lucidrains/magvit2-pytorch

Issues

• Is there anyone success to train this model?

  #34 opened 10 months ago by Jihun999
  34

• Why is magvitv2 different from the description in the paper? Am I understanding it wrong?

  #40 opened 7 months ago by hefeicyp
  6

• About GroupNorm described in the MAGVIT V2 paper

  #41 opened 4 months ago by sen-ye
  2

• Is this a bug at magvit2_pytorch.py#L1340?

  #43 opened 5 months ago by AIWanderer-X
  1

• Maybe a small bug at magvit2_pytorch.py#L1236?

  #44 opened 5 months ago by AIWanderer-X
  1

• About the correctness

  #42 opened 6 months ago by shinshiner
  1

• Running multi-gpu training

  #35 opened 10 months ago by joe-sht
  5

• About training speed.

  #39 opened 7 months ago by kxgong
  0

• recon images is black

  #31 opened a year ago by syyxsxx
  9

• About training steps and correctness.

  #36 opened 10 months ago by becauseofAI
  3

• Training difficulties

  #4 opened a year ago by LouisSerrano
  51

• Running multi-gpu hangs after first step

  #18 opened a year ago by jpfeil
  9

• Is there any pretrained weights for debug?

  #38 opened 9 months ago by clownrat6
  1

• Pixelated image reconstruction

  #29 opened a year ago by coolbunnyx
  7

• Error while loading the states of optimizer in Trainer - def load(self, path)

  #37 opened 9 months ago by OliviaWang123456
  0

• The results for CausalConv3d

  #11 opened a year ago by Epiphqny
  8

• train on video dataset

  #7 opened a year ago by Y-ichen
  4

• Large scale training

  #17 opened a year ago by iejMac
  17

• Question about Imagenet Parameters

  #6 opened a year ago by jpfeil
  4

• The configuration of training

  #33 opened 10 months ago by RobertLuo1
  0

• Unsuccessful image reconstruction

  #25 opened a year ago by coolbunnyx
  3

• Question about casual 3d cnn

  #32 opened a year ago by eii-lyl
  1

• ‘video_contains_first_frame’ in encoder

  #30 opened a year ago by Du-Yao
  1

• pretrained weights

  #28 opened a year ago by wileewang
  0

• weights

  #23 opened a year ago by HongliXiao
  1

• object has no attribute 'has_multiscale_discrs'

  #22 opened a year ago by Ted-developer
  2

• expired discord invitation

  #26 opened a year ago by xyzhang626
  3

• Reconstruction image is always a solid color

  #12 opened a year ago by jpfeil
  20

• Is there any requirement on the training images?

  #19 opened a year ago by wd255
  3

• using different samples for gan generator and discriminator

  #8 opened a year ago by teou
  3

• Discriminator loss converges to zero early in training

  #16 opened a year ago by jpfeil
  9

• Running with GAN raises RuntimeError

  #15 opened a year ago by jpfeil
  8

• Flash attention not working on A100 GPU

  #9 opened a year ago by jpfeil
  2

• Example for Images

  #3 opened a year ago by jpfeil
  2

• implementation

  #2 opened a year ago by 0nutation
  1