaurus/threestudio-mvimg-gen).

> That's odd, I have a i9-13900KS and 128GB RAM. Could it be because I'm running it in WSL? > > When I try threestudio-mvimg-gen I get the following error:...

HI @GggggiitHHHHuuub `DeepFloyd/IF-I-XL-v1.0` needs to have an access token. Please login into huggingface to get that token or you can use `dreamfusion-sd` to replace guidance with stable diffusion.

Hi @JunzheJosephZhu , I have invited you to join threestudio project. It would be better if you could create a branch and submit a new pull request.

Hi, @jon-chuang That's interesting. I think it is a CFG problem. Also, the gradient of `sd-turbo` may be too large, which causes instability in 3D generation.