157315cd68
Add Vertex AI express mode support ( #3977 )
...
* Add Vertex AI express mode support
Split Google AI Studio and Vertex AI
* Add support for Vertex AI, including updating default models and related settings, modifying frontend HTML to include Vertex AI options, and adjusting request processing logic in the backend API.
* Log API name in the console
* Merge sysprompt toggles back
* Use Gemma tokenizers for Vertex and LearnLM
* AI Studio parity updates
* Add link to express mode doc. Also technically it's not a form
* Split title
* Use array includes
* Add support for Google Vertex AI in image captioning feature
* Specify caption API name, add to compression list
---------
Co-authored-by: Cohee <18619528+Cohee1207@users.noreply.github.com >
2025-05-22 20:10:53 +03:00
db2971c82d
Remove Pygmalion instruct template (duplicate of Metharme)
...
ST already applies user sequence as stop string, so Pygmalion's <|user|> stop_sequence is meaningless.
2025-05-01 03:24:28 -05:00
8dc7aa0c20
Add post_history field to default prompts
2025-04-30 21:07:06 +03:00
775ae0f557
TC sysprompt: Add Post-History Instructions control
...
Closes #3920
2025-04-28 00:14:57 +03:00
76aa17e08f
Merge pull request #3911 from cloak1505/staging
...
Normalize instruct names behavior and repair Lightning 1.1's system prompt
2025-04-26 00:10:01 +03:00
82c86c9ce6
Clean Lightning 1.1
2025-04-25 14:57:42 -05:00
cf7edd99a7
trivial: remove extraneous \n after sop token
2025-04-26 00:08:02 +09:00
2151ae7aaa
Normalize instruct "names_behavior" to "force" for those that don't require "none" or "always
2025-04-25 09:40:49 -05:00
81fec97f54
Repair Lightning 1.1's system prompt
2025-04-25 09:22:10 -05:00
2d366117dd
chat preset: GLM-4
2025-04-25 15:22:51 +09:00
897632b583
fix: system prompt not the same as user prompt for L3/4
2025-04-13 15:12:09 +01:00
fc9b2173c0
fix: add <|begin_of_text|> to story string
2025-04-13 14:14:43 +01:00
61ca7775d2
fix: start/end header tags
2025-04-13 14:09:54 +01:00
f95077ac9f
chore: add llama 4 chat templates
2025-04-13 14:04:52 +01:00
271c93a504
Rename DeepSeek template, add Blank reasoning template
2025-03-18 00:32:31 +02:00
62342b35e2
Reasoning template
2025-03-16 15:01:31 +02:00
070de9df2d
(CC) Move continue nudge at the end of completion ( #3611 )
...
* Move continue nudge at the end of completion
Closes #3607
* Move continue message together with nudge
2025-03-09 18:17:02 +02:00
7d568dd4e0
Generic generate methods ( #3566 )
...
* sendOpenAIRequest/getTextGenGenerationData methods are improved, now it can use custom API, instead of active ones
* Added missing model param
* Removed unnecessary variable
* active_oai_settings -> settings
* settings -> textgenerationwebui_settings
* Better presetToSettings names, simpler settings name in getTextGenGenerationData,
* Removed unused jailbreak_system
* Reverted most core changes, new custom-request.js file
* Forced stream to false, removed duplicate method, exported settingsToUpdate
* Rewrite typedefs to define props one by one
* Added extractData param for simplicity
* Fixed typehints
* Fixed typehints (again)
---------
Co-authored-by: Cohee <18619528+Cohee1207@users.noreply.github.com >
2025-03-03 10:30:20 +02:00
09aaa9181c
Good night 'Black Magic Time'
2025-02-13 23:34:20 +02:00
2b00cdce7b
system same as user
2025-01-22 17:07:23 +09:00
636e79c438
fixes
2025-01-22 16:03:32 +09:00
7877e6601d
add presets for DeepSeek R1 models
2025-01-22 13:13:41 +09:00
7f94cb4bee
CC: Simplify default wrappers for personality and scenario
2024-12-22 23:36:58 +02:00
17df259afd
Update ChatML-Names.json
...
Remove brackets from `ChatML-Names`
2024-12-22 15:00:00 -06:00
964437ed13
Update Llama-3-Instruct-Names.json
...
