lucidrains/meshgpt-pytorch

Issues

• Recommended Hyperparamters

  #96 opened a month ago by NickPJQ
  0

• does the “Int['b nv 3']” mean？

  #95 opened a month ago by Syuao
  0

• Would be possible to apply on objects with a lotmore meshes?

  #94 opened 2 months ago by 17740201
  0

• hyper-parameter suggestion

  #93 opened 3 months ago by chinmay5
  20

• Continuous range for discretize area

  #91 opened 5 months ago by sbriseid
  2

• Area for quads

  #92 opened 5 months ago by sbriseid
  6

• Transformer keeps predicting the same token

  #82 opened 5 months ago by MagicianWu
  8

• Version conflict with x-transformer-1.31.6

  #89 opened 5 months ago by hyeonjang
  1

• [Critical] Very high loss rate at first few tokens (classifier free guidance not working)

  #80 opened 5 months ago by MarcusLoppe
  66

• distirbution training

  #70 opened 8 months ago by kingofrubbish2
  1

• Pretrained checkpoints?

  #76 opened 8 months ago by agaier
  2

• Area calculation for each triangle is same (with value of 0.5)

  #83 opened 6 months ago by aakash-saboo
  6

• Seeking Collaboration

  #81 opened 7 months ago by xhsyy
  3

• Mesh conditioning instead of text conditioning

  #77 opened 7 months ago by pathquester
  6

• New/replace attention mechanism

  #79 opened 7 months ago by MarcusLoppe
  0

• Guide for generating novel meshes based on the trained models

  #78 opened 7 months ago by HenryLhc
  0

• Loss=nan when training transformertrainer

  #68 opened 8 months ago by HenryLhc
  5

• Transformer - High VRAM, context length

  #39 opened a year ago by MarcusLoppe
  4

• `ResidualLFQ` was successful, but `ResidualVQ` failed severely!

  #73 opened 8 months ago by fighting-Zhang
  2

• High RAM cost in derive_face_edges_from_faces

  #75 opened 8 months ago by inmny
  0

• Mesh completion task

  #71 opened 8 months ago by jrmgr
  2

• A question about the code in 'meshgpt_pytorch.py'

  #72 opened 8 months ago by LeeYan1
  0

• Mesh intra face vertex id ordering convention

  #62 opened 8 months ago by jrmgr
  16

• TypeError: MessagePassing.__init__() got an unexpected keyword argument 'sageconv_dropout'

  #65 opened 8 months ago by StephenYangjz
  12

• ImportError: cannot import name 'MeshDataset' from 'meshgpt_pytorch'

  #67 opened 9 months ago by StephenYangjz
  5

• Transformer - token_embed outputs nan values

  #44 opened 8 months ago by MarcusLoppe
  11

• GroupNorm in ResnetBlocks

  #64 opened 9 months ago by shanemankiw
  13

• Sliding window for transformer

  #61 opened 9 months ago by MarcusLoppe
  1

• Question

  #57 opened 10 months ago by nicolasdonati
  10

• Classifier-Free Guidance, cond_drop_prob=1.0, attn_mask=False: Error!!!

  #63 opened 9 months ago by fighting-Zhang
  2

• when `transformer.generate(prompt=None)`,empty code is passed to the decoder. Error!!

  #55 opened 9 months ago by fighting-Zhang
  14

• Is there a pretrained model and if not, how to train the model

  #56 opened 9 months ago by jonasweimar
  12

• Just to get started

  #60 opened 9 months ago by studiohuahong
  2

• Correction in the meshgpt paper.

  #58 opened 10 months ago by fire
  0

• text_condition_model_types

  #59 opened 10 months ago by Ginjing-Yuan
  0

• Commit loss is negative

  #43 opened a year ago by ZekaiGalaxy
  8

• Support for quads

  #54 opened 10 months ago by pathquester
  38

• Isn't MeshGPT a VQ-VAE?

  #53 opened 10 months ago by rexxxx1234
  2

• Simple training script for toy data?

  #46 opened a year ago by xiao-xian
  6

• is gaussian_blur_1d correct?

  #52 opened 10 months ago by thuliu-yt16
  1

• Residual quantization - VRAM bottleneck

  #36 opened 10 months ago by MarcusLoppe
  13

• Recipe for overfitting a single obj

  #38 opened a year ago by ZekaiGalaxy
  6

• transformer.generate()

  #41 opened a year ago by Genshin-Impact-king
  3

• derive_face_edges_from_faces high ram usage

  #34 opened a year ago by MarcusLoppe
  5

• MeshTransformer.generate does not work with a prompt if kv cache is enabled

  #48 opened a year ago by Kurokabe
  15

• How to preprocess ShapeNet models?

  #35 opened a year ago by xuanyaoming
  18

• Training on my own data

  #49 opened a year ago by adeerAI
  1

• gateloop_use_heinsen=True on MeshTransformer results in NaN loss

  #47 opened a year ago by Kurokabe
  6

• `face_mask` rearrange in Autoencoder Forward Pass

  #40 opened a year ago by qixuema
  2

• Typo Error: `dim` Parameter in `MeshAutoencoder`

  #33 opened a year ago by qixuema
  1