Commit Graph

117 Commits

Author SHA1 Message Date
Timothy J. Baek
797afd0b72 fix: embedding metadata issue 2024-10-13 03:25:11 -07:00
Timothy J. Baek
92605fd59f refac 2024-10-13 03:16:18 -07:00
Timothy J. Baek
5ffd216fca refac 2024-10-13 03:02:02 -07:00
Timothy J. Baek
dff3732fcd enh: tiktoken/token splitter support 2024-10-13 02:07:50 -07:00
Timothy J. Baek
5273dc4535 refac 2024-10-13 00:21:06 -07:00
Timothy J. Baek
333317a7ce refac: add embedding metadata 2024-10-12 18:30:21 -07:00
Jannik Streidl
f47c9c69e3 Merge branch 'upstream-dev' into dev 2024-10-12 15:18:59 +02:00
Jannik Streidl
741230bcdb fix 2024-10-10 17:20:50 +02:00
Jannik Streidl
89c77f05a8 chromadb switch to cosine similarity 2024-10-10 16:46:14 +02:00
Timothy Jaeryang Baek
d7a71f3b34
Merge pull request #6050 from RobinBially/main
feat: add qdrant vector db connector
2024-10-09 12:12:48 -07:00
Timothy J. Baek
c157004e07 refac 2024-10-09 12:05:16 -07:00
Timothy J. Baek
451f1bae15 refac: embeddings function 2024-10-09 11:41:35 -07:00
Robin Bially
34150fc3ed add NO_LIMIT var 2024-10-09 18:34:04 +02:00
Robin Bially
54dc94317c improvements 2024-10-09 18:29:14 +02:00
Robin Bially
2c59f2dcaf fix critical bug 2024-10-09 13:23:04 +02:00
Robin Bially
b56f77ed47 improvements 2024-10-09 13:10:23 +02:00
Robin Bially
878a570a2c add qdrant as vector db 2024-10-09 12:51:43 +02:00
Timothy J. Baek
3391a855f0 fix 2024-10-08 17:03:42 -07:00
Timothy J. Baek
dcab991d44 fix: model dump issue 2024-10-08 11:05:57 -07:00
Peter De-Ath
885b9f1ece refactor: Update GenerateEmbeddingsForm to support batch processing
refactor: Update embedding batch size handling in RAG configuration

refactor: add query_doc query caching

refactor: update logging statements in generate_chat_completion function

change embedding_batch_size to Optional
2024-10-08 00:04:35 +01:00
Timothy J. Baek
48e7f47558 refac 2024-10-07 14:03:42 -07:00
Jannik Streidl
9fc813cfa6 fix: only append if distances are available 2024-10-07 21:13:13 +02:00
Jannik Streidl
b105efa05f enh: append citations with distance scores 2024-10-07 21:11:04 +02:00
Timothy J. Baek
ee22ba9676 fix: web, youtube attachment issue 2024-10-06 19:44:02 -07:00
Timothy J. Baek
05c15b017d fix: milvus 2024-10-06 17:58:09 -07:00
Timothy J. Baek
378223aedb refac: comments 2024-10-05 10:08:48 -07:00
Timothy J. Baek
61b147441c refac 2024-10-05 10:05:12 -07:00
Timothy J. Baek
1f9b5b6456 refac: retain metadata for collection 2024-10-05 09:58:46 -07:00
Timothy J. Baek
0bd88090bb refac: DOCS_DIR deprecated 2024-10-05 01:45:22 -07:00
Timothy J. Baek
493745a70b refac 2024-10-04 19:32:33 -07:00
Timothy J. Baek
a6c797d4c2 refac: process docs dir 2024-10-04 17:22:00 -07:00
Timothy J. Baek
d1b8af6220 chore: format 2024-10-04 01:04:04 -07:00
Timothy Jaeryang Baek
0876c9b5ef
Merge pull request #5829 from jannikstdl/query-embedding-perf-fix
fix:  performance issues on large collections
2024-10-04 10:01:17 +02:00
Timothy Jaeryang Baek
ebc7da6f82
Merge pull request #5861 from open-webui/projects
feat: knowledge/projects
2024-10-04 10:00:47 +02:00
Timothy J. Baek
1fe1c27220 refac: legacy support 2024-10-04 00:59:19 -07:00
Timothy J. Baek
8013c152d0 feat: edit file content support 2024-10-04 00:46:32 -07:00
Timothy J. Baek
630a78cead fix 2024-10-04 00:24:51 -07:00
Timothy J. Baek
17c772831d refac 2024-10-04 00:23:14 -07:00
Timothy J. Baek
05970157f6 refac 2024-10-03 23:06:47 -07:00
Timothy J. Baek
b291271df3 refac 2024-10-03 22:22:22 -07:00
Timothy J. Baek
b862dff185 refac 2024-10-03 21:10:33 -07:00
Timothy J. Baek
6747478f67 refac 2024-10-03 21:05:55 -07:00
Timothy J. Baek
124a17e826 refac 2024-10-03 20:58:56 -07:00
Timothy J. Baek
57360b7a61 refac 2024-10-03 20:51:21 -07:00
Timothy J. Baek
2fc07fd6a2 enh: vector db hash collision check 2024-10-03 06:53:21 -07:00
Timothy J. Baek
d394f8b7be enh: add to vector db support 2024-10-03 06:44:17 -07:00
Timothy J. Baek
325ca98773 enh: vector db delete filter support 2024-10-03 06:43:50 -07:00
Timothy J. Baek
351b1dbf31 refac 2024-10-02 21:14:58 -07:00
Jun Siang Cheah
318f61161e refac: lazy load big dependencies to reduce min memory usage 2024-10-02 22:18:42 +01:00
Timothy J. Baek
a0fb4a9b84 refac 2024-10-01 13:13:39 -07:00
Jannik Streidl
6bff5a4d09 fix: do not embed the query every single time 2024-09-30 16:18:02 +02:00
Timothy J. Baek
8206c47a47 refac 2024-09-29 23:20:37 +02:00
Timothy J. Baek
6afc686e17 refac 2024-09-29 23:08:55 +02:00
Timothy J. Baek
677c36c3aa refac 2024-09-29 22:55:53 +02:00
Timothy J. Baek
6d764ee55e feat: retrieval whole document mode 2024-09-29 22:52:27 +02:00
Timothy J. Baek
92dd173b27 refac 2024-09-29 18:55:26 +02:00
Timothy J. Baek
276ce3374d refac 2024-09-28 19:17:21 +02:00
Timothy J. Baek
e77c3ab043 refac 2024-09-28 19:16:52 +02:00
Timothy J. Baek
90ec458c4c enh: show extracted file content 2024-09-28 10:53:25 +02:00
Timothy J. Baek
9d2ed3d2be refac 2024-09-28 02:56:56 +02:00
Timothy J. Baek
b8b994a820 refac 2024-09-28 02:49:18 +02:00
Timothy J. Baek
00eb022450 refac 2024-09-28 02:38:59 +02:00
Timothy J. Baek
2428878f42 refac 2024-09-28 02:29:08 +02:00
Timothy J. Baek
af57a2c153 refac 2024-09-28 02:23:09 +02:00
Timothy J. Baek
a52e8cd537 refac 2024-09-28 01:35:31 +02:00
Timothy J. Baek
5b7cf88915 refac 2024-09-28 01:28:45 +02:00
Timothy J. Baek
e1103305f5 refac: "rag" endpoints renamed to "retrieval" 2024-09-28 01:27:46 +02:00