danielward27/flowjax

Issues

• compatibility with python<3.10

  #194 opened a month ago by jackaraz
  1

• Q: Normal Distribution Compatibility with Reparameterization Trick?

  #181 opened 2 months ago by gil2rok
  2

• Doctest fails

  #155 opened 3 months ago by danielward27
  1

• Autodiff problem with block_neural_autoregressive_flow

  #176 opened 3 months ago by aseyboldt
  1

• AutoregressiveBisectionInverter bounds picked up as trainable parameters

  #173 opened 3 months ago by mdmould
  1

• Implementing inverse in Planar flow

  #169 opened 5 months ago by weiyaw
  4

• masked autoregressive flow with mixed transformer types

  #161 opened 7 months ago by llaurabatt
  4

• ENH: use jaxtyping

  #150 opened 6 months ago by nstarman
  4

• Infinite Loss When Using Uniform Base Distribution to Model Gaussian Data with MAF

  #147 opened 9 months ago by llaurabatt
  7

• Add docs section on serialization

  #149 opened 9 months ago by nstarman
  1

• Cross-validation/Shuffling Test/Train split option in fitting utility functinos

  #146 opened 9 months ago by joshuawchen
  1

• Concatenate transform incompatible with jax tracing

  #132 opened a year ago by mdmould
  0

• Concatenate fails when stacking more than two bijections

  #130 opened a year ago by mdmould
  0

• Citing your package

  #124 opened a year ago by mdmould
  1

• Transformation for conditioning variables

  #98 opened a year ago by vallis
  1

• BlockNeuralAutoregressiveFlow not initially normalized correctly

  #102 opened a year ago by mdmould
  7

• flowjax not available after install

  #99 opened a year ago by mdmould
  1

• Add a "sticking the landing" loss for VI

  #66 opened a year ago by Tennessee-Wallaceh
  3

• Better support for bounded/constrained distributions

  #74 opened 2 years ago by danielward27
  0

• Consider adding orthogonal transformations

  #70 opened 2 years ago by Tennessee-Wallaceh
  2

• Order of arguments to Distribution.sample

  #62 opened 2 years ago by Tennessee-Wallaceh
  4

• Add Variational Inference training helper

  #60 opened 2 years ago by Tennessee-Wallaceh
  3

• Loss function in train utils isn't "jitted"

  #35 opened 2 years ago by Tennessee-Wallaceh
  1

• Add more distributions

  #29 opened 2 years ago by Tennessee-Wallaceh
  0