Skip to content

Conversation

@OmarWagih1
Copy link
Contributor

The state of the repository right now is that most tests pass (only about 30 tests fail) with a default backend of CuPy.

All failed tests are mentioned in this document with the reason of their failure

https://docs.google.com/document/d/1Ag7Yh6Ovi1zKwZSv2p-ckwC6gTuZrzb5CUUt3f7iShc/edit

Some tests fail because the randomization between CuPy and NumPy is different, so while the code works correctly the tests do not pass. Some functions are not implemented with CuPy backend and others are not compatible with CuPy.

The code needs working on to be better looking as some Numpy/Cupy switches were crudely done.

MichaelMauderer and others added 17 commits August 24, 2020 12:44
…, coordinates, temperature, difference, biochemistry, colorimetry and algebra. Currently parts of the last two modules do not work.
…se functions, continous/signal-Multisignal, rgb/derivation, temperature CCT_to_xy, XYZto_sd_Meng2015, notation/munsell and primitive_grid/cube
…y instead of numpy, redefined the as_float for numpy to avoid the zero division error, currently all tests work on numpy and some of them do not work on cupy as shared in the document
@KelSolaar KelSolaar changed the title Feature/ndarray backend PR: CuPy Based GPU Backend Aug 29, 2020
@KelSolaar
Copy link
Member

Thanks @OmarWagih1!

Would it be possible to rebase this on develop so that there are only the commits pertaining to the GPU backend please?

@KelSolaar KelSolaar changed the base branch from feature/ndarray_backend to develop August 29, 2020 23:25
@KelSolaar KelSolaar marked this pull request as draft November 4, 2020 08:50
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants