Commit Graph

269 Commits

Author SHA1 Message Date
Jyong
0d9ce1bab0
fix multi retrieval with resource score issue (#1578)
Co-authored-by: jyong <jyong@dify.ai>
2023-11-21 13:47:55 +08:00
Jyong
0e08526428
fix hybrid search reranking check (#1563)
Co-authored-by: jyong <jyong@dify.ai>
2023-11-18 17:06:28 +08:00
Jyong
4588831bff
Feat/add retriever rerank (#1560)
Co-authored-by: jyong <jyong@dify.ai>
2023-11-17 22:13:37 +08:00
takatost
d654770732
feat: supports for new version of openllm (#1554) 2023-11-17 14:07:36 +08:00
Garfield Dai
8835435558
fix: change model mode. (#1520) 2023-11-13 23:13:01 +08:00
takatost
41d0a8b295
feat: [backend] vision support (#1510)
Co-authored-by: Garfield Dai <dai.hai@foxmail.com>
2023-11-13 22:05:46 +08:00
crazywoola
d0e1ea8f06
1506 remove duplicated code (#1511) 2023-11-13 19:05:32 +08:00
takatost
9de67c586f
feat: update free plan rules of spark (#1515) 2023-11-13 17:00:36 +08:00
takatost
d7ae86799c
feat: support basic feature of OpenAI new models (#1476) 2023-11-07 04:05:59 -06:00
takatost
059c089f93
fix: external data tool batch retrieve bug (#1472) 2023-11-07 01:28:22 -06:00
Garfield Dai
db43ed6f41
feat: add api-based extension & external data tool & moderation backend (#1403)
Co-authored-by: takatost <takatost@gmail.com>
2023-11-06 19:36:16 +08:00
takatost
4dfbcd0b4e
feat: support chatglm_turbo model #1443 (#1460) 2023-11-06 04:33:05 -06:00
takatost
076f3289d2
feat: add spark v3.0 llm support (#1434) 2023-10-31 03:13:11 -05:00
takatost
e122d677ad
fix: return wrong when init 0 quota in trial provider (#1394) 2023-10-21 14:02:38 -05:00
takatost
4c63cbf5b1
feat: adjust anthropic (#1387) 2023-10-20 02:27:46 -05:00
Garfield Dai
fe14130b3c
refactor advanced prompt core. (#1350) 2023-10-18 20:02:52 +08:00
wayne.wang
52ebffa857
fix: app config zhipu chatglm_std model, but it still use chatglm_lit… (#1377)
Co-authored-by: wayne.wang <wayne.wang@beibei.com>
2023-10-18 05:07:36 -05:00
takatost
7c9b585a47
feat: support weixin ernie-bot-4 and chat mode (#1375) 2023-10-18 02:35:24 -05:00
takatost
07285e5f8b
feat: optimize completion model agent (#1364) 2023-10-17 06:54:59 -05:00
Garfield Dai
695841a3cf
Feat/advanced prompt enhancement (#1340) 2023-10-13 16:47:01 +08:00
takatost
3efaa713da
feat: use xinference client instead of xinference (#1339) 2023-10-13 02:46:09 -05:00
takatost
9822f687f7
fix: max tokens of OpenAI gpt-3.5-turbo-instruct to 4097 (#1338) 2023-10-13 02:07:07 -05:00
takatost
f4be2b8bcd
fix: raise error in minimax stream generate (#1336) 2023-10-12 23:48:28 -05:00
Garfield Dai
42a5b3ec17
feat: advanced prompt backend (#1301)
Co-authored-by: takatost <takatost@gmail.com>
2023-10-12 10:13:10 -05:00
takatost
2d1cb076c6
fix: dataset segment not exist return agent response (#1329) 2023-10-12 04:40:20 -05:00
Jyong
289c93d081
Feat/improve document delete logic (#1325)
Co-authored-by: jyong <jyong@dify.ai>
2023-10-12 13:30:44 +08:00
takatost
9cba1c8bf4
fix: retriever_resource missing (#1317) 2023-10-11 14:37:11 -05:00
takatost
cbf095465c
feat: remove llm client use (#1316) 2023-10-11 14:02:53 -05:00
Jyong
7f6ad9653e
Fix/grpc gevent compatible (#1314)
Co-authored-by: jyong <jyong@dify.ai>
2023-10-11 20:48:35 +08:00
takatost
2851a9f04e
feat: optimize minimax llm call (#1312) 2023-10-11 07:17:41 -05:00
takatost
c536f85b2e
fix: compatibility issues with the tongyi model. (#1310) 2023-10-11 05:16:26 -05:00
takatost
b1352ff8b7
feat: using random sampling to check if it violates the review mechan… (#1308) 2023-10-11 04:11:20 -05:00
Jyong
5003db987d
milvus secure check fix (#1305)
Co-authored-by: jyong <jyong@dify.ai>
2023-10-11 13:11:06 +08:00
Jyong
07aab5e868
Feat/add milvus vector db (#1302)
Co-authored-by: jyong <jyong@dify.ai>
2023-10-10 21:56:24 +08:00
takatost
875dfbbf0e
fix: openllm completion start with prompt, remove it (#1303) 2023-10-10 04:44:19 -05:00
takatost
8480b0197b
fix: prompt for baichuan text generation models (#1299) 2023-10-10 13:01:18 +08:00
takatost
4ab4bcc074
feat: support openllm embedding (#1293) 2023-10-09 23:09:35 -05:00
takatost
1d4f019de4
feat: add baichuan llm support (#1294)
Co-authored-by: zxhlyh <jasonapring2015@outlook.com>
2023-10-09 23:09:26 -05:00
takatost
024250803a
feat: move login_required wrapper outside (#1281) 2023-10-08 05:21:32 -05:00
takatost
373e90ee6d
fix: detached model in completion thread (#1269) 2023-10-02 22:27:25 +08:00
takatost
41d4c5b424
fix: count down thread in completion db not commit (#1267) 2023-10-02 10:19:26 +08:00
takatost
8606d80c66
fix: request timeout when openai completion (#1265) 2023-10-01 16:00:23 +08:00
takatost
a31466d34e
fix: db session not commit before long llm call running (#1251) 2023-09-27 21:40:26 +08:00
takatost
d38eac959b
fix: wenxin model name invalid when llm call (#1248) 2023-09-27 16:29:13 +08:00
Jyong
46154c6705
Feat/dataset service api (#1245)
Co-authored-by: jyong <jyong@dify.ai>
Co-authored-by: StyleZhang <jasonapring2015@outlook.com>
2023-09-27 16:06:32 +08:00
Garfield Dai
18c710c906
feat: support binding context var (#1227)
Co-authored-by: Joel <iamjoel007@gmail.com>
2023-09-27 14:53:22 +08:00
zxhlyh
227f9fb77d
Feat/api jwt (#1212) 2023-09-25 12:49:16 +08:00
Jyong
724e053732
Fix/qdrant data issue (#1203)
Co-authored-by: jyong <jyong@dify.ai>
2023-09-22 14:21:26 +08:00
Garfield Dai
e409895c02
Feat/huggingface embedding support (#1211)
Co-authored-by: StyleZhang <jasonapring2015@outlook.com>
2023-09-22 13:59:02 +08:00
takatost
32d9b6181c
fix: transaction not commit during long LLM calls (#1213) 2023-09-22 12:43:06 +08:00
takatost
2b018fade2
fix: transaction hangs due to message commit block during long LLM calls (#1206) 2023-09-21 11:22:10 +08:00
takatost
435f804c6f
fix: gpt-3.5-turbo-instruct context size to 8192 (#1196) 2023-09-19 02:10:22 +08:00
takatost
ae3f1ac0a9
feat: support gpt-3.5-turbo-instruct model (#1195) 2023-09-19 02:05:04 +08:00
Jyong
269a465fc4
Feat/improve vector database logic (#1193)
Co-authored-by: jyong <jyong@dify.ai>
2023-09-18 18:15:41 +08:00
takatost
827c97f0d3
feat: add zhipuai (#1188) 2023-09-18 17:32:31 +08:00
takatost
c8bd76cd66
fix: inference embedding validate (#1187) 2023-09-16 03:09:36 +08:00
takatost
f9082104ed
feat: add hosted moderation (#1158) 2023-09-12 10:26:12 +08:00
Jyong
642842d61b
Feat:dataset retiever resource (#1123)
Co-authored-by: jyong <jyong@dify.ai>
Co-authored-by: StyleZhang <jasonapring2015@outlook.com>
2023-09-10 15:17:43 +08:00
Joel
2d5ad0d208
feat: support optional query content (#1097)
Co-authored-by: Garfield Dai <dai.hai@foxmail.com>
2023-09-10 00:12:34 +08:00
takatost
c4d8bdc3db
fix: hf hosted inference check (#1128) 2023-09-09 00:29:48 +08:00
takatost
a7cdb745c1
feat: support spark v2 validate (#1086) 2023-09-01 20:53:32 +08:00
takatost
73c86ee6a0
fix: prompt of title generation (#1084) 2023-09-01 14:55:58 +08:00
takatost
33562a9d8d
feat: optimize prompt (#1080) 2023-09-01 11:46:06 +08:00
takatost
a199fa6388
feat: optimize high load sql query of document segment (#1078) 2023-09-01 10:52:39 +08:00
