-
90ff5281 · Try 1) to increase epochs number 2) to use the trained encoder of t...
- ... and 1 more commit. Compare 6315ecb6...90ff5281
-
6315ecb6 · better naming for activation function
- ... and 78 more commits. Compare dfdce859...6315ecb6
-
dfdce859 · Play with the init of (planar) encoder NN weights to avoid constant...
- ... and 10 more commits. Compare 97f0d18d...dfdce859
-
97f0d18d · working version of notebook on toy example
- ... and 7 more commits. Compare a2c9a812...97f0d18d
-
a2c9a812 · Modif expe with learnable Gaussian base distribution.
- ... and 7 more commits. Compare ab71b17f...a2c9a812
-
ab71b17f · simulation setting with (2, 2) as offset, no covariate. PlnSto work...
- ... and 4 more commits. Compare 7207abf8...ab71b17f
-
e3663db3 · different optimizers for variational vs model parameters
- ... and 3 more commits. Compare 850d8e98...e3663db3
-
850d8e98 · implement GD on covaraince (C C^T parameterization) and coef
- ... and 2 more commits. Compare 2d431fbc...850d8e98
created project
JOUVIN Nicolas / pyPLNmodels