Alibaba-MIIL/ImageNet21K

Issues

• 请教一下，Winter21 Imagenet21k在国内服务器上如何才能快速下载？

  #72 opened 6 months ago by 2catycm
  1

• winter21

  #71 opened 7 months ago by xi203040
  0

• CIFAR-100 pretrained ViT-B-16 weight

  #70 opened a year ago by HengGao12
  2

• Dependencies to run code

  #69 opened a year ago by meghbhalerao
  0

• When using your ImageNet21K pretrained ResNet50 model in Detectron2, performance degrades

  #57 opened 3 years ago by miznchimaki
  1

• ask for the Transfer Learning Code of train.py on cifar100

  #68 opened 2 years ago by xiaohua-chen
  0

• ask for the pretrained model on ImageNet-21k-P from single label

  #67 opened 2 years ago by hongge831
  0

• How to test Imagenet1K with pretrained backbone MobilenetV3_large_100? Could you release the testing script? Thanks a lot.

  #66 opened 2 years ago by ywang-37
  0

• I see no normalization of images.

  #58 opened 2 years ago by Phuoc-Hoan-Le
  0

• Anyone here have trouble reaching the mentioned accuracy for ViT-B?

  #65 opened 2 years ago by Phuoc-Hoan-Le
  0

• Does there any decriptions abount classes in "imagenet21k_small_classes"

  #64 opened 2 years ago by JH-Xie
  0

• Where can I find the data?

  #62 opened 2 years ago by johnypark
  0

• Single label pretraining on in21k using ViT

  #45 opened 3 years ago by cissoidx
  0

• What is the teacher model when using semantic softmax with KD?

  #61 opened 2 years ago by Phuoc-Hoan-Le
  0

• Missing details on Dropout and momentum value used for SGD when fine tuning on ImageNet1k

  #60 opened 2 years ago by Phuoc-Hoan-Le
  0

• Hyperparameters to finetune ResNet50 from IN21k to IN1k

  #59 opened 2 years ago by andreamaracani
  0

• For a customized dataset, how to build a semantic tree and how to match them with the label defined by pytorch ImageFolder?

  #23 opened 3 years ago by aiot-tech
  2

• could you please provide image-label map directly?

  #56 opened 3 years ago by liming-ai
  0

• Any label map for ImageNet-1K?

  #55 opened 3 years ago by ckddls1321
  0

• No parent for n09450163 (sun)

  #54 opened 3 years ago by jvlmdr
  0

• About hierarchy balancing

  #53 opened 3 years ago by hankyul2
  2

• Can you share mean, std of imagenet21k?

  #51 opened 3 years ago by ooodragon94
  2

• Is the 1k validation set included in the 21k data？

  #52 opened 3 years ago by Bo396543018
  2

• The strategy of transferring ImageNet-21k ViT model to cifar100

  #40 opened 3 years ago by Dyfine
  5

• why no data normlization in data pre-processing？when I use data normlization in data pre-processing， rate of convergence of the network is slow

  #50 opened 3 years ago by dmuqlzhang
  1

• ImageNet-1K Metadata

  #49 opened 3 years ago by Auth0rM0rgan
  3

• The question about the relative number of occurrences of hierarchy k in the loss function

  #46 opened 3 years ago by ckx506772099
  3

• ResNet50 model fine-tuned on the ILSVRC-2012 dataset

  #48 opened 3 years ago by interstate50
  1

• Detected call of `lr_scheduler.step()` before `optimizer.step()`

  #47 opened 3 years ago by zbwxp
  1

• labeling and classes of images

  #42 opened 3 years ago by Tortoise17
  9

• could I transferred the paper's idea to object detection task?

  #44 opened 3 years ago by cs-heibao
  1

• data augmantations

  #43 opened 3 years ago by cissoidx
  1

• imagenet1k pretrain model

  #26 opened 3 years ago by cissoidx
  24

• About downstream tasks

  #41 opened 3 years ago by SJLeo
  1

• The strategy of transferring ImageNet-21k models to ImageNet-1K.

  #38 opened 3 years ago by whai362
  5

• pretrain head of vit

  #39 opened 3 years ago by cissoidx
  5

• About Training Setting Parameters

  #35 opened 3 years ago by SJLeo
  5

• Which version is imagenet21k_resized.tar.gz ?

  #37 opened 3 years ago by Felix-Petersen
  2

• Line 60 in `train_semantic_softmax.py`

  #36 opened 3 years ago by Felix-Petersen
  1

• casting error in AccuracySemanticSoftmaxMet

  #34 opened 3 years ago by cissoidx
  1

• General fail in preprocessing

  #25 opened 3 years ago by cissoidx
  13

• Detection and segmentation performance

  #30 opened 3 years ago by akuxcw
  3

• How long does one epoch take?

  #31 opened 3 years ago by cissoidx
  4

• About Multi-label Training Scheme

  #24 opened 3 years ago by SJLeo
  15

• How to download semantic tree: winter21_imagenet21k_tree.pth?

  #32 opened 3 years ago by zlenyk
  2

• the Adam Optimizer

  #33 opened 3 years ago by Dyfine
  2

• Model weights trained on Stanford cars

  #28 opened 3 years ago by ma-siddiqui
  5

• Normalize

  #29 opened 3 years ago by akuxcw
  2

• 如何用resnet50加载21k pertained

  #27 opened 3 years ago by daixiangzi
  5

• data deduplicate

  #22 opened 3 years ago by pawopawo
  3