main -> develop (Release 0.4.1) #209
Merged
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
release -> main (for v0.4.0)
Try allowing to skip coverage comment
The test execution shows warnings about traces being potentially incorrect because the Python3 control flow is not completely recorded. This includes conditions on the shape of the integration domain tensor. Since the only arguments of the compiled integration function are the integrand and integration domain, and the dimensionality of this integration domain is constant, we can ignore the warnings. After this change, the two `get_jit_compiled_integrate` functions hide PyTorch trace compilation warnings with `warnings.filterwarnings`.
Hide PyTorch trace compilation warnings
Update .readthedocs.yml
fixed a bug.
I fixed a bug for tutorial.rst.
Update tutorial.rst
Update tutorial.rst
The previous check if code is currently being compiled no longer works with new TensorFlow versions because the `Tensor` type is now called `SymbolicTensor`. This change adds a helper function to check if code is being compiled for JAX, TensorFlow or PyTorch. If tf.is_symbolic_tensor() is available, i.e. if the TensorFlow version is high enough, we use this function to check if code is being compiled. To avoid inconsistencies between backends, the check for integration domain values is disabled if code is being compiled with PyTorch even if the check works with PyTorch.
Fix JIT compilation with TensorFlow >= 2.14.0
Switch black version
Fixing mamba setup in CI
|
Merging this and will continue to fix CI errors on |
Merged
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Outdated suggestions cannot be applied.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.
Description
Summary of changes
This pull request includes several changes to update dependencies, improve compatibility, and add new utility functions. The most important changes include updating Python versions in workflows, improving integration functions, and adding new utility functions for handling compilation checks and tracing.
Workflow Updates:
.github/workflows/autoblack.yml: Updated Python version from 3.8 to 3.11 and specified the version of Black to 24.4.2..github/workflows/run_tests.yml: Addedcontinue-on-error: trueto the pytest coverage comment step.Integration Function Improvements:
docs/source/tutorial.rst: Updated calls tointegrator.calculate_resultto include theintegration_domainparameter.torchquad/integration/grid_integrator.py: Replaced direct calls totorch.jit.tracewith_torch_trace_without_warningsto avoid warnings. [1] [2]torchquad/integration/monte_carlo.py: Replaced direct calls totorch.jit.tracewith_torch_trace_without_warningsto avoid warnings. [1] [2]New Utility Functions:
torchquad/integration/utils.py: Added_is_compilingto check if code is being compiled with PyTorch, JAX, or TensorFlow.torchquad/integration/utils.py: Added_torch_trace_without_warningsto executetorch.jit.tracewhile hiding tracer warnings.Documentation and Testing:
.readthedocs.yml: Updated build configuration to use Ubuntu 22.04 and mambaforge-22.9 for Python.torchquad/tests/utils_integration_test.py: Added tests for the new_is_compilingutility function.Resolved Issues
N/A
How Has This Been Tested?
Related Pull Requests
N/A