heat

heat copied to clipboard

Provide containerized HeAT

3

**Related** We've received several requests of potential users who would like to use HeAT from containers. The applications range from development a simple development setup to embedding HeAT in complex...

bhagemeier

Replace scattered scikit-learn references with central BSD-3 clause reference

9

**Feature functionality** Remove the scattered references to scikit learn (in the docstrings) and replace it with a BSD-3 clause reference.

Markus-Goetz

Documentation - add programming model page

5

**Feature functionality** Add a page describing HeAT's programming model.

Markus-Goetz

Documentation - add 'Hello World' page

2

**Feature functionality** Add a page in the documentation showing a small 'Hello World' usage example.

Markus-Goetz

ht.allclose should factor in the dtype of the inputs when determining the limits

1

**Feature functionality** see title **Motivation** When using ht.allclose with float32, we expect to have lower precision. But the current limits on `ht.allclose` are derived from numpy (https://numpy.org/doc/stable/reference/generated/numpy.allclose.html) which uses 64...

coquelin77

batch normalization

6

Implementation of a parallel version of Batch Normalisation for Model and Data parallization

krajsek

setitem does not account for key changing the split axis

3

**Description** Setitem checks `self.split` which is not correct because `key` may change the split axis if it is an `int` and smaller than split Correct behaviour would be to check...

ben-bou

**Related** #891 The `heat` function `stride_tricks.broadcast_shape` differs unnecessarily from its numpy and torch counterparts: 1. naming (heat `broadcast_shape` vs. numpy/torch `broadcast_shapes`) 2. calling (`ht.stride_tricks.broadcast_shape()` vs. `np.broadcast_shapes()` or `torch.broadcast_shapes()`) 3. arguments:...

ClaudiaComito

Improve load-functionality: load multiple files into one DNDarray

4

**Feature functionality** In HPC data analytics we often encounter the problem that there is not one large .h5-file to be processed, but instead many many single files (e.g., csv or...

coquelin77

**Feature functionality** Inspired by the [General dataset API](https://scikit-learn.org/stable/datasets.html#dataset-loading-utilities) of scikit-learn, it would be good to have something similar for HeAT. I do not mean to re-implement what's already there, but...

bhagemeier