Home
Artists
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register
Partychan Telegram Live Sex ThePornDude
Home Artists Posts Import Register

PLUGIN UPDATE: Alive v44 (files) (Patreon)

Published:
2023-06-22 21:25:01
Imported:
2023-06

Content

NEW IN ALIVE V44

This is a very big update code-wise, mainly due to the character system framework that I needed to build to be able to create some of the features many of you requested. I still have like 30 notes of small things I wanted to do or fix in this update alone, so that part it's still very much a work in progress, but I wanted to upload it because I know many of you want some of the fixes or to try some of the new features:

VR UI TOGGLE

You can toggle the Aeternum UI in VR with X/A now. Given the lack of button options and since that's used by VaM already, maybe by other plugins too, it will likely get triggered unintentionally too which is not great. I'll test this more and look into making it better. For me for now this is better than before, it allows changing the UI hand too. But if you have any problems with it or there are any other plugins in conflict, let me know.

VR GOLF

It's pretty cool and I don't even like golf! It's not built in VaM as you might expect, it's actually a unity game that I made to work as standalone and then ported it into vam, to make the physics work. To play, you can activate it on the wall in the dome room in Alive home. To disable it, select the empty box on the same wall, or you can use the VR Golf UI app. You can chage the course from the app as well as the settings. The club will be sticked to the selected hand while playing.
AI CAM

This is meant mostly for desktop. It works in VR too but the performance might not be too good. In VR if the main goal is to take photos, it's recommended to lower VaM's settings, as the quality becomes very irrelevant anyway for what the AI generates. I put together a very quick & improvised preview system, to better frame the shots. It might be a bit buggy and difficult to use. The performance for it also is not too good and I hope I'll get to replace it with something that works and looks like vam native screenshots or MacGruber's SuperShot, but for now best I could to given the circumstances.

To get it to work edit Automatic1111's webui_user.bat and add  "--api" to COMMANDLINE_ARGS. I have this for reference:

COMMANDLINE_ARGS=--xformers --api

--xformers I think helps with performance. Then you just copy the url in the Services app. It should be by default already set to the default A1111 url.  IMPORTANT: if you run text AI oobabooga too, since they're both built on the same ui thing (gradio) they'll both try to use the same port (127.0.0.1:7860). But if one's already open, the other one will jump to the next available port, 7861. So basically the urls change depending on the order on which you open oobabooga and Automatic1111. I'll try to get rid of having to deal with that, with api urls entirely most likely, and just focus on local use and do pings to identify if they're running and on which port, but for now dealing with that has to be manually.

CHAT & ANDROIDS

In order to do the kind of features people want for chat, and to try to make them work and be cool, I had to build a character system to tie things together. It's a very complicated feature but I experimented with it before and it's advancing quickly. There's still a lot of work to be done here, but new exciting things are possible with it already:
- having individual chats with different characters, persistent chat history
- characters information system (age & name for now)
- description & chat sample (similer to AI cards)
- traits (3 basic personality traits for now)
- states that improve or degrade, like sex-drive, health & energy
- some very crude gameplay mechanics

There's still a lot to do here like custom voices per character, editing personalities, loading and editing the bodies (looks), improving the character bottom menu and much more.

BETTER SETTINGS

I redid the settings app and build new ones to organize things better. There's a new Services app where the settings for AI stuff are separated based on purpose. There's also a new Alive app where the settings for Alive now have their own tabs and are easier to use. Also the app screen was updated to be less messy. And to make it easier to switch between functionalities ( Diffusion Cam + poses for example) I added a Recent Apps button, clicking it will list the open apps. Double clicking it will will quickly switch to the previous app. It's like a cycle between the last 2 recent apps.

PERFORMANCE UPDATES

Lights affect performance very much in vam and the headlights all had 2 or 3 lights. I added a new option to enable one light one for the light presets. This should notably improve performance when activated.

The Poses app was slow due to having 1k+ images. I created sprites images with all the poses, took it down to 4 image while looking & feeling the same, so the UI should be much more quick & responsive.

I made also other changes for the UI to be more responsive, like removing the click ripple visual effect and making changing person atoms much more quicker.

BUGS

I'll likely post v45 later this week with any bugs I find and can fix, hopefully a few more features for the character system also to make chat a bit better.


