Remove ACE in-Memory Path#1604
Closed
julianmi wants to merge 10 commits intorapidsai:mainfrom
Closed
Conversation
- Added `cuvsHnswAceParams` structure for ACE configuration. - Implemented `cuvsHnswBuild` function to facilitate index construction using ACE. - Updated HNSW index parameters to include ACE settings. - Created new tests for HNSW index building and searching using ACE. - Updated documentation to reflect the new ACE parameters and usage.
- Add heuristic to automatically derive the number of partitions based on host and device memory requirements. - Increase the user-profided `npartitions` if it does not fit memory. - Introduced `max_host_memory_gb` and `max_gpu_memory_gb` fields to `cuvsAceParams` and `cuvsHnswAceParams` structures for controlling memory usage during ACE builds. - Added tests to verify that small memory limits trigger disk mode correctly for both CAGRA and HNSW index builds.
- ACE (Augmented Core Extraction) now always uses disk-based storage for consistent behavior and memory efficiency when building large indices. - Remove 'use_disk' parameter from ace_params struct in C++, C, Python, Java - Remove in-memory code path from build_ace() in cagra_build.cuh - Remove ace_gather_partition_dataset() (no longer needed) - Consolidate ace_adjust_sub_graph_ids_disk() into ace_adjust_sub_graph_ids() - Remove conditional memory vs disk branches throughout ACE build - Update documentation to reflect disk-only behavior
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
- Renamed parameter `m` to `M` in HNSW structures and related functions for consistency. - Removed `ef_construction` from `cuvsHnswAceParams` and related classes, as it is no longer needed. - Load the HNSW index from file before search if needed.
cjnolet
added
improvement
github-project-automation
Bot
moved this from In Progress
to Done
in Unstructured Data Processing
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
2 participants
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 removes the optional in-memory path for ACE (Augmented Core Extraction) builds, making disk-based storage the only supported mode. The all neighbors API can be used in cases where partitioning is required due to limited device but sufficient host memory.
This simplifies the ACE implementation and reduces the complexity of maintaining two code paths.
Changes
use_diskparameter fromace_params.ace_gather_partition_dataset()function (in-memory only).ace_adjust_sub_graph_ids_disk()intoace_adjust_sub_graph_ids().