[ENH] Parallelize solver input preparation using ThreadPoolExecutor#55
Open
Leguark wants to merge 4 commits intooptimize_grad_vs_scalarfrom
Open
[ENH] Parallelize solver input preparation using ThreadPoolExecutor#55Leguark wants to merge 4 commits intooptimize_grad_vs_scalarfrom
Leguark wants to merge 4 commits intooptimize_grad_vs_scalarfrom
Conversation
This was referenced Mar 13, 2026
Member
Author
|
Warning This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
This stack of pull requests is managed by Graphite. Learn more about stacking. |
This was referenced Mar 13, 2026
Leguark
force-pushed
the
optimize_yield_cov
branch
from
b246cfc to
93929d4
Compare
Leguark
force-pushed
the
optimize_grad_vs_scalar
branch
from
e67c8a4 to
29a2e9a
Compare
- Refactored stack operations to enable parallel preparation of solver inputs. - Introduced thread-safe handling of `stack_structure` to avoid race conditions. - Improved efficiency and scalability by replacing sequential loops with concurrent execution.
- Added conditionals for GPU usage to streamline tensor preparation and movement. - Improved contiguity enforcement and asynchronous GPU transfers in evaluator workflows. - Enhanced backend configuration logic by refining `BackendTensor` methods for modularity. - Fixed minor formatting and consistency issues across backend and evaluator modules.
…ization - Implemented `_compute_weights_for_stacks_single_thread` to support stack weight calculations. - Added explicit GPU synchronization for enhanced eGPU stability. - Resolved minor formatting inconsistencies across evaluator and stack modules.
- Enhanced tensor handling by enabling `.to("cuda")` during tensor preparation in `symbolic_evaluator`.
- Introduced `BackendTensor.clear_gpu_memory()` to explicitly manage GPU memory.
- Added `gc.collect()` for improved resource cleanup after computations.
Leguark
force-pushed
the
optimize_grad_vs_scalar
branch
from
29a2e9a to
b45b42e
Compare
Leguark
force-pushed
the
optimize_yield_cov
branch
from
93929d4 to
641cb34
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
[ENH] Parallelize solver input preparation using ThreadPoolExecutor
stack_structureto avoid race conditions.[ENH] Refactor GPU tensor handling and evaluator logic
BackendTensormethods for modularity.[ENH] Add optimized stack weight computation and improve GPU synchronization
_compute_weights_for_stacks_single_threadto support stack weight calculations.