ayaanzhaque / instruct-nerf2nerf

Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions (ICCV 2023)
https://instruct-nerf2nerf.github.io/
MIT License
769 stars 64 forks source link

in2n is not registered in the method #6

Closed pacifinapacific closed 1 year ago

pacifinapacific commented 1 year ago

Create nerfstudio by pulling the official docker image and then installed in2n. pip install git+https://github.com/ayaanzhaque/instruct-nerf2nerf

The nerfacto training completes successfully, but when I use in2n, I get the following error: ns-train: error argument {nerfacto,depth-nerfacto,instant-ngp,instant-ngp-bounded,mipnerf,semantic-nerfw,vanilla-nerf,tensorf,dnerf,phototourism}: Can you tell me why in2n is not added t the method?

Launch python3 and import in2n works fine

ayaanzhaque commented 1 year ago

It seems to work fine on my end when i create a conda environment, install the dependencies, and then do the pip install. Could you try the instructions exactly as they are listed in the readme?

frankjiang commented 1 year ago

You may not install the suitable version of nerfstudio. nerfstudio==0.1.19 (lastest) works fine for me. Use centos or other system, the nerfstudio may be 0.1.8 or early. In the early version, the plugins mode is not develop yet.

hugosbnarciso commented 1 year ago

what's the return of ns-train -h

ayaanzhaque commented 1 year ago

It should have the three new methods listed at the bottom, in2n, in2n-small, in2n-tiny

pacifinapacific commented 1 year ago

Thank you very much. It was due to the fact that I was using nerfstudio docker version 1.7. Using the latest version (2.0) will work correctly.

cwwjyh commented 1 year ago

It should have the three new methods listed at the bottom, in2n, in2n-small, in2n-tiny

hello, I also meet this question. i find ns-trian -h, no return you said in2n, in2n-small, in2n-tiny. my version of nerfstudio=0.2.1. how I to fix this problem. 1682337704242 image