250 Commits

Author SHA1 Message Date
3089cdb72d Dont cache the social media domain checks. i can probably put this in other places but just here for now works 2024-11-26 15:39:46 -08:00
7e653681a7 Comment out the llm stuff since its turned off 2024-11-22 10:51:41 -08:00
d9f7a7a426 Add boss portraits 2024-11-03 18:18:22 -08:00
49ed58b25f Move social media domain corrections out of the bot so they can be updated out of band 2024-11-01 19:14:45 -07:00
a553949776 Dont need to fix tiktok anymore, they play natively 2024-10-29 21:54:33 -07:00
4c8a4f992d No-op commit to get CI to run 2024-10-24 14:39:26 -07:00
86e3634dcd Adding a domain corrector for instagram videos 2024-09-28 10:18:51 -07:00
907679eaa5 Auto correct reddit to rxddit for embed 2024-08-06 11:04:52 -07:00
038ec6e691 fix hardcoding of ein's IP 2024-08-05 15:23:05 -07:00
73690d9d35 Change vxtwitter to fixupx 2024-06-17 08:59:47 -07:00
364bbd2483 Comment out the social media post edits 2024-06-17 08:57:26 -07:00
038c7f9eff Add a try/except to the llm 2024-06-06 08:03:07 -07:00
014f34903b small change, removing the command prefix from the bot 2024-05-17 09:57:20 -07:00
58439e18a5 Add autocomplete to the /meme function 2024-05-03 10:50:11 -07:00
345559ecbc update action file 2024-05-02 19:38:58 -07:00
ed72385cb7 Move the roles blurb to the web endpoint as well 2024-05-01 13:06:51 -07:00
176896a5c9 Remove palworld from the blurb and merge tarkov and grayzone 2024-05-01 10:02:01 -07:00
164d567976 open the llm up to everyone 2024-04-24 16:20:56 -07:00
3e4d0a53c8 Make it look like dale-bot is typing 2024-04-23 10:49:47 -07:00
eecd0d60df fix dale bot to work with llama3, also move stop tokens to a list i can configure 2024-04-23 10:42:39 -07:00
ca43550b20 Allow ME to use the LLM in any server 2024-04-22 11:20:07 -07:00
5513a8727b Allow ME to use the LLM in any server 2024-04-22 11:19:18 -07:00
6ca3087821 Allow ME to use the LLM in any server 2024-04-22 10:59:46 -07:00
7b1ae9c32b Not sure why i marked goon as the stop word 2024-04-20 16:46:01 -07:00
d28ce3aaee Move dracula flow to the kubernetes endpoint so i can rapidly change, and set the json endpoint as a var in core_utils 2024-04-19 21:55:22 -07:00
79f2a18675 I'm dumb, dont make 3 rquests for the same json file 2024-04-19 10:31:13 -07:00
1fea8c1c2c move the model to the endpoint as well 2024-04-19 09:58:11 -07:00
533442fa4e move bot instructions to k8s endpoint and also add a switch for the URL so i can test without worrying 2024-04-19 09:39:54 -07:00
746611c87c Add responseds for timed out people for llm 2024-04-18 14:10:34 -07:00
1cda3042c1 update llm url 2024-04-18 11:05:03 -07:00
362ecffeba no more request caching 2024-04-16 13:12:18 -07:00
41fd73d598 dont cast to string 2024-04-16 12:32:37 -07:00
088ee014ed fix url for json 2024-04-16 12:29:18 -07:00
97cfcdfe1d Format 2024-04-16 12:19:24 -07:00
29c6a197f4 Have dale-bot pull from a json endpoint to get gpt users 2024-04-16 12:17:10 -07:00
014e258469 shut down jason harder 2024-04-16 11:06:32 -07:00
f138b98872 typo 2024-04-15 21:16:22 -07:00
ecd8811260 changing the prompt even though it doesnt do anything, also ignoring jason again 2024-04-15 20:14:08 -07:00
188172ce38 Fucking with the llm prompt, i dont know if it even does anything 2024-04-15 17:04:06 -07:00
b6d7eea957 Fixing dale's brain 2024-04-14 17:24:57 -07:00
8336ae9ddf Split long answers into two mesages 2024-04-10 20:17:50 -07:00
2042e7e930 silence jason 2024-04-10 14:28:48 -07:00
eec2dd78dc Lock gpt to my servers 2024-04-10 14:14:04 -07:00
630a9b8029 Change gpt model 2024-04-10 07:35:20 -07:00
24f1ec7a02 adjusting chatgpt prompt 2024-04-09 15:04:49 -07:00
b1ff1c09c1 Insruct the bot to keep things under 2000 characters 2024-04-08 14:18:01 -07:00
f0a9deb00e Turn gpt back on 2024-04-08 13:44:45 -07:00
f0f4758290 Turn off gpt for now, its not working 2024-04-08 10:06:50 -07:00
dc7e9c7936 Adding in the first pass at llm 2024-04-05 23:16:56 -07:00
1062c9707b Fixing a bunch of bad stuff using an actual linter 2024-04-04 22:00:08 -07:00