Define __matmul__ operator.

[codecov[bot]]

Codecov Report

Merging #37 (cf6696a) into master (4656839) will not change coverage. The diff coverage is 100.00%.

@@            Coverage Diff            @@
##            master       #37   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           16        16           
  Lines         1753      1743   -10     
=========================================
- Hits          1753      1743   -10     

Impacted Files	Coverage Δ
eagerpy/tensor/jax.py	100.00% <ø> (ø)
eagerpy/tensor/numpy.py	100.00% <ø> (ø)
eagerpy/tensor/pytorch.py	100.00% <ø> (ø)
eagerpy/tensor/tensorflow.py	100.00% <ø> (ø)
eagerpy/tensor/tensor.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 4656839...cf6696a. Read the comment docs.

[jonasrauber]

Nice, this allows us to get rid of the framework-specific code duplication for matmul 👍 I guess this breaks compatibility with some older versions of PyTorch, TensorFlow, NumPy that didn't support __matmul__, but I think we can live with that.

[jonasrauber]

resolves #35