lucidrains/gigagan-pytorch

Issues

• support 256x256 images

  #55 opened 3 months ago by YoucanBaby
  1

• Video?

  #59 opened 2 months ago by darkzbaron
  2

• How to adapt this network to image to image translation task

  #57 opened 2 months ago by stephano41
  0

• Loss turned `NaN` when custom training

  #56 opened 3 months ago by yasohasakii
  1

• Multi GPU with gradient accumulation

  #37 opened a year ago by dprze
  1

• Standard Attention vs ViTGAN Attention

  #54 opened 4 months ago by jfischoff
  1

• [News] Videogigagan is published.

  #53 opened 4 months ago by O-O1024
  4

• Has Anyone Trained This Model Yet?

  #48 opened a year ago by PaulFidika
  2

• Training plans?

  #17 opened a year ago by nbardy
  122

• NaN losses after hours of training (UPSAMPLER)

  #40 opened a year ago by sansavision
  16

• the loss became nan after a few train steps

  #52 opened 9 months ago by boya34
  2

• question about code in unet_upsampler.py

  #51 opened 10 months ago by 8639sung
  1

• Config to reproduce paper

  #50 opened a year ago by dawgster
  0

• The text-to-image tasks

  #49 opened a year ago by SYZhang0805
  0

• How to use this model for SR ?

  #47 opened a year ago by LilyTheBear
  0

• Gradient Penalty is very high in the start

  #46 opened a year ago by inspirit
  10

• to_rgb branch has only 1 learnable kernel

  #45 opened a year ago by inspirit
  7

• Wrong order of resolutions list

  #44 opened a year ago by inspirit
  1

• Turn on/off gradients computation between generator/discriminator

  #43 opened a year ago by inspirit
  2

• How to implement this model to enhance my input images? Do I have to train the model to use?

  #41 opened a year ago by Mr-Nobody-dey
  2

• Weights of Gigagan Upscaler

  #42 opened a year ago by dheerajpant700
  1

• [Question] About the upscaler

  #35 opened a year ago by LouSparfell
  2

• Confused about this project?

  #39 opened a year ago by Maki9009
  0

• [Request] Please provide a replicate.com version

  #38 opened a year ago by dreamflasher
  0

• Multi GPU training

  #36 opened a year ago by dprze
  4

• Possible Discrepancies

  #33 opened a year ago by randintgenr
  3

• The training code not deal with paired data yet?

  #34 opened a year ago by mychina75
  2

• What makes the gan upsampler so effective

  #7 opened a year ago by salvadog
  10

• Question about the timing and complexity of GigaGAN replication

  #9 opened a year ago by ARTUROSING
  3

• Some doubts about the multi-scale input and output of the super-resolution part

  #31 opened a year ago by boya34
  5

• Hardware Requirements

  #32 opened a year ago by kaykyr
  4

• Why change the collection of rgb in upsampler?

  #30 opened a year ago by XavierXiao
  1

• How do I run inference with this model?

  #13 opened a year ago by kct22aws
  2

• Some minor questions regarding the design

  #29 opened a year ago by XavierXiao
  2

• Questions regarding the discriminator

  #19 opened a year ago by XavierXiao
  10

• fidelity

  #15 opened a year ago by Inferencer
  1

• Generator/Discriminator dim Argument

  #26 opened a year ago by CerebralSeed
  2

• rgb info in the generator is not collected in the correct way?

  #18 opened a year ago by XavierXiao
  8

• L2 attention is implemented wrong!

  #14 opened a year ago by PeterL1n
  10

• Question about discriminator

  #12 opened a year ago by Zijian007
  8

• Insert the attention mechanism in the codes to the stylegan2-ada but fail

  #10 opened a year ago by landian60
  7

• If I can just use the super resolution module to do some SR task? How long can you finish the sr moddules

  #11 opened a year ago by Shawn-Yu-1
  2

• License Violation Issue

  #8 opened a year ago by CerebralSeed
  5

• basic setup/install documentation

  #5 opened a year ago by djo267
  2

• Is this project ready to train?

  #6 opened a year ago by kaykyr
  1

• How can I use this?

  #4 opened a year ago by AceBeaker2
  5

• How far is this codebase?

  #3 opened a year ago by stencilman
  1

• Implement dynamic convulsion

  #2 opened a year ago by Jainam2130
  1

• Pretrained models

  #1 opened a year ago by IceClear
  4