|
|
a5ae9cc7fb
|
Simplify message owner checks a bit
|
2023-05-06 00:25:53 +02:00 |
|
|
|
e639666b34
|
Move chat renaming logic to server side. Add "quiet" reply generation mode.
|
2023-05-03 21:02:23 +03:00 |
|
|
|
fcd8a8886b
|
Add OAI preset deletion button. Nicer OAI models dump. Adjust auto-connect delay
|
2023-05-02 22:42:39 +03:00 |
|
|
|
d159ddf4d1
|
Properly highlight last OpenAI message
|
2023-05-02 18:05:24 +03:00 |
|
|
|
2f55b11afe
|
Atomic treatment of message examples blocks in OpenAI
|
2023-05-01 20:46:28 +02:00 |
|
|
|
3f37de699a
|
Revert "#210 [OAI] Don't insert example dialogues without bots"
This reverts commit 7b9e042589.
|
2023-05-01 16:56:27 +02:00 |
|
|
|
7b9e042589
|
#210 [OAI] Don't insert example dialogues without bots
|
2023-05-01 17:15:16 +03:00 |
|
|
|
22f2c11d6b
|
Merge pull request #208 from bf62963/dev
|
2023-05-01 12:25:40 +03:00 |
|
|
|
29cfa97ddc
|
Fix group chat member names in openai.js
Group members are not by card name, so look up the actual character name instead
|
2023-05-01 04:14:42 -04:00 |
|
|
|
7a1c543ec5
|
Quick Fix to Enable Using gpt-4-0314
|
2023-05-01 03:53:27 +00:00 |
|
|
|
6615d94051
|
Rename character as a group member.
Unify CSRF headers usage in fetch request.
|
2023-04-30 22:36:18 +03:00 |
|
|
|
e57f53c2fb
|
- smart OAI reverse proxy warning
|
2023-05-01 02:00:13 +09:00 |
|
|
|
7babc1cafe
|
Handholding message for broke
|
2023-04-29 23:21:19 +03:00 |
|
|
|
57a5fc8ea8
|
Minor spelling mistake
|
2023-04-29 20:45:04 +03:00 |
|
|
|
d625f1386e
|
Do not act as if forEach can iterate beyond the chat array
|
2023-04-29 13:31:26 +02:00 |
|
|
|
303e183cda
|
Fix disappearing personality and unreliable anchors when swiping or with system messages
|
2023-04-29 13:31:26 +02:00 |
|
|
|
a71b8967ac
|
Deduplicate code for substitution in the story and the first message
|
2023-04-29 13:29:15 +02:00 |
|
|
|
717c14801d
|
Define bottom anchor threshold in one place
|
2023-04-29 13:13:42 +02:00 |
|
|
|
f2d4b5ba76
|
Remove redundant checks for personality insertion and fix a stray space in OpenAI prompt
|
2023-04-29 13:13:41 +02:00 |
|
|
|
443682a7f3
|
Add import/export/deletion of logit bias presets
|
2023-04-24 19:40:02 +03:00 |
|
|
|
141a60f5f0
|
Add logit bias
|
2023-04-23 21:54:21 +03:00 |
|
|
|
99b6571e32
|
Change default JB prompt
|
2023-04-23 19:24:50 +03:00 |
|
|
|
388d982040
|
Add OpenAI API usage
|
2023-04-23 18:48:46 +03:00 |
|
|
|
f25ecbd95c
|
Add proper processing of streaming aborting
|
2023-04-21 20:29:18 +03:00 |
|
|
|
4f9cbe5a5d
|
Add GPT4 32K option
|
2023-04-21 23:11:35 +08:00 |
|
|
|
a563a45bf0
|
Send example dialogues as system messages with names
|
2023-04-15 20:05:58 +03:00 |
|
|
|
e3b9c24d8c
|
Save OAI preset based on current settings
|
2023-04-15 19:47:41 +03:00 |
|
|
|
648f2e9ceb
|
Bug fixes and optimizations
|
2023-04-15 13:45:25 +03:00 |
|
|
|
e350c460b8
|
fixed typo in OAI impersonation string
|
2023-04-14 13:21:38 +09:00 |
|
|
|
31521f41aa
|
Adjust styles
|
2023-04-14 01:19:20 +03:00 |
|
|
|
b72d6566ea
|
Impersonation mode for OAI
|
2023-04-14 00:35:23 +03:00 |
|
|
|
48e4e408a4
|
- ai config ranges more compact
- fixed misuse of range-block-range class on text areas
|
2023-04-12 14:58:00 +09:00 |
|
|
|
2040b8d3ff
|
OAI streaming
|
2023-04-12 01:33:33 +03:00 |
|
|
|
dee813dfa7
|
Foundation for token streaming (non-functional at the moment)
|
2023-04-11 22:49:05 +03:00 |
|
|
|
f2d7f74d5a
|
OAI reverse proxy (untested)
|
2023-04-10 18:44:11 +03:00 |
|
|
|
72efa08a88
|
Update OAI defaults
|
2023-04-06 13:05:14 +03:00 |
|
|
|
b295f5a49d
|
Preset settings for OAI #23
|
2023-04-04 20:50:24 +03:00 |
|
|
|
488edf38c9
|
Fix OAI preset not displaying on save
|
2023-04-03 12:13:30 +03:00 |
|
|
|
d79d7576bd
|
Migrate advanced formatting settings to settings.json
|
2023-04-02 21:59:28 +03:00 |
|
|
|
cf9af364fb
|
Preselect preset based on character name
|
2023-04-02 20:50:45 +03:00 |
|
|
|
006c5b63cc
|
OpenAI presets saving (presumably)
|
2023-04-02 17:13:20 +03:00 |
|
|
|
885ef0c37c
|
Save some OAI tokens
|
2023-04-01 03:11:20 +03:00 |
|
|
|
3acb43a7a4
|
Placeholders are not working in A/N / some extra suggestions about it SillyLossy/TavernAI#18
|
2023-03-31 20:56:41 +03:00 |
|
|
|
ef52e20986
|
made bias triggering stricter
{ } >>> {{ }}
|
2023-03-31 23:31:08 +09:00 |
|
|
|
94abd80bb2
|
Experimental: always pad tokens in OpenAI
|
2023-03-31 10:43:11 +03:00 |
|
|
|
1541683492
|
Add caching of OAI messages tokens
|
2023-03-30 19:11:42 +03:00 |
|
|
|
ca889b420a
|
Add jailbreak as a system message option
|
2023-03-28 21:13:52 +03:00 |
|
|
|
cd4f699e74
|
Allow zero depth for floating prompt
|
2023-03-28 17:49:02 +03:00 |
|
|
|
ae7e21d6c5
|
Async OAI tokenization
|
2023-03-28 17:44:28 +03:00 |
|
|
|
2c4876ac39
|
Adjust max tokens based on OAI model
|
2023-03-28 16:44:26 +03:00 |
|