Remove brackets from `Llama-3-Instruct-Names`
2024-12-22 14:59:35 -06:00
ef35adb9e4
tulu template: added \n to input_suffix and system_suffix to match model card
2024-11-28 13:04:40 +09:00
8d67bdee84
Add Tulu templates
2024-11-27 11:58:12 +09:00
9382845dee
Claude: remove user filler from prompt converter
2024-11-24 19:05:41 +02:00
5d2858268e
Mistral-2411 -> Mistral V7
2024-11-19 21:12:01 +01:00
30269e8ec6
Mistral-2411 context+instruct presets
2024-11-19 20:10:10 +01:00
8c39e20ed9
Merge branch 'staging' into deprecate-cc-or-instruct
2024-09-24 21:19:03 +03:00
bc6520c392
Tabs? Why does it have to be tabs?
2024-09-24 12:56:59 +03:00
e730acbbe1
Port Erato presets
2024-09-24 12:55:25 +03:00
26f4d1a4ad
Remove CFG for Novel, fix logitbias and text adventure bans for Erato
2024-09-24 08:12:50 +00:00
182756fcb2
Erato ( #2896 )
...
* Erato
erato
2024-09-24 03:10:31 +03:00
3812e3bdcb
Merge branch 'staging' into sysprompt-divorce
2024-09-22 19:56:29 +03:00
a18dae8f69
Remove "include newlines" checkbox from context formatting settings
2024-09-22 19:55:43 +03:00
e122dd2d7a
Merge branch 'staging' into sysprompt-divorce
2024-09-21 02:11:42 +03:00
125498cfd5
Fixing Mistral Templates ( #2883 )
...
* Update Mistral.json
* Rename Mistral.json to Mistral V1.json
* Create Mistral V2 & V3.json
* Create Mistral V3-Tekken.json
* Update and rename Mistral.json to Mistral V1.json
* Update Mistral V1.json
* Update Mistral V2 & V3.json
* Update Mistral V2 & V3.json
* Update Mistral V3-Tekken.json
* Update Mistral V1.json
* Create Mistral V2 & V3.json
* Create Mistral V3-Tekken.json
* Update Mistral V1.json
* Update Mistral V2 & V3.json
* Update Mistral V3-Tekken.json
* Update Mistral V1.json
* Update Mistral V2 & V3.json
* Update Mistral V3-Tekken.json
* Update Mistral V1.json
* Update Mistral V2 & V3.json
* Update Mistral V1.json
* Update Mistral V3-Tekken.json
* Update Mistral V1.json
* Update Mistral V2 & V3.json
* Update Mistral V3-Tekken.json
* Add to content index
* Remove example separators
* Simplify context fillers
* Add a hack for proper V1 format with names
* Readability
---------
Co-authored-by: Cohee <18619528+Cohee1207@users.noreply.github.com >
2024-09-21 01:56:02 +03:00
e2f88e6243
Add Unified sampling for Novel
2024-09-21 00:24:52 +03:00
8efe0b6437
Add Min P for Novel
2024-09-20 22:42:56 +03:00
5001187b3c
Revert default prompt
2024-09-20 08:15:23 +00:00
589dc8a858
Revert "Remove square brackets from default prompts"
...
This reverts commit 48cd881972
2024-09-20 08:15:23 +00:00
4e5a997e63
Set default sysprompt to neutral chat
2024-09-20 00:35:47 +03:00
48cd881972
Remove square brackets from default prompts
2024-09-19 23:54:16 +03:00
b377a2b7d3
Decouple system prompts from instruct mode
2024-09-17 10:37:36 +00:00
ca4d3c5ffe
Refactor instruct names behavior controls ( #2837 )
...
* Refactor instruct names behavior controls
Closes #2819
* Remove current persona from force names
* Don't add current user name in groups
* Revert to old text
* Add titles to names behavior
* Proper default for names_behavior
2024-09-16 00:46:03 +03:00
30af741c3e
Deprecated forced instruct on OpenRouter for Chat Completion
2024-09-15 10:54:12 +03:00
7d4b3e0800
Use J1.5 Large by default
2024-08-26 12:11:29 +03:00
5fc16a2474
New AI21 Jamba + tokenizer
2024-08-26 12:07:36 +03:00
