Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Update copyright | Jon Bratseth | 2023-10-09 | 1 | -1/+1 |
| | |||||
* | Control the auto-adjustment of targetHits in ANN using post-filtering. | Geir Storli | 2023-08-15 | 1 | -4/+18 |
| | | | | | | | | | | | | | | When searching the HNSW index in a post-filtering case, targetHits is auto-adjusted in an effort to still expose targetHits hits to first-phase ranking after post-filtering. The following formula is now used to ensure an upper bound of adjustedTargetHits, avoiding that the search in the HNSW index takes too long. adjustedTargetHits = min(targetHits / estimatedHitRatio, targetHits * targetHitsMaxAdjustmentFactor). The target-hits-max-adjustment-factor can be set in a rank profile and overriden per query. The value is in the range [1.0,inf], with the default being 20.0. When setting this to 1.0, auto-adjustment of targetHits is effectively disabled. | ||||
* | Add Doom to hnsw index. | Tor Egge | 2023-05-31 | 1 | -2/+9 |
| | |||||
* | Store max squared norm in file header during hnsw index save when using | Tor Egge | 2023-05-23 | 1 | -4/+37 |
| | | | | dotproduct distance metric. Adjust hnsw index load. | ||||
* | add mimimal version of BoundDistanceFunction | Arne Juul | 2023-04-19 | 1 | -7/+10 |
| | |||||
* | Add SerializedTensorRef. | Tor Egge | 2023-02-22 | 1 | -1/+55 |
| | |||||
* | Report address space usage for hnsw nodeid mapping. | Tor Egge | 2023-01-13 | 1 | -7/+24 |
| | |||||
* | Update address space component names. | Tor Egge | 2023-01-11 | 1 | -2/+2 |
| | |||||
* | Remove stacksize from the thread pools and thread executors. | Henning Baldersheim | 2022-12-20 | 1 | -1/+1 |
| | |||||
* | Remove out of range search::BitVector::setBit() in unit test. | Tor Egge | 2022-11-30 | 1 | -1/+0 |
| | |||||
* | Setup hnsw index for mixed tensor types. | Tor Egge | 2022-11-23 | 1 | -17/+108 |
| | |||||
* | Add enum class HnswIndexType. | Tor Egge | 2022-11-16 | 1 | -2/+3 |
| | |||||
* | Rename HnswNode to HnswTestNode. | Tor Egge | 2022-11-09 | 1 | -2/+2 |
| | |||||
* | Pass VectorBundle to NearestNeighborIndex::prepare_add_document member function. | Tor Egge | 2022-11-04 | 1 | -2/+4 |
| | |||||
* | Pass subspace to DocVectorAccess::get_vector member function. | Tor Egge | 2022-11-02 | 1 | -3/+3 |
| | |||||
* | Rename trimHoldLists() -> reclaim_memory(). | Geir Storli | 2022-10-12 | 1 | -2/+2 |
| | |||||
* | Rename transferHoldLists() -> assign_generation(). | Geir Storli | 2022-10-12 | 1 | -1/+1 |
| | |||||
* | Use paged setting for SerializedFastValueAttribute. | Tor Egge | 2022-10-10 | 1 | -33/+63 |
| | |||||
* | add support for multi-bitvector global filter | Håvard Pettersen | 2022-09-13 | 1 | -3/+3 |
| | | | | | move some testing convenience into GlobalFilter make more realistic bitvectors for filter testing | ||||
* | GlobalFilter is now an interface | Håvard Pettersen | 2022-09-12 | 1 | -1/+1 |
| | | | | instead of a shared optional BitVector | ||||
* | Refactor validation code for setting up a distance calculator for re-use in ↵ | Geir Storli | 2022-07-06 | 1 | -6/+7 |
| | | | | rank features. | ||||
* | Reduce code duplication by using fef::QueryValue in RequestContext ↵ | Geir Storli | 2022-07-05 | 1 | -4/+10 |
| | | | | implementation. | ||||
* | Revert "Revert "Don't make changes to the hnsw index when the inserted ↵ | Geir Storli | 2022-06-22 | 1 | -0/+30 |
| | | | | tensor is unchanged"" | ||||
* | Revert "Don't make changes to the hnsw index when the inserted tensor is ↵ | Geir Storli | 2022-06-13 | 1 | -30/+0 |
| | | | | unchanged" | ||||
* | Merge pull request #23051 from ↵ | Geir Storli | 2022-06-13 | 1 | -0/+30 |
|\ | | | | | | | | | vespa-engine/geirst/dont-update-hnsw-with-unchanged-tensors Don't make changes to the hnsw index when the inserted tensor is unchanged | ||||
| * | Don't make changes to the nearest neighbor index (hnsw) when the inserted ↵ | Geir Storli | 2022-06-10 | 1 | -0/+30 |
| | | | | | | | | | | | | tensor is unchanged. With this optimization we avoid doing unnecessary costly work, first removing the vector point, then inserting the same point. | ||||
* | | fix undefined behavior in unit tests | Håvard Pettersen | 2022-06-09 | 1 | -1/+1 |
|/ | |||||
* | Remove most use of vespalib::rmdir in searchlib. | Tor Egge | 2022-06-02 | 1 | -6/+6 |
| | |||||
* | Include searchcommon/attribut/config.h yourself and prepare for hiding large ↵ | Henning Baldersheim | 2022-05-25 | 1 | -6/+7 |
| | | | | | | implementation. Add setters to vespalib::GrowStrategy. | ||||
* | Auto-adjust targetHits for approximate nearest neighbor search when using ↵ | Geir Storli | 2022-05-10 | 1 | -11/+20 |
| | | | | | | | | post-filtering. The goal is to expose 'targetHits' hits to first-phase ranking. Before searching the HNSW index, targetHits is adjusted based on the estimated hit ratio of the query to compensate for the hits that will be removed in post-filtering. | ||||
* | Improve and extend visit trace for nearest neighbor blueprint. | Geir Storli | 2022-04-22 | 1 | -5/+0 |
| | |||||
* | Simplify calculation of global filter and fallback to brute force when using ↵ | Geir Storli | 2022-04-11 | 1 | -2/+13 |
| | | | | | | | | nearest neighbor search. With this change only two parameters are used to adjust the algorithm: - vespa.matching.global_filter.lower_limit - vespa.matching.global_filter.upper_limit | ||||
* | - assert result of Close() in destructor of FastOS_File and FastOS_BufferedFile. | Henning Baldersheim | 2022-01-26 | 1 | -1/+0 |
| | | | | | | | - Check result of Close() - Check result of Sync() - Scope FastOS_File to avoid explicit Close(). - NULL -> nullptr | ||||
* | Calculate need for compaction when updating stats. | Tor Egge | 2021-12-09 | 1 | -1/+2 |
| | |||||
* | Move CompactionStrategy from searchcommon to vespalib. | Tor Egge | 2021-12-08 | 1 | -1/+1 |
| | |||||
* | Update 2017 copyright notices. | gjoranv | 2021-10-07 | 1 | -1/+1 |
| | |||||
* | Change loading of nearest neighbor index to use direct I/O instead of mmapping. | Geir Storli | 2021-09-15 | 1 | -8/+6 |
| | | | | This should reduce memory spike during loading. | ||||
* | Reduce memory spike during loading of hnsw index by committing at regular ↵ | Geir Storli | 2021-09-10 | 1 | -4/+22 |
| | | | | | | | intervals. This ensures that memory buffers on hold lists are trimmed while loading, keeping the excess memory usage at a minimum. | ||||
* | The global filter is only needed when having a nearest neighbor index (hnsw) ↵ | Geir Storli | 2021-08-30 | 1 | -17/+44 |
| | | | | | | and doing approximate calculation. This avoids costly calculation of the global filter in cases it is not needed. | ||||
* | Report address space usage for shared string repo for non-dense tensor ↵ | Geir Storli | 2021-08-23 | 1 | -2/+8 |
| | | | | attributes. | ||||
* | Report address space usage for components in tensor attributes. | Geir Storli | 2021-08-20 | 1 | -0/+23 |
| | |||||
* | Compact HNSW index when ratio of dead bytes / address space is too high | Tor Egge | 2021-08-18 | 1 | -0/+8 |
| | | | | relative to used bytes / address space. | ||||
* | Use the executor for the part that can be parallell when rebuilding index on ↵ | Henning Baldersheim | 2021-08-13 | 1 | -0/+33 |
| | | | | load. | ||||
* | swappable -> paged | Henning Baldersheim | 2021-08-12 | 1 | -2/+2 |
| | |||||
* | A swappable attribute will use a file backed memory allocator. | Henning Baldersheim | 2021-08-12 | 1 | -0/+23 |
| | |||||
* | Reenable tensor attribute compaction test. | Tor Egge | 2021-06-01 | 1 | -8/+10 |
| | |||||
* | let the distance function decide how it wants the query tensor converted | Arne Juul | 2021-04-12 | 1 | -1/+1 |
| | |||||
* | fix NNS distance for new cell types | Arne Juul | 2021-04-12 | 1 | -1/+1 |
| | | | | This reverts commit f167fe4362c5e4e20a7605b99205cfbee77c569a. | ||||
* | Revert "fix NNS distance for new cell types" | Arnstein Ressem | 2021-04-12 | 1 | -1/+1 |
| | |||||
* | fix NNS distance for new cell types | Arne Juul | 2021-04-09 | 1 | -1/+1 |
| |