takatost
4c8608dc61
feat: optimize conversation title generation output must be a valid JSON (#1077) 2023-09-01 10:31:42 +08:00
takatost
df6604a734
feat: optimize generation of conversation title (#1075) 2023-09-01 02:28:37 +08:00
takatost
78e26f8b75
fix: summary no docs (#1073) 2023-08-31 20:19:26 +08:00
takatost
2191312bb9
fix: segments query missing idx hit (#1072) 2023-08-31 19:39:44 +08:00
takatost
fcc6b41ab7
feat: decrease claude model request time by set max top_k to 10 (#1071) 2023-08-31 18:23:44 +08:00
takatost
d75e8aeafa
feat: disable anthropic retry (#1067) 2023-08-31 16:44:46 +08:00
takatost
2eba98a465
feat: optimize anthropic connection pool (#1066) 2023-08-31 16:18:59 +08:00
takatost
a7a7aab7a0
fix: csv import error (#1063) 2023-08-31 15:42:28 +08:00
yezhwi
d33a269548
refactor(file extractor): file extractor (#1059) 2023-08-31 14:45:31 +08:00
yezhwi
8b8e510bfe
fix: handle AttributeError for datasets and index (#1052) 2023-08-30 11:14:16 +08:00
takatost
417c19577a
feat: add LocalAI local embedding model support (#1021)
Co-authored-by: StyleZhang <jasonapring2015@outlook.com>
2023-08-29 22:22:02 +08:00
Jyong
a43e80dd9c
add qdrant migration (#1046)
Co-authored-by: jyong <jyong@dify.ai>
2023-08-29 10:37:04 +08:00
WangBooth
ad5f27bc5f
fix openpyxl dimensions error (#1041) 2023-08-29 10:36:48 +08:00
takatost
7b3314c5db
fix: dataset desc (#1045) 2023-08-29 09:07:27 +08:00
Jyong
a55ba6e614
Fix/ignore economy dataset (#1043)
Co-authored-by: jyong <jyong@dify.ai>
2023-08-29 03:37:45 +08:00
takatost
02452421d5
fix: pub generate message text return null (#1037) 2023-08-28 16:43:54 +08:00
takatost
0796791de5
feat: hf inference endpoint stream support (#1028) 2023-08-26 19:48:34 +08:00
takatost
6c148b223d
fix: dataset query truncated (#1026) 2023-08-26 17:35:17 +08:00
Uranus
2d9616c29c
fix: xinference last token being ignored (#1013) 2023-08-25 18:15:05 +08:00
Jyong
2d604d9330
Fix/filter empty segment (#1004)
Co-authored-by: jyong <jyong@dify.ai>
2023-08-25 15:50:29 +08:00
Jyong
b1fd1b3ab3
Feat/vector db manage (#997)
Co-authored-by: jyong <jyong@dify.ai>
2023-08-24 21:27:31 +08:00
takatost
9ae91a2ec3
feat: optimize xinference request max token key and stop reason (#998) 2023-08-24 18:11:15 +08:00
takatost
701e2b334f
feat: remove unnecessary prompt of baichuan (#993) 2023-08-24 15:30:59 +08:00
takatost
bd3a9b2f8d
fix: xinference-chat-stream-response (#991) 2023-08-24 14:39:34 +08:00
takatost
18d3877151
feat: optimize xinference stream (#989) 2023-08-24 13:58:34 +08:00
takatost
53e83d8697
feat: optimize baichuan prompt (#988) 2023-08-24 12:07:10 +08:00
takatost
2c30d19cbe
feat: add baichuan prompt (#985) 2023-08-24 10:22:36 +08:00
takatost
9b247fccd4
feat: adjust hf max tokens (#979) 2023-08-23 22:24:50 +08:00
takatost
a76fde3d23
feat: optimize hf inference endpoint (#975) 2023-08-23 19:47:50 +08:00
Jyong
1fc57d7358
normalize embedding (#974)
Co-authored-by: jyong <jyong@dify.ai>
2023-08-23 19:10:11 +08:00
takatost
5cba2e7087
fix: web reader tool retrieve content empty (#957) 2023-08-22 18:01:16 +08:00
Jyong
5623839c71
update document segment (#950)
Co-authored-by: jyong <jyong@dify.ai>
2023-08-22 17:59:24 +08:00
takatost
78d3aa5fcd
fix: embedding init err (#956) 2023-08-22 17:43:59 +08:00
takatost
4f3053a8cc
fix: xinference chat completion error (#952) 2023-08-22 15:58:04 +08:00
takatost
866ee5da91
fix: openllm generate cutoff (#945) 2023-08-22 13:43:36 +08:00
takatost
e0a48c4972
fix: xinference chat support (#939) 2023-08-21 20:44:29 +08:00