66 Commits

Author	SHA1	Message	Date
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	41d0a8b295	feat: [backend] vision support (#1510) ... Co-authored-by: Garfield Dai <dai.hai@foxmail.com>	2023-11-13 22:05:46 +08:00
takatost	d7ae86799c	feat: support basic feature of OpenAI new models (#1476)	2023-11-07 04:05:59 -06:00
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	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
Garfield Dai	42a5b3ec17	feat: advanced prompt backend (#1301) ... Co-authored-by: takatost <takatost@gmail.com>	2023-10-12 10:13:10 -05:00
takatost	cbf095465c	feat: remove llm client use (#1316)	2023-10-11 14:02:53 -05:00
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	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	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
Garfield Dai	e409895c02	Feat/huggingface embedding support (#1211) ... Co-authored-by: StyleZhang <jasonapring2015@outlook.com>	2023-09-22 13:59:02 +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
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	2eba98a465	feat: optimize anthropic connection pool (#1066)	2023-08-31 16:18:59 +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
takatost	0796791de5	feat: hf inference endpoint stream support (#1028)	2023-08-26 19:48:34 +08:00
takatost	9ae91a2ec3	feat: optimize xinference request max token key and stop reason (#998)	2023-08-24 18:11:15 +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
takatost	78d3aa5fcd	fix: embedding init err (#956)	2023-08-22 17:43:59 +08:00
takatost	e0a48c4972	fix: xinference chat support (#939)	2023-08-21 20:44:29 +08:00
takatost	6c832ee328	fix: remove openllm pypi package because of this package too large (#931)	2023-08-21 02:12:28 +08:00
takatost	25264e7852	feat: add xinference embedding model support (#930)	2023-08-20 19:35:07 +08:00
takatost	18dd0d569d	fix: xinference max_tokens alisa error (#929)	2023-08-20 19:12:52 +08:00
takatost	3ea8d7a019	feat: add openllm support (#928)	2023-08-20 19:04:33 +08:00
takatost	da3f10a55e	feat: server xinference support (#927)	2023-08-20 17:46:41 +08:00
takatost	95b179fb39	fix: replicate text generation model validate (#923)	2023-08-19 21:40:42 +08:00