wassname/stampy-chat
1
0
Fork 0
mirror of https://github.com/wassname/stampy-chat.git synced 2026-09-25 14:10:17 +08:00
Code Issues Packages Projects Releases Wiki Activity
357 Commits 1 Branch 0 Tags
7558dbbfa9274fa0a79443aa50eb0703cd5eb320
Commit Graph
44 Commits
Author SHA1 Message Date
Fraser 7558dbbfa9 implement changes from @Thomas-Lemoine review 2023-07-11 23:32:16 -04:00
FraserLee df40ca4e1d Implement changes from @henri123lemoine review 2023-07-11 23:25:42 -04:00
Fraser 8303bf196e Add plex suggestion 2023-07-10 22:06:30 -04:00
Fraser f7ce715ac4 create level modes 2023-07-10 21:59:51 -04:00
FraserLee e092b70a6c Merge pull request #50 from StampyAI/prompt-eng 
Prompt Eng
 2023-07-09 14:45:03 -04:00
Fraser f839ff6863 drop unused f-strings 2023-07-09 14:44:31 -04:00
Fraser 348cb5940e Revert "Improve responses." 
This reverts commit af274dadb1.
 2023-07-09 14:40:53 -04:00
Fraser b292e76fae fix logging times 2023-07-08 15:36:14 -04:00
Fraser af274dadb1 Improve responses. 2023-07-08 15:19:02 -04:00
Fraser 361f1e1499 basic submodule working 2023-07-06 09:56:07 -04:00
Fraser 2b92a501fd include letter key in citation form 2023-06-12 21:36:53 -04:00
Fraser 97ebac8b07 add simplified chat API 2023-06-12 21:33:11 -04:00
Fraser 9ed99441e6 Split out json conversion into wrapping coroutine 2023-06-12 21:15:59 -04:00
Fraser 0496864593 set temperature = 0 2023-06-12 20:27:23 -04:00
Fraser eee29179db pass followups to front-end on finish response 2023-06-02 00:03:03 -04:00
Fraser e2fd7def4b search for followups based on both initial query and generated response 2023-06-01 23:38:39 -04:00
Fraser 820d70a3ab very basic followups check in backend 2023-06-01 23:31:11 -04:00
Fraser 6fbc1171a8 decrease amount of info logged 
also add disclaimer on front-end.
 2023-05-07 21:01:02 -04:00
Fraser d750a08e0e add logging via discord webhook 2023-05-07 20:47:30 -04:00
Fraser dee78258d6 fully working streaming 2023-04-16 21:12:33 -04:00
Fraser 6d91e42363 stream answer 2023-04-16 19:37:57 -04:00
Fraser 10c3068aa5 error passing and logging in SSE 2023-04-16 19:01:12 -04:00
Fraser 421b755a75 somewhat more complete integration 2023-04-16 18:52:58 -04:00
Fraser 0a7f68f1fc working SSE streaming 2023-04-16 18:26:25 -04:00
Fraser 20c9369209 log response time 2023-04-06 20:58:31 -04:00
Fraser 0fe3651f6a error catch around larger section in chat. 2023-04-06 20:25:36 -04:00
Fraser 5e49d90654 fix typo in pormpt 2023-04-06 20:02:00 -04:00
henri123lemoine a561404bef Changed back to using ChatGPT by default 2023-04-05 13:17:07 -04:00
henri123lemoine bfe01889ab Increased k in gpt-4 2023-04-05 11:22:03 -04:00
henri123lemoine 2fa1e51012 Moved to Pinecode instead of pickle 2023-04-05 09:55:03 -04:00
Fraser 2eabf1ac04 disable printing 2023-04-03 20:21:52 -04:00
Fraser 386614fa68 Finish restructure prompt 
I've A/B tested *a lot*, and find doing it this way (sources initially, then truncated conversation, then final question) ends up allowing the user to ask clarifying questions better, with minimal if any drop in citation knowledge and relevance
 2023-04-03 20:20:36 -04:00
Fraser ebbf5e81ac attempt to restructure prompt code 2023-04-03 20:16:51 -04:00
Fraser dc83b8b357 fix bug in prompt 2023-04-03 19:32:46 -04:00
Fraser ebeba0cfbb fix margin 2023-04-01 03:22:00 -04:00
Fraser d49e18d4d0 add gpt 4 toggle 2023-04-01 01:20:59 -04:00
Fraser 74fa2b7b1a fix bug 2023-03-31 23:30:36 -04:00
Fraser 26f7d4411e get chat slightly more compiling 2023-03-31 23:13:44 -04:00
henri123lemoine b197ee086c Big fix total_tokens 2023-03-31 22:10:53 -04:00
henri123lemoine 5b64ae1efc Reduced odds of context length error. 2023-03-31 21:45:15 -04:00
Fraser 52585d7991 add error passing mechanism 2023-03-31 18:32:52 -04:00
Fraser b80194c380 switch to remote datasource 2023-03-31 17:02:29 -04:00
Fraser b9f88f7c9d list citations in obvious way 2023-03-31 13:28:28 -04:00
Fraser e0c7eb8cc7 chat code working in railways 2023-03-30 09:35:23 -04:00