Commit Graph

3223 Commits

Author SHA1 Message Date
a85a6cf606 Allow displaying unreferenced macro in message texts 2023-12-21 20:49:03 +02:00
39e0b0f5cb Remove custom Handlebars helpers for extensions. 2023-12-21 20:33:50 +02:00
343c33e331 Stricter Horde prompt sanitation 2023-12-21 20:22:21 +02:00
8fb26284e2 Clean up Generate(), part 2 (#1578)
* Move StreamingProcessor constructor to the top

Typical code style is to declare the constructor at the top of the class
definition.

* Remove removePrefix

cleanupMessage does this already.

* Make message_already_generated local

We can pass it into StreamingProcessor so it doesn't have to be a global
variable.

* Consolidate setting isStopped and abort signal

Various places were doing some combination of setting isStopped, calling
abort on the streaming processor's abort controller, and calling
onStopStreaming. Let's consolidate all that functionality into
onStopStreaming/onErrorStreaming.

* More cleanly separate streaming/nonstreaming paths

* Replace promise with async function w/ handlers

By using onSuccess and onError as promise handlers, we can use normal
control flow and don't need to remember to use try/catch blocks or call
onSuccess every time.

* Remove runGenerate

Placing the rest of the code in a separate function doesn't really do
anything for its structure.

* Move StreamingProcessor() into streaming code path

* Fix return from circuit breaker

* Fix non-streaming chat completion request

* Fix Horde generation and quiet unblocking

---------

Co-authored-by: Cohee <18619528+Cohee1207@users.noreply.github.com>
2023-12-21 20:20:28 +02:00
bddccd0356 Missed several context bind cases 2023-12-21 17:19:42 +02:00
fac4169dd8 Merge pull request #1568 from DonMoralez/staging
(claude)reworked prefix, sysprompt, console messages, sequence check
2023-12-21 17:02:05 +02:00
ffb711d802 Unify Claude request logging with other API sources 2023-12-21 16:59:43 +02:00
b5e59c819c Merge branch 'staging' into claude-rework 2023-12-21 16:52:43 +02:00
e1afe41c91 Fix custom expression duplication 2023-12-21 16:50:30 +02:00
b3dfe16706 #1575 Fix clean-up WI depth injections 2023-12-21 16:33:21 +02:00
e087f29496 Log MistralAI prompts to server console 2023-12-21 16:08:58 +02:00
dd661cf879 Instruct "Bind to context" is now an option 2023-12-21 15:12:30 +02:00
ee75adbd2d Update persona name if it is bound by user name input 2023-12-21 14:56:32 +02:00
f3099ac270 Remove model icon fill colors 2023-12-21 14:43:36 +02:00
747867c6f4 Merge pull request #1580 from SillyTavern/custom
Custom API endpoint type for Chat Completion
2023-12-21 14:42:18 +02:00
afdd9d823e Merge branch 'staging' of https://github.com/Cohee1207/SillyTavern into staging 2023-12-21 12:43:59 +09:00
348cc5f2a5 placeholder API icon for tabby 2023-12-21 12:43:57 +09:00
3001db3a47 Add additional parameters for custom endpoints 2023-12-20 23:39:10 +02:00
e42daa4098 Add caption ask prompt mode 2023-12-20 21:23:59 +02:00
ae64c99835 Add custom caption source 2023-12-20 21:05:20 +02:00
cf8d7e7d35 Merge branch 'staging' into custom 2023-12-20 18:37:47 +02:00
ebec26154c Welcome message fixed 2023-12-20 18:37:34 +02:00
5734dbd17c Add custom endpoint type 2023-12-20 18:29:03 +02:00
041b9d4b01 Add style sanitizer to message renderer 2023-12-20 17:03:37 +02:00
34decf1c05 add creating of new QR sets 2023-12-20 14:04:28 +00:00
c212a71425 Fix ignore list of preset manager 2023-12-20 15:51:00 +02:00
b0a4341571 Merge pull request #1574 from artisticMink/feature/before-combine-event
Allow extensions to alter the context order.
2023-12-20 15:46:34 +02:00
69d6b9379a implement QR basics 2023-12-20 13:40:44 +00:00
e19bf1afdd clean out QR extension 2023-12-20 13:39:09 +00:00
93db2bf953 Simplify extras summary settings 2023-12-20 01:56:35 +02:00
4b131067e4 Add local multimodal caption sources 2023-12-20 00:45:45 +02:00
029cf598ce Fix /peek command 2023-12-19 23:12:14 +02:00
8d63ce5559 Log Novel Ai prompt to console
When prompt logging is enabled.
2023-12-19 19:27:24 +01:00
f30f75b310 Add GENERATE_BEFORE_COMBINE_PROMPTS event
Allows for context to be ordered by extensions
2023-12-19 19:11:36 +01:00
da1e9cb3b2 Use const where possible 2023-12-19 19:48:42 +02:00
a78875ca08 Use native color util 2023-12-19 19:47:23 +02:00
3b22159f53 Fix spelling 2023-12-19 19:45:28 +02:00
423c2b70dc Camel case variable name 2023-12-19 19:44:52 +02:00
6859e4443e Fix ollama chunk wrapper 2023-12-19 19:17:19 +02:00
c7b93b690f Merge pull request #1573 from StefanDanielSchwarz/Llama-2-Chat-separator-fix
Llama 2 Chat separator fix
2023-12-19 19:14:54 +02:00
44318fef22 Fix double logging of non-streamed replies 2023-12-19 16:49:21 +02:00
67dd52c21b #1309 Ollama text completion backend 2023-12-19 16:38:11 +02:00
SDS
d9fc396b39 Update Llama 2 Chat.json
Official separator is a space instead of a newline. That format itself is still FUBAR, but at least we get a little closer to what it was intended to look, I guess.
2023-12-19 00:48:42 +01:00
0a649a395d Add 'synchronous' event emitter 2023-12-18 21:55:33 +01:00
edd737e8bd #371 Add llama.cpp inference server support 2023-12-18 22:38:28 +02:00
37807acc60 Merge remote-tracking branch 'upstream/staging' into staging 2023-12-18 22:01:38 +02:00
6e8104873e #1569 Add logit bias for text completions 2023-12-18 18:57:10 +02:00
08ea2095f8 Refactor Novel logit bias 2023-12-18 17:32:10 +02:00
cc27bcb076 UNRELATED: this never happened. 2023-12-18 15:01:45 +02:00
be5d428706 Merge pull request #1565 from SillyTavern/togetherai
Add TogetherAI as a text completion source
2023-12-18 14:52:36 +02:00