muslll/neosr

Issues

• [Question] How Nomos-v2 is collected? Is there a benchmark on performance?

  #86 opened 11 days ago by Luciennnnnnn
  8

• Loss calculation should indicate which loss(es) returned NaN

  #87 opened 24 days ago by pokepress
  3

• Suggestions/corrections for setup/walkthrough page

  #85 opened a month ago by pokepress
  3

• Support safetensors as pretrained models

  #36 opened 2 months ago by RunDevelopment
  2

• NaN errors occurring constantly

  #83 opened 2 months ago by wdsongama
  1

• ValueError at 79 epoch

  #82 opened 3 months ago by someitalian123
  3

• All architectures with `rgb_mean = (0.5, 0.5, 0.5)` are incompatible

  #73 opened 3 months ago by RunDevelopment
  9

• GPU memory not fixed when training

  #81 opened 3 months ago by ProBURN-E
  1

• can not start train

  #80 opened 4 months ago by linkhins122
  3

• Question regarding net_d

  #78 opened 4 months ago by seanko29
  4

• Python version problem

  #77 opened 4 months ago by seanko29
  4

• Mismatched resume state issue

  #72 opened 5 months ago by Kim2091
  1

• Automatically remove old checkpoints and visualizations

  #66 opened 4 months ago by RunDevelopment
  4

• Modifications for CRAFT, SwinIR, and DCTLAS are not detectable

  #64 opened 6 months ago by RunDevelopment
  8

• Add tile option

  #63 opened 5 months ago by marcobarbierato
  2

• Document which architectures have been modified and are incompatible with the original architecture

  #61 opened 6 months ago by RunDevelopment
  4

• imwrite raise IOError when trying to write an image where the filename contains unicode characters

  #59 opened 6 months ago by mogrol
  2

• Add Anime4K models

  #53 opened 7 months ago by arianaa30
  26

• model zoo

  #79 opened 4 months ago by ttys0001
  1

• assets

  #68 opened 5 months ago by muslll
  3

• dataset_enlarge_ratio description

  #70 opened 5 months ago by wdsongama
  1

• How to save a model fully in pth? Not state dict

  #60 opened 6 months ago by v0vanch0
  3

• and again rgb_mean

  #67 opened 6 months ago by umzi2
  1

• `train_realesrgan.yml` is named "train_esrgan"

  #58 opened 7 months ago by RunDevelopment
  1

• wrapper around train.py, ... scripts and release on pypi

  #56 opened 7 months ago by alsa64
  0

• Exiting early doesn't save latest training state

  #12 opened a year ago by awused
  2

• Huge performance drop after update

  #54 opened 7 months ago by VandersonQk
  3

• Models and states saving/loading issue

  #37 opened 8 months ago by terrainer
  1

• incorrect conversion to grayscale

  #28 opened 8 months ago by umzi2
  1

• Non-RGB SPAN models

  #25 opened 9 months ago by RunDevelopment
  13

• train.py throws an error, unable to proceed to training.

  #24 opened 9 months ago by youzrneym
  3

• Feature request: allow toggling attention implementation via options

  #20 opened 9 months ago by Mikewando
  2

• Wiki feedback

  #21 opened 10 months ago by RunDevelopment
  3

• Real-CUGAN License

  #18 opened 10 months ago by Kim2091
  2

• Cannot resume training ESRGAN

  #19 opened 10 months ago by Vega-KH
  4

• just wanted to express my respect

  #17 opened 10 months ago by sanbuphy
  1

• SPAN incompatible with official arch code

  #16 opened 10 months ago by RunDevelopment
  2

• Perceptual Loss and AMP - discussion

  #4 opened a year ago by muslll
  1

• Print file path that causes read error upon dataloader crash

  #15 opened 10 months ago by Kim2091
  3

• Validation fails for HAT models

  #13 opened a year ago by awused
  2

• cv2.error: Caught error in DataLoader worker process 2.

  #5 opened a year ago by radry
  3

• Validation doesn't appear to be working.

  #3 opened a year ago by Sirosky
  1

• AMP enabled info spamming the console

  #1 opened a year ago by Sirosky
  1

• OmniSR variable passing bug.

  #2 opened a year ago by Sirosky
  1