Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Feature/499 datainf #582

Merged
merged 44 commits into from
Jun 6, 2024
Merged

Feature/499 datainf #582

merged 44 commits into from
Jun 6, 2024

Conversation

schroedk
Copy link
Collaborator

@schroedk schroedk commented May 21, 2024 •
edited
Loading

Description

This PR closes #499 closes #585

Changes

Checklist

  • Wrote Unit tests (if necessary)
  • Updated Documentation (if necessary)
  • Updated Changelog
  • If notebooks were added/changed, added boilerplate cells are tagged with "tags": ["hide"] or "tags": ["hide-input"]

schroedk added 18 commits May 21, 2024 16:54
@schroedk
Add new module influence.types containing generic types to build up i…
93b6ecd 
…nfluence computation

in a very flexible way
@schroedk
Refactor influence.base_influence_function_model:
db0fa07 
* move types to influence.types
* add fit_required decorator
* add class ComposableInfluence, based on generic types from influence.types
* add class SumAggregator to influence.array
@schroedk
Merge remote-tracking branch 'origin/develop' into feature/499-datainf
23fed56
@schroedk
Fix linting issues in influence.types
c4b7d3a
@schroedk
Add class ModelInfoMixin to influence.torch.util
c189848
@schroedk
Add class TorchBatch to influence.torch.util
bda3504
@schroedk
Add classes TorchChunkAverageAggregator and TorchPointAverageAggregat…
ddec527 
…or to influence.torch.util
@schroedk
Add functions rank_one_mvp and inverse_rank_one_update to influence.t…
7472840 
…orch.util
@schroedk
Add LossType (annotation), enum BlockMode and class ModelParameterDic…
e32ae46 
…tBuilder to influence.torch.util
@schroedk
Add dtype property to LowRankProductRepresentation
d804e6b
@schroedk
Add new subpackage influence.torch.operator with submodule operator.g…
993bcfa 
…radient_provider
@schroedk
Add operator subpackages batch_operation, add tests for bnatch_operat…
3292253 
…ion and gradient_provider
@schroedk
Add submodules operator.base, operator.bilinear_form and operator.solve
ec57282
@schroedk
Add tests for influence.torch.util
5e37a4b
@schroedk
Add implementation of generic types to influence.torch.influence_func…
29eb69a 
…tion_model, implement InverseHarmonicMeanInfluence and add corresponding test
@schroedk
Fix linting issues in operator subpackage
07f80bd
@schroedk
Fix type-checking issues
3b7289c
@schroedk
Add and improve docstrings
531aacc
@schroedk schroedk force-pushed the feature/499-datainf branch from b0e297a to 531aacc Compare May 22, 2024 15:40
schroedk added 11 commits May 22, 2024 19:45
@schroedk
Simplify package structure:
1e5f18b 
* move operator submodules to influence.torch level
* move implementations of generic classes to influence.torch.base
@schroedk
Fix linting issues
964f566
@schroedk
Move test modules
3fa9a95
@schroedk
Add module description to pydvl.influence.types
a1d4c9a
@schroedk
Update reference bibtex file
545308a
@schroedk
Refactor influence package:
6337a5f 
* improve naming
* add and extend documentation
* simplify classes
@schroedk
Add inverse rank-1 update approximation for dictionary input
f007507
@schroedk
Refactor new structure:
ed5f14d 
* renaming classes
* add cocept 'TensorDictOperator', which can act on tensor dictioniaries, to avoid
  intermeditate flatten and concat to reduce memory consumption
@schroedk
Move functions from util to static methods, increase tolerance for test
f4f5931
@schroedk
Simplify operator interface
01791d5
@schroedk
Simplify tensor summation
79f7569
@schroedk schroedk marked this pull request as ready for review May 29, 2024 16:11
@schroedk schroedk changed the title WIP: Feature/499 datainf Feature/499 datainf May 29, 2024
@schroedk schroedk added this to the v0.11.0 milestone May 29, 2024
@schroedk schroedk mentioned this pull request Jun 3, 2024
Closed
This was referenced Jun 3, 2024
Closed
Closed
Closed
Closed
@schroedk schroedk requested a review from janosg June 3, 2024 09:34
@schroedk schroedk mentioned this pull request Jun 3, 2024
Merged
4 tasks
@janosg janosg mentioned this pull request Jun 5, 2024
Open
janosg
janosg reviewed Jun 5, 2024
View reviewed changes
Copy link
Collaborator

@janosg janosg left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think the new abstractions in the types module are really good and will also work well for JAX. The rest are minor comments.

src/pydvl/influence/torch/base.py Show resolved Hide resolved
src/pydvl/influence/torch/base.py Show resolved Hide resolved
src/pydvl/influence/torch/util.py Show resolved Hide resolved
src/pydvl/influence/torch/util.py Outdated Show resolved Hide resolved
tests/influence/torch/test_batch_operation.py Outdated Show resolved Hide resolved
@schroedk
Copy link
Collaborator Author

schroedk commented Jun 5, 2024 •
edited
Loading

Fix issue in GradientProvider: not exposing dict structure of gradients, as this might be framework specific.

Done

@schroedk schroedk requested a review from janosg June 5, 2024 16:42
@schroedk schroedk merged commit 347b55a into develop Jun 6, 2024
35 checks passed
@schroedk schroedk deleted the feature/499-datainf branch June 6, 2024 08:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@janosg janosg janosg approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
v0.11.0
2 participants
@schroedk @janosg