I added --shapes=True to your hints, but in the end I didn't find the ply file, what can I do to get the 3d model?

[XuanNui]

I added --shapes=True to your hints, but in the end I didn't find the ply file, what can I do to get the 3d model?

[MustafaHilmiYAVUZHAN]

%cd /content/PanoHead
!python shape_utils.py /content/output/easy-khair-180-gpc0.8-trans10-025000.pkl/0/PTI_render/ --level 25

[MustafaHilmiYAVUZHAN]

/content/output/easy-khair-180-gpc0.8-trans10-025000.pkl/0/PTI_render/

[MustafaHilmiYAVUZHAN]

@XuanNui

[wangchenwc]

/content/output/easy-khair-180-gpc0.8-trans10-025000.pkl/0/PTI_render/

run this,but still not find the ply file

[BigRabit12138]

hello, there are some reasons for this:
firstly, --num-steps=200 --num-steps-pti=200, these paras should set blow 300, for eaxample i set 200;
sencondly, in projector_withseg.py: 391, you shoud set max_batch = 100000;
finaly, in projector_withseg.py: 427, set output_ply = True.
in the end, you will find /content/output/easy-khair-180-gpc0.8-trans10-025000.pkl/0/geometry.ply

[aspiring3dmodeler]

hello, there are some reasons for this: firstly, --num-steps=200 --num-steps-pti=200, these paras should set blow 300, for eaxample i set 200; sencondly, in projector_withseg.py: 391, you shoud set max_batch = 100000; finaly, in projector_withseg.py: 427, set output_ply = True. in the end, you will find /content/output/easy-khair-180-gpc0.8-trans10-025000.pkl/0/geometry.ply

I can't find the code?, where do you add it and when?

[aspiring3dmodeler]

Checked it and all the text you wrote is the same, it still didn't work

[BigRabit12138]

Checked it and all the text you wrote is the same, it still didn't work

hi , you should add --shapes=True firstly.

[aspiring3dmodeler]

Checked it and all the text you wrote is the same, it still didn't work

hi , you should add --shapes=True firstly.

Hi, already did that

[BigRabit12138]

Hi, already did that

hi, I tried again.I got a Ctrl-C in the ends without a ply file.With opening Resource Panel, I found System RAM was overhead, so the program was terminated(that why it was ended with Ctrl-C).Therefore I turned down "max_batch = 1000 voxel_resolution = 256" in projector_withseg.py in 391 392.After that, it works.
By the way, you could not get a nice 3D model in such a setting.If you want to get a useful 3D model , you need more powerful machine than google free T4.

[aspiring3dmodeler]

Ah ok, so free isn't good. Got it!

[aspiring3dmodeler]

Is it possible to skip the mp4 and only generate the ply?

[aspiring3dmodeler]

Hi, already did that

hi, I tried again.I got a Ctrl-C in the ends without a ply file.With opening Resource Panel, I found System RAM was overhead, so the program was terminated(that why it was ended with Ctrl-C).Therefore I turned down "max_batch = 1000 voxel_resolution = 256" in projector_withseg.py in 391 392.After that, it works. By the way, you could not get a nice 3D model in such a setting.If you want to get a useful 3D model , you need more powerful machine than google free T4.

Which gpu would i need and how many compute units does each use cost?

EDIT: facepalm question but should i clone it to drive first?

[BigRabit12138]

Hi, already did that

hi, I tried again.I got a Ctrl-C in the ends without a ply file.With opening Resource Panel, I found System RAM was overhead, so the program was terminated(that why it was ended with Ctrl-C).Therefore I turned down "max_batch = 1000 voxel_resolution = 256" in projector_withseg.py in 391 392.After that, it works. By the way, you could not get a nice 3D model in such a setting.If you want to get a useful 3D model , you need more powerful machine than google free T4.

Which gpu would i need and how many compute units does each use cost?

EDIT: facepalm question but should i clone it to drive first?

1. If you only want to get ply file , you need to comment out the related code in projector_withseg.py.
2. I didn't do a try for more powerful GPU because there is no free except Google T4.
3. No matter in local PC or Google Colab, colning the code is the first step(the file of colab notebook version already has cloning step).