google-research/big_transfer

Issues

• How many learnable parameters does the ResNet-152x4 BiT model has?

  #86 opened 2 years ago by Lampent
  0

• 2 extra output classes in ImageNet-21K

  #18 opened 5 years ago by kracwarlock
  3

• Possibly Wrong Model Name in README?

  #84 opened 2 years ago by m-parchami
  0

• Is "post pre-training (Imagenet weight init) on custom dataset" possible with this git code?

  #83 opened 2 years ago by flipflop98
  0

• 'StandardizedConv2D' object has no attribute '_convolution_op'

  #80 opened 2 years ago by spradeep28
  2

• Pretraining hyperparameter conflict between code and paper

  #79 opened 2 years ago by sgunasekar
  0

• pad then pool in root defintion

  #74 opened 2 years ago by sgunasekar
  0

• module must have its parameters and buffers on device cuda:0 (device_ids[0]) but found one of them on device: cuda:1

  #73 opened 2 years ago by RuojiWang
  1

• is there any trick to improve bit's perfermance on cifar10?

  #72 opened 2 years ago by RuojiWang
  0

• convert pth to onnx will give error"Exporting the operator var_mean to ONNX opset version {version} is not supported"

  #69 opened 3 years ago by nakairika
  0

• Reproducing problems

  #70 opened 3 years ago by Sukeysun
  0

• GRAD CAM

  #42 opened 4 years ago by movaldivia
  2

• "flip_left_right" in input_pipeline_tf2_or_jax.py

  #62 opened 3 years ago by exiarepairii
  0

• Unable to replicate published benchmark accuracy on cifar10 dataset

  #59 opened 3 years ago by abdul53saifee
  0

• Linear warmup learning rate is 0 when step = 0

  #58 opened 3 years ago by danielamassiceti
  0

• TF2 weights shape wrong for 2 architectures

  #47 opened 4 years ago by shkarupa-alex
  2

• what's the meaning of Ex/cls

  #55 opened 4 years ago by GuoQuanhao
  1

• Top-1 accuracy on few-shot tasks

  #54 opened 4 years ago by fengliu90
  0

• Fine tuning from specific layer number

  #53 opened 4 years ago by sumedhvdatar
  0

• Jax from flax.nn to flax.linen?

  #52 opened 4 years ago by dave-andersen
  0

• Saving model after finetuning

  #33 opened 4 years ago by sahilwalke17
  4

• how to use my custom trained model

  #51 opened 4 years ago by kyuhong0206
  0

• do not find models that trained on JFT300M?

  #50 opened 4 years ago by lichun-wang
  0

• [doc] Clarification on the input normalisation format.

  #49 opened 4 years ago by Conchylicultor
  0

• is it possible to speed up the prediction?

  #48 opened 4 years ago by ibrahim-ozcan
  0

• How to download ImageNet21K dataset?

  #41 opened 4 years ago by FredHuang16
  2

• Training my own dataset

  #36 opened 4 years ago by charliewang789
  6

• Experiment with SWA to improve the state of the art

  #38 opened 4 years ago by LifeIsStrange
  0

• BiT as Image Classification on Mobile Application

  #37 opened 4 years ago by ngptruong
  0

• [Question] why do we need PaddingFromKernelSize instead of ConV padding?

  #35 opened 4 years ago by chychen
  0

• PyTorch BiT-M-ResNet152x4 model bug

  #34 opened 4 years ago by junsukchoe
  0

• Transfer Learning For Detectors

  #30 opened 4 years ago by abhiagwl4262
  3

• Invalid argument: slice index 0 of dimension 0 out of bounds.

  #32 opened 4 years ago by sahilwalke17
  0

• AttributeError: Layer block2 has no inbound nodes.

  #31 opened 4 years ago by himanshurawlani
  0

• Loss function for pretraining

  #29 opened 4 years ago by kritiagg
  2

• Request for more detail for pre-training on ImageNet-21k

  #26 opened 4 years ago by junsukchoe
  5

• Extract Features

  #27 opened 5 years ago by wyp19930313
  2

• pre-trained bit_models for EfficientNet

  #12 opened 5 years ago by dashesy
  3

• Object detection pre-trained weights

  #20 opened 5 years ago by arunponnusamy
  2

• Pre-trained Resnet v1 model

  #21 opened 5 years ago by dulucas
  1

• Last layer of tf2 model is dense and a conv2d in pytorch

  #22 opened 5 years ago by Molaire
  1

• Input size for r50x1?

  #23 opened 5 years ago by PenroseTiles
  5

• Release New Model

  #25 opened 5 years ago
  1

• Model taking entire GPU

  #28 opened 5 years ago by sumedhvdatar
  1

• Model download link is broken

  #24 opened 5 years ago by mjafari-1qb
  4

• how to change this class model to segmentation

  #19 opened 5 years ago by Ontheroad123
  0

• How to fine-tune on my local dataset?

  #17 opened 5 years ago by mohhao
  1

• Clarity on range normalization of inputs

  #16 opened 5 years ago by hari-aindra
  4

• What is the BiT-M-***-ILSVRC2012

  #13 opened 5 years ago by duchengyao
  1

• Contribution

  #11 opened 5 years ago by digs1998
  0