eagerpy

New Features

• added ep.utils.get_dummy
• TensorFlowTensor.sum now automatically casts boolean to int (like PyTorchTensor and others)
• other improvements and new tests

New Features

• added __mod__
• fixed ep.mean (was sum instead of mean)
• reorganized the tests
• completed our test suite to reach 100% test coverage

New Features

• 99% test coverage
  • all tests rewritten from scratch
  • removed all framework-specific duplication
  • added nice helper functions
• truely lazy imports of the different frameworks
• added magic framework wrappers that automatically apply ep.astensor and preform lazy imports
  • ep.torch
  • ep.tensorflow
  • ep.numpy
  • ep.jax
• added pow()
• added log_softmax()
• added kl_div_with_logits()
• lots of bugfixes and improvements

New Features

• added support for automatic differentation of arbitrary functions that take and return EagerPy tensors
  • value_and_grad (function and method)
  • value_aux_and_grad (function and method)
  • value_and_grad_fn (function)
• added crossentropy
• improved the __repr__() shown for EagerPy tensors
• other improvements and bug fixes

New Features

• added __format__ support to tensors
• added full
• added full_like
• added arange
• added index_update
• added cumsum
• added newaxis
• added flip
• added meshgrid
• added pad
• added nan
• added inf
• added isnan
• added isinf

Plus: Lot's of bug fixes and improvements.

New Features

• added softmax
• added stack
• added squeeze and expand_dims

New Features

• added where
• added tile
• added matmul
• added logical_not

New Features

• added exp, log, log2, log10, log1p
• bug fixes

New Features

• added bool()
• added all() and any()
• added logical_and() and logical_or()

New Features

• added support for concatenate
• added support for transpose
• fixed abs
• other bug fixes

New Features

• added support for onehot_like, from_numpy, ones, zeros, ones_like, zeros_like, __rsub__
• improved __getitem__ support for TensorFlow (to match PyTorch and JAX)
• lot's of bugfixes
• astensor can now handle JAX arrays

New Features

• Support for JAX 0.1.48 (JAXTensor)
• support for Tensor.item()
• support for Tensor.float32()
• support for Tensor.__bool__()
• new istensor() function
• and lot's of bugfixes

New Features

• eagerpy.astensor() that automatically wraps it's input using the appropriate EagerPy tensor

Supported Frameworks

• PyTorch 1.3.0 (PyTorchTensor)
• TensorFlow 2.0.0 (TensorFlowTensor)
• NumPy 1.17.3 (NumPyTensor)