From f868c7352c64aad4c2bcc34dc4769e80c7893782 Mon Sep 17 00:00:00 2001 From: Vahan Stepanyan Date: Wed, 29 Jan 2025 12:06:32 +0000 Subject: [PATCH 1/3] Updated BM25, text and hybrid search for Activeloop Deeplake --- docs/tools/vdb_table/data/activeloop.json | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/docs/tools/vdb_table/data/activeloop.json b/docs/tools/vdb_table/data/activeloop.json index 5384da8ee..8ad7c5d33 100644 --- a/docs/tools/vdb_table/data/activeloop.json +++ b/docs/tools/vdb_table/data/activeloop.json @@ -33,9 +33,9 @@ "comment": "" }, "hybrid_search": { - "support": "", - "source_url": "https://docs.activeloop.ai/performance-features/querying-datasets/query-syntax", - "comment": "While you can run embedding search + contains(text, 'keywoard') or multiple those (keyword search inside text tensor) since BM25 not available I wouldn't call a full hybrid search. https://docs.activeloop.ai/performance-features/querying-datasets/query-syntax" + "support": "full", + "source_url": "https://docs.deeplake.ai/4.0/guide/rag/", + "comment": "BM25 with ANN" }, "facets": { "support": "", @@ -58,14 +58,14 @@ "comment": "no native sparse vector support, although it supports all numpy arrays hence can also store sparse numpy arrays" }, "bm25": { - "support": "none", - "source_url": "", + "support": "full", + "source_url": "https://docs.deeplake.ai/latest/", "comment": "" }, "full_text": { - "support": "partial", - "source_url": "", - "comment": "you can search keywords with TQL `contains(...)` function" + "support": "full", + "source_url": "https://docs.deeplake.ai/4.0/guide/rag/", + "comment": "via Inverted Index and BM25" }, "embeddings_text": { "support": "partial", From e0258100e9e20dec75e4a765149a01e14eaf05bd Mon Sep 17 00:00:00 2001 From: Vahan Stepanyan Date: Fri, 31 Jan 2025 13:33:32 +0400 Subject: [PATCH 2/3] Update docs/tools/vdb_table/data/activeloop.json Co-authored-by: Prashanth Rao <35005448+prrao87@users.noreply.github.com> --- docs/tools/vdb_table/data/activeloop.json | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/tools/vdb_table/data/activeloop.json b/docs/tools/vdb_table/data/activeloop.json index c48bb5dfc..0916fa424 100644 --- a/docs/tools/vdb_table/data/activeloop.json +++ b/docs/tools/vdb_table/data/activeloop.json @@ -59,7 +59,7 @@ }, "bm25": { "support": "full", - "source_url": "https://docs.deeplake.ai/latest/", + "source_url": "https://docs.deeplake.ai/latest/#key-features", "comment": "" }, "full_text": { From a0d67df9b14248360d22927e3cb2fbcd4a5d1792 Mon Sep 17 00:00:00 2001 From: Prashanth Rao <35005448+prrao87@users.noreply.github.com> Date: Sat, 1 Feb 2025 16:07:24 -0500 Subject: [PATCH 3/3] Apply suggestions from code review --- docs/tools/vdb_table/data/activeloop.json | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/tools/vdb_table/data/activeloop.json b/docs/tools/vdb_table/data/activeloop.json index 0916fa424..9c0aa603a 100644 --- a/docs/tools/vdb_table/data/activeloop.json +++ b/docs/tools/vdb_table/data/activeloop.json @@ -34,7 +34,7 @@ }, "hybrid_search": { "support": "full", - "source_url": "https://docs.deeplake.ai/4.0/guide/rag/", + "source_url": "https://docs.deeplake.ai/latest/guide/rag/", "comment": "BM25 with ANN" }, "facets": { @@ -64,7 +64,7 @@ }, "full_text": { "support": "full", - "source_url": "https://docs.deeplake.ai/4.0/guide/rag/", + "source_url": "https://docs.deeplake.ai/latest/api/query/", "comment": "via Inverted Index and BM25" }, "embeddings_text": {