This makes it so that SocketIO uses long polling to set up the
connection before switching to websocket, instead of immediately using
websocket.
This seems to resolve issues where the browser sometimes can't connect
to the websocket server until the window has been open for a minute.
kobold.outputs must be a 1D list of strings, but sometimes its still blank. In those cases rather than throwing an error and crashing the scripting its better if it does nothing.
Allow people to enter a prompt without generating anything by the AI. Combined with the always add prompt this is a very useful feature that allows people to write world information first, and then do a specific action. This mimics the behavior previously seen in AI Dungeon forks where it prompts for world information and then asks an action and can be particularly useful for people who want the prompt to always be part of the generation.
Automatically converts Huggingface cache models to full models on (down)load.
WARNING: Does wipe old cache/ dir inside the KoboldAI folder, make a backup before you run these models if you are bandwith constraint.
This seems to be related to the model config files, because only certain
models have this problem, and replacing ALL configuration files of a
"bad" model with those of a "good" model of the same type would fix the
problem.
Shouldn't be required anymore.
Updates KoboldAI to the latest official version, if you want to use United you first manually need to be on that git branch otherwise it gets overwritten with the old one.
Apparently Colab does not properly have netbase which we now use for the proper websocket support, now the installer forces it to be correctly installed so we don't crash on launch.
We may want to bundle these at some point, but in that case you should make an exception like we do for the sample story. The same applies to Softprompts.
