thuanz123/enhancing-transformers

Issues

• Could anyone reproduce the rFID scores from the small-large enc-dec version?

  #24 opened 3 months ago by shashankg7
  0

• Incomplete implementation of RQ-VAE

  #23 opened 9 months ago by JJJYmmm
  0

• why not used pretrained discriminator

  #22 opened 10 months ago by BaohaoLiao
  0

• Learning rate and scheduler for stage 1 training.

  #21 opened a year ago by duchenzhuang
  1

• The question is done.

  #18 opened a year ago by chengjiaxiangbytedance
  1

• Reconstruction results

  #20 opened a year ago by Marcelo5444
  4

• Reconstruction Visualization

  #12 opened 2 years ago by iamwangyabin
  2

• ImageNet version.

  #19 opened a year ago by chengjiaxiangbytedance
  0

• stage1 pretraining

  #14 opened a year ago by mactavish91
  7

• about training loss

  #17 opened a year ago by CrossLee1
  1

• stage2 transform

  #16 opened 2 years ago by chengjiaxiangbytedance
  9

• Model license

  #15 opened 2 years ago by Aspie96
  4

• stage2 transformer

  #13 opened 2 years ago by Mdahao
  2

• An inplace operation in the forward process

  #11 opened 2 years ago by VanderHua
  1

• Pretrained Stage 2 Transformer for ViT-VQGAN

  #8 opened 2 years ago by manuelknott
  8

• Pre-norm and Post-norm

  #10 opened 2 years ago by zyf0619sjtu
  6

• Training time and number of GPUS

  #9 opened 2 years ago by JialeTao
  6

• Smaller images

  #6 opened 2 years ago by snarb
  10

• OOM for imagenet_gpt_vitvq_base and a 100M params GPT on A100 40G

  #7 opened 2 years ago by BlinkDL
  3

• Results

  #5 opened 2 years ago by snarb
  5

• raise CalledProcessError(retcode, process.args, subprocess.CalledProcessError: Command '['where', 'cl']' returned non-zero exit status 1.

  #4 opened 2 years ago by bastvdn
  5