Cufinufft torch #69
Closed
Cufinufft torch #69
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
paquiteau
force-pushed
the
cufinufft-torch
branch
from
d8422c3 to
0ec9080
Compare
Closed
* refactor: homogeneize interfaces. * feat: add bart and sigpy implementation. * feat(test): refactor test to follow changes of api. * update ci * fix(test): use truncated gaussian for sample generation. * feat(test): show available/selected backend in table format. * update CPU backend tests in CI * ignore bart and pynfft. * feat(ci): use pip cache. * install all deps in linter * update coverage steps * report steps * feat(ci): add example to test CI. * install libnfft only for pynfft. * fix typo * fix(tests): correct shape * fix(cufinufft): error in gradient computation * fix(tests): relax adjoint condition. * feat: just test for reproductibility. * remove density tests. * feat: use new interface of gpunufft. * style: only cupyx is needed. * fix: update gpunufft to follow mem-efficient. * fix(ci): add cupy to dependencies of gpunufft. * fix: remove prints. * feat: pin smaps in python code. * feat: make the pinning available for outside use. * feat: disable pinned memory pooling. * feat: use a custom allocator to share memory across context. This is mostly cargo culting. * Update src/mrinufft/operators/interfaces/gpunufft.py * fix: ruff * Fix on order and copy old stuff * Added make copy vback * Added corrections * style: run black. * style: ruff. --------- Co-authored-by: Pierre-antoine Comby <[email protected]> Co-authored-by: chaithyagr <[email protected]>
paquiteau
force-pushed
the
cufinufft-torch
branch
from
387f1d6 to
43a17f3
Compare
paquiteau
force-pushed
the
density_register
branch
from
e257a13 to
98f7a8a
Compare
paquiteau
force-pushed
the
cufinufft-torch
branch
from
43a17f3 to
18fe853
Compare
paquiteau
force-pushed
the
cufinufft-torch
branch
from
18fe853 to
459f472
Compare
paquiteau
force-pushed
the
density_register
branch
4 times, most recently
from
677055a to
63e5d1a
Compare
2 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR add support for pytorch in mri-nufft
Currently only cufinufft can use torch's Tensor on GPU (i.e. device="cuda").
Feature to implement:
The point of this PR is NOT to add autodiff support.
NB: This is built on top of the
density_registerbranch.