rixty
/
KoboldAI-Client
mirror of https://github.com/KoboldAI/KoboldAI-Client.git
1
1
Fork 0
Code Issues Projects Releases Wiki Activity

Further Polishing 

Multiple smaller changes to get 1.17 in shape for its release.
This commit is contained in:
henk717 2022-02-01 11:15:44 +01:00
parent 36b6dcb641
commit ecd7b328ec
3 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
aiserver.py
View File

@ -2309,6 +2309,8 @@ def loadsettings():
vars.autosave = js["autosave"] vars.autosave = js["autosave"]
if("newlinemode" in js): if("newlinemode" in js):
vars.newlinemode = js["newlinemode"] vars.newlinemode = js["newlinemode"]
if("welcome" in js):
vars.welcome = js["welcome"]
if("antemplate" in js): if("antemplate" in js):
vars.setauthornotetemplate = js["antemplate"] vars.setauthornotetemplate = js["antemplate"]
@ -4632,8 +4634,6 @@ def randomGameRequest(topic, memory=""):
emit('from_server', {'cmd': 'setmemory', 'data': vars.memory}, broadcast=True) emit('from_server', {'cmd': 'setmemory', 'data': vars.memory}, broadcast=True)
# Load desired settings from both the model and the users config file # Load desired settings from both the model and the users config file
if(not vars.model in ["InferKit", "Colab", "OAI", "ReadOnly", "TPUMeshTransformerGPTJ"]):
loadmodelsettings()
loadsettings() loadsettings()
# Prevent tokenizer from taking extra time the first time it's used # Prevent tokenizer from taking extra time the first time it's used

2
static/custom.css
View File

@ -447,7 +447,7 @@ body.connected #popupfooter, #popupfooter.always-available {
#rspopup { #rspopup {
width: 800px; width: 800px;
background-color: #262626; background-color: #262626;
margin-top: 150px; margin-top: 50px;
} }
/*================= Classes =================*/ /*================= Classes =================*/

4
templates/index.html
View File

@ -353,9 +353,9 @@
<div class="aidgpopuplistheader"> <div class="aidgpopuplistheader">
<br> <br>
Story quality and topic depends on the model and your settings/suggestion (Around 0.5 temp is recommended).<br> Story quality and topic depends on the model and your settings/suggestion (Around 0.5 temp is recommended).<br>
This feature works best with finetuned models like GPT-Neo-AID or GPT-Neo-Horni but is limited to what the AI knows.<br> This feature works best with heavily themed models like the Adventure model and can also be influenced with softprompts .<br>
If you get random spam then your model is not capable of using this feature and if you get unrelated stories it does not understand the topic.<br> If you get random spam then your model is not capable of using this feature and if you get unrelated stories it does not understand the topic.<br>
Generated results are unfiltered and can be offensive or unsuitable for children.<br><br> Generated results are unfiltered and can be offensive or unsuitable for children, the AI can make connections the model/softprompt creator did not intend.<br><br>
Unsaved data will be lost.<br><br> Unsaved data will be lost.<br><br>
Below you can input a genre suggestion for the AI to loosely base the story on (For example Horror or Cowboy).<br> Below you can input a genre suggestion for the AI to loosely base the story on (For example Horror or Cowboy).<br>
</div> </div>