- Removed redundant code from
reduce_precision
. - Added
types
argument toPseudoParameter.parametrize_module
for better selection for Parameterising the Layers.
- Fixed
GeLU
backward function equation.
Model
is subclass ofBackwardModule
for additional functionality.- Using
inspect.isclass
to check ifbackward_class
is a class inLinear.set_backward_function
. - Repr using
self.__class__.__name__
in all classes.
- Removed unnecessary
PseudoParameter.grad
property. - Patch for Pytorch 2.0.1, add filtering inputs in
BackwardGraph._calculate_gradients
.
- Combined
PseudoParameter
andPseudoParameterModule
for better visibility.- BugFix: fixed save and load of state_dict of
PseudoParameter
and transformation module.
- BugFix: fixed save and load of state_dict of
- Removed redundant class
analogvnn.parameter.Parameter
.
- Added support for no loss function in
Model
class.- If no loss function is provided, the
Model
object will use outputs for gradient computation.
- If no loss function is provided, the
- Added support for multiple loss outputs from loss function.
- Bugfix: removed
graph
fromLayer
class.graph
was causing issues with nestedModel
objects.- Now
_use_autograd_graph
is directly set while compiling theModel
object.
- added
grad.setter
toPseudoParameterModule
class.
- Public release.