leftthomas/SimCLR

Issues

• Problem in dictionary keys

  #36 opened 10 months ago by deepankarvarma
  0

• Hello. How long did it take to pretrain your model with an NVIDIA TESLA V100(32G) GPU?

  #35 opened a year ago by duoqingxiaowangzi
  1

• Batch size doesn't affect

  #34 opened a year ago by momendoufu
  1

• Question about positive similarity in training process

  #33 opened 2 years ago by uqyqiu
  0

• loss function problem

  #32 opened 2 years ago by hendredlorentz
  2

• Pretrain model

  #31 opened 2 years ago by Holmes2002
  1

• Training

  #30 opened 2 years ago by Holmes2002
  1

• Why no upscale back to original size in augmentations?

  #29 opened 2 years ago by vedantroy
  1

• DataParrallel to avoid memory issue

  #28 opened 2 years ago by zhoubin-me
  4

• Calculation of Loss in train() seems wrong?

  #27 opened 3 years ago by wangruoyu1126
  2

• 下游任务训练的数据量

  #26 opened 3 years ago by lizhilin-ustc
  1

• what does total_num*100 mean

  #25 opened 3 years ago by xxuujjiiee
  1

• Low accuracy after linear evaluation

  #9 opened 4 years ago by ryancampbell514
  2

• I don't know how part of the code works

  #17 opened 3 years ago by libingDY
  0

• How can I perform data enhancement for other data sets?

  #24 opened 3 years ago by Kenny-Li2023
  1

• Useless download link!

  #23 opened 3 years ago by arashno
  1

• Question about linear finetune

  #22 opened 4 years ago by tangyuelm
  1

• results reported in your repo are finetuning and not linear evalaution

  #21 opened 4 years ago by fawazsammani
  1

• Data augmentation in linear evaluation

  #20 opened 4 years ago by fawazsammani
  1

• question about learning rate

  #19 opened 4 years ago by homelifes
  2

• Pretrained model load error

  #18 opened 4 years ago by ThomasMrY
  3

• Batch size 1024 in V100?

  #16 opened 4 years ago by FerranAlet
  1

• Have you tried the Global BN mentioned in the paper?

  #15 opened 4 years ago by MrChenFeng
  1

• Contrastive loss value

  #14 opened 4 years ago by abduallahmohamed
  1

• Why do you modify 'conv1' and remove maxpooling from ResNet50?

  #13 opened 4 years ago by AkihiroSasabe
  2

• Planning to run SimCLR on my dataset

  #12 opened 4 years ago by Jessejx
  2

• How to carry out semi supervised training？

  #11 opened 4 years ago
  0

• Using V100(32G) still get RuntimeError: CUDA out of memory.

  #10 opened 4 years ago by abdalgader-a-zz
  1

• cifar10上linear的结果超过了论文中的结果(92.0vs90.6)

  #1 opened 5 years ago by xieenze
  2

• Training time.

  #8 opened 4 years ago by MalharJ
  0

• In linear evaluation stage, 2048-embeddings are normalized before fc?

  #7 opened 4 years ago by haohang96
  0

• Benchmarking on CIFAR-100

  #6 opened 4 years ago by abhinavagarwalla
  1

• How can you load 512 batch size ?

  #5 opened 5 years ago by shamanez
  1

• Pretrained Model Links

  #4 opened 5 years ago by abhinavagarwalla
  1

• Augmentation during linear evaluation

  #3 opened 5 years ago by chenwydj
  3

• Adam or SGD?

  #2 opened 5 years ago by chenwydj
  1