heat

heat copied to clipboard

Implement sparse matrices

1

**Feature functionality** Design and implement a distributed sparse DNDarray class. Come up with one or more reasonable storage formats (e.g. COO, CRS, CSC, ...) **Additional context** PyTorch implements COO only...

Markus-Goetz

Dockerfile and accompanying documentation

2

The Dockerfile provides some flexibility in selecting which version of HeAT should be inside the Docker image. Also, one can choose whether to install from source or from PyPI. ##...

bhagemeier

Features/30 convolve2d

2

## Description Issue/s resolved: # ## Changes proposed: - - - - ## Type of change ## Memory requirements ## Performance ## Due Diligence - [ ] All split configurations...

krajsek

Features/589 batch normalization

4

## Description Issue/s resolved: # ## Changes proposed: - - - - ## Type of change ## Due Diligence - [ ] All split configurations tested - [ ] Multiple...

krajsek

**Related** Related to advanced indexing in the `setitem`. **Feature functionality** [See torch definition.](https://pytorch.org/docs/stable/generated/torch.Tensor.index_put_.html) Essentially, its the same as advanced indexing, however with the option to deal with duplicate indices by...

ben-bou

Currently branches off `main` **Expected behavior** Should branch off `releases/1.2.x`

ClaudiaComito

## Description Issue/s resolved: # ## Changes proposed: - - - - ## Type of change ## Memory requirements ## Performance ## Due Diligence - [ ] All split configurations...

krajsek

Cut down memory requirements for same-split reshape where possible

3

## Description When reshaping distributed DNDarrays: - if `new_split` is the same as the original split, and - if distribution (lshapes) allows then reshape locally via pytorch, stitch `local_reshaped` tensors...

ClaudiaComito

Add float16 support to types

5

Markus-Goetz

Extend kwargs `out`, `where` to all binary operations

2

**Related** #870 **Feature functionality** As per #945, it is possible to pass a condition to `_operations.__binary_op` via the kwarg `where`. This feature is available in `ht.divide`. Extend feature to all...

ClaudiaComito