IntelLabs/bayesian-torch

Issues

• There are questions about function fitting (variational Bayesian network)

  #37 opened a month ago by dashangyanhua
  2

• Regression tasks: Epistemic and Aleatoric Uncertainty Estimation

  #22 opened 2 years ago by feracero
  8

• Inconsistent use of mean & sum when calculating KL divergence?

  #36 opened 8 months ago by profPlum
  0

• number of MC samples for training and inference

  #31 opened a year ago by dzy666fly
  2

• problem in conv_transpose2d

  #34 opened a year ago by amirshamaei
  1

• Error in trans_conv2d when model is reparametrized

  #21 opened a year ago by rameshamurtaza
  2

• local variable 'kl' referenced before assignment when using flipout in dnn_to_bnn()

  #24 opened a year ago by barghit
  3

• Implement quant prepare function

  #29 opened a year ago by ranganathkrishnan
  0

• dnn_to_bnn() does not support Conv2d with non-square kernels

  #32 opened a year ago by bryanmooremd
  4

• Quantizatize bayesian cifar runtime error

  #27 opened a year ago by szhaoesat
  4

• Calculation of ELBO in Bayesian layers seems to mismatch the equation in the original paper

  #26 opened a year ago by dzy666fly
  3

• How bayesian network be used in quantization?

  #5 opened a year ago by LeopoldACC
  2

• Wrong output for regression

  #25 opened a year ago by AmosJoseph
  3

• conv_transpose2d parameter order problem

  #13 opened 2 years ago by pierreosselin
  4

• KL divergence not changing during training

  #19 opened 2 years ago by gkwt
  4

• Kernel_size

  #17 opened 2 years ago by flydephone
  1

• Batchnorm layer

  #16 opened 2 years ago by fzimmermann89
  1

• Variance/standard deviation

  #15 opened 2 years ago by jeiglsperger
  1

• The average should be taken over log probability rather than logits

  #10 opened 2 years ago by Nebularaid2000
  5

• dnn_to_bnn() with LSTM network

  #14 opened 2 years ago by jeiglsperger
  1

• kl_loss function in Convolutional flipout layers

  #12 opened 3 years ago by ranganathkrishnan
  1

• FR: Enable forward method of Bayesian Layers to return value only for smoother integration with PyTorch

  #7 opened 3 years ago by piEsposito
  2

• Enable the Bayesian layer to freeze the parameters to their mean values

  #8 opened 3 years ago by Nebularaid2000
  2

• modifing the Scripts for any arbitrary output size

  #4 opened 3 years ago by narminGhaffari
  2

• Pytorch version issue

  #6 opened 3 years ago by hebowei2000
  4

• add mu_kernel to of delta_kernel in flipout layers?

  #3 opened 3 years ago by burntcobalt
  2

• Shape error during LSTM forward pass

  #1 opened 4 years ago by anirudhs96
  0