CHANGELOG

  • NEW: added a new head lights setting to help with performance: one headlight only. When enabled, it will enable only one of the headlights, not 2 or 3 as it is by default, reducing the performance hit.
  • NEW: added help info to the headlights settings
  • NEW: added logic to connect to Automatic1111 for stable diffusion AI images; created new app Diffusion camera that can capture a vam screenshot and use AI to enhance it; added sampler & seed settings; added preview cam for better positioning; added image ratio options; started logic for Vr but more work needed there
  • NEW: added logic to save data/states based on Unity scene interaction
  • NEW: added new Construct scene, alternative to VaM's empty scene: simple white room with 1m x 1m grid on the floor, room tone noise (low volume static to make ambience more normal and less like floating in void), no lights but custom rendering light settings for better default visibility even without lights
  • NEW: added UI apps quick switch functionality to top menu to quickly switch between last 2 apps - double clicking on recent apps will swith to the previous app
  • NEW: added UI logic to add location links for easier text notifications that include a link to go to a specific place, added this as a link to the dome room in the vr golf app
  • NEW: Alive Home added support for modular multi-level game system
  • NEW: Alive Home added new VR mini golf game to Alive Home dome room , added 10 mini golf courses
  • NEW: Alive Home added persistent states, things load like they were left: kitchen AC, hallway fan, ambient lights, balcony tarp, elevator doors
  • NEW: Chat added logic to send character specific chat example and character description to AI
  • NEW: Chat added persistent messages & delete messages functionalities
  • NEW: connected UI apps to Alive home mini games allowing UI action & settings; added VR Golf app
  • NEW: created new Alive Settings app and moved alive settings there in grouped tabs
  • NEW: created new Services Settings app, reorganized services settings and added speech-to-text setting
  • NEW: extended Alive backend functionality to work with character based configs, allow storing more meta data that is character based
  • NEW: new apps & functionality for androids, buy & sell, customization, business money & debt system
  • NEW: SPQR ExtraLibraries dll added events for steam VR buttons down/up to be forwarded to vam and be accessible to normal plugins; added steam vr hotkey for mainmenu
  • NEW: Text AI moved as android tab to allow meta data specific to that character to be saved and sent to the AI
  • NEW: VR added toggle UI on/off on pressing A or X (and not focused on VaM UI), move UI to left/right hand based on what button was pressed
  • CHANGE: added 90 framerate to ui framerate limit dropdown
  • CHANGE: added notification (quick 1s thought bubble) on person selection to easier identify the selected person
  • CHANGE: added UI to VaM app change event with old & new app for smarter apps
  • CHANGE: change text to speech functionality to work with general audio as fallback (no person), changed settings test to play as general text-audio instead of forcing a person to be in the scene and selected
  • CHANGE: Chat added logic to clean up repeating characters and punctuation sign floods by chicken brain ai models
  • CHANGE: Chat added logic to prevent chicken brain AI models getting stuck in starting all their messages the same way once they get in a loop
  • CHANGE: cleaned up some not used UI files
  • CHANGE: created poses thumbs sprite images to improve UI performance & get rid of 1k+ files & UI calls
  • CHANGE: fixed and readded old Recent Apps screen, added to top menu to work with the quick switch app logic
  • CHANGE: posing removed All category, too many poses making it not very useful
  • CHANGE: removed old voice snippets & dialogue messages, soundpad app, voice & dialogue to be handled in a single place through the chat screen
  • CHANGE: removed ripple visual effect animation on clicks to reduce performance cost
  • CHANGE: removed setting for alive voice feedback, voice stuff to be controlled differently
  • CHANGE: removed settings app from permanent top menu, added to apps and dropdown top menu
  • CHANGE: renamed & made first tabs on apps bigger to serve as the titles of the apps to make it easier to understand in which screen you're currently in
  • CHANGE: reworked actions system (vibe,twerk,etc) for better performance and to prevent constant calls between the UI and vam with redundant data
  • CHANGE: reworked person select flow to improve performance
  • CHANGE: UI added fullscreen top button for desktop sidebar mode (easier & quicker maximize of the UI), removed right menu popup fullscreen button
  • CHANGE: UI darker overlay background for popups and autohide popups on app change
  • CHANGE: UI forms updates for style & range preview from settings to be usable globally in other apps too
  • CHANGE: UI moved debug app to hidden top menu button
  • CHANGE: UI removed app logic of unselecting app/button & person when clicking on one that's already selected
  • CHANGE: UI removed Enable Dialogue setting, no longer used
  • CHANGE: UI simplified unselect person process, removed background click unselecting person, added unselect button next to people buttons
  • FIX: fixed character switch button appearing when no person is selected
  • FIX: fixed scrolling in VR, when scrolling with right VR hand x and y axis were reversed
  • FIX: fixed setting for background color not updating in vr in realtime
  • FIX: fixed VR version turning off default vam navigation even when alive vr locomotion mode is not enabled
  • FIX: UI fixed help app layout
  • FIX: UI fixed person bottom buttons tooltips




INSTRUCTIONS

- Alive.cslist  - add to person
- Aeternum.cslist - add to session for the UI

The Alive Home scene needs Aeternum to be loaded to bring it to life (buttons, sky, cars, etc).

Alive doesn't need internet access, but it uses the web browser to render the UI in it so a few settings might be needed to get it to work. In User preferences Security these need to be enabled or there might be errors: Allow Plugin Network Access, Enable Web Browsers, Allow All Sites.

If you want to keep the sites whitelist enabled, you can add this to your whitelist.json file in vam's main folder, to whitelist just the UI file needed for Alive: "Saves/PluginData/Aeternum/Ae/Html/mainmenu.html"

AI STUFF

These are optional for people that want to add more life to VaM. By default Alive doesn't require any of these but they can be activated from the Services app.

  • Local text AI (Chat) - requires OobaBooga (quick guide)
  • Local image AI - requires Automatic1111 (quick guide)
  • Local Text to speech & speech recognition - require my Text Audio Tool (quick guide)
  • There's also support for online ElevenLabs AI voices, that requires an account on elevenlabs

CREDITS

Assets used: https://drive.google.com/file/d/1M_F6iWz1zK1Av2hgLNsR-6WOrBOqJs5s/view (3d objects from sketchfab, music from soundcloud)

Ashauryn - morