2543 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
Kishore Nallan	83171082b4	Tweak test.	2023-10-06 14:58:23 +05:30
Kishore Nallan	8766d87e0b	Allow wildcard with empty vector (for convenience).	2023-10-06 14:13:22 +05:30
Kishore Nallan	f175d8cce6	Add test for dropping for explicit dynamic field.	2023-10-05 20:54:52 +05:30
Kishore Nallan	6d1e9f07cf	Fix dropping a nested child field.	2023-10-05 20:17:16 +05:30
Kishore Nallan	de6e6e4a8a	Restore concurrent vec indexing.	2023-10-05 14:50:54 +05:30
Kishore Nallan	b3afb5d12c	Finer locking for res cache.	2023-10-05 14:35:54 +05:30
Kishore Nallan	67b6b00b37	Fix float stats with negative only value.	2023-10-05 14:31:06 +05:30
Kishore Nallan	3d98c1ce2d	Merge branch 'v0.25-multi-vec-patch' into v0.25-join	2023-10-02 15:33:07 +05:30
Kishore Nallan	184758954f	Merge pull request #1265 from ozanarmagan/v0.25-join ... Fix reindexing all fields on alter	2023-10-02 15:16:57 +05:30
Ozan Armağan	72bef0e776	Rename search_schemato indexable_schema	2023-10-02 12:24:15 +03:00
ozanarmagan	48b2c6bba7	Fix reindexing all fields on alter	2023-10-01 19:34:51 +03:00
Kishore Nallan	d5d9fcad1d	Add flag for not keeping missing group field values distinct.	2023-09-29 16:29:23 +05:30
Kishore Nallan	e80eba7b69	Merge pull request #1262 from ozanarmagan/v0.25-join ... Prevent using vector fields in query_by	2023-09-29 07:26:58 +05:30
ozanarmagan	5d7c0b1565	Prevent using vector fields in query_by	2023-09-29 01:09:28 +03:00
Kishore Nallan	ab59656c99	Merge pull request #1245 from ozanarmagan/v0.25-join ... Don't return `vector_distance` for keyword only matches in hybrid search	2023-09-28 18:14:41 +05:30
ozanarmagan	2d0b976f03	Do not flatten topster if not grouping	2023-09-28 14:28:09 +03:00
ozanarmagan	e7d6b56736	Use topster hashmap if not grouping	2023-09-28 14:21:07 +03:00
ozanarmagan	e747b53257	Fix using seq_id instead of kv->key	2023-09-28 13:14:27 +03:00
ozanarmagan	8eb38883b5	Use flattened KVs to find KV	2023-09-27 13:43:53 +03:00
ozanarmagan	8d481d53fd	Remove unnecessary topster sort	2023-09-27 13:12:25 +03:00
ozanarmagan	a99baf7e0a	Fix grouping	2023-09-27 11:38:04 +03:00
Kishore Nallan	b06d0d4ad0	Expose prioritize_num_matching_fields to api.	2023-09-27 12:30:14 +05:30
Kishore Nallan	4dd35835e3	Add flag prioritize_num_matching_fields.	2023-09-27 12:12:01 +05:30
Kishore Nallan	fd85c897e5	Use a single thread of indexing vectors.	2023-09-26 20:48:46 +05:30
Kishore Nallan	86c4712345	Merge pull request #1249 from krunal1313/range_facets_groupby ... adding group_by support for range_facets	2023-09-26 20:18:27 +05:30
krunal	1b4452a6e1	correcting indent	2023-09-26 14:52:46 +05:30
krunal	c42970db13	correcting group_by with range_facets	2023-09-26 14:38:28 +05:30
krunal	501e2e1c9e	adding group_by support for range_facets	2023-09-25 18:44:13 +05:30
ozanarmagan	0050134f1c	Add check for vector_distance to be positive	2023-09-25 09:02:00 +03:00
ozanarmagan	d42590c638	Don't return vector_distance for keyword only matches in hybrid search	2023-09-24 18:52:43 +03:00
Kishore Nallan	44ad5fdbcf	Merge pull request #1238 from ozanarmagan/v0.25-join ... Fix updating old documents with embeddings on alter	2023-09-24 18:20:36 +05:30
ozanarmagan	e54f680b22	Add test for one embedding and one keyword field that have same prefix	2023-09-24 14:52:48 +03:00
ozanarmagan	88f53c7f32	Add assert for .flat	2023-09-24 14:33:24 +03:00
ozanarmagan	eaef5cb8c6	Fix reindexing old documents with embeddings on alter	2023-09-21 16:38:35 +03:00
Kishore Nallan	c798966a50	Merge pull request #1234 from ozanarmagan/v0.25-join ... Fix extracting field names for two embedding fields with same prefix	2023-09-20 16:56:39 +05:30
ozanarmagan	a976127767	Fix extracting field names for two embedding fields with same prefix	2023-09-20 12:55:34 +03:00
Kishore Nallan	3abe841b7d	Add test to verify modification of vector value on upsert.	2023-09-15 20:38:50 +05:30
Kishore Nallan	5c879ccc49	Fix analytics aggregstion with interval > 60s	2023-09-15 11:43:29 +05:30
Kishore Nallan	0db3dd0d00	Don't remove vector during update: use update api directly.	2023-09-14 19:16:33 +05:30
Kishore Nallan	2af676916a	Use drop token iters directly for text match scoring. v0.25.1	2023-09-09 14:17:44 +05:30
Kishore Nallan	dc780c0f58	Add more test for partial vector update	2023-09-03 17:28:05 +05:30
Kishore Nallan	004dfa6a0f	Merge pull request #1201 from ozanarmagan/v0.25-join ... Fix tokenizing XLM-RoBERTa models with 3 inputs and altering embeddin…	2023-09-02 21:27:33 +05:30
ozanarmagan	781c5348d8	Use f instead of diff_fields.back()	2023-09-02 17:26:24 +03:00
ozanarmagan	dedb8e213d	Remove validate_and_init_embed_fields	2023-09-02 17:21:07 +03:00
ozanarmagan	16a553dd3e	Add tests & add validate_and_init_embed_field	2023-09-02 16:14:39 +03:00
ozanarmagan	29613ad054	Fix tokenizing XLM-RoBERTa models with 3 inputs and altering embedding fields	2023-09-02 01:12:21 +03:00
Kishore Nallan	3b157f6c61	Fix duplicate embed fields init.	2023-08-31 15:12:49 +05:30
Kishore Nallan	633ec69aed	Fix upsert on unchanged docs with embedding fields.	2023-08-31 13:38:08 +05:30
Kishore Nallan	6a9d5efc94	Rollback glog stdout/stderrr separation. ... This does not respect buffer levels yet: https://github.com/google/glog/issues/943	2023-08-28 12:21:38 +05:30
Kishore Nallan	c5d2efa36d	Fix preset usage regression on multi search endpoint.	2023-08-28 12:00:21 +05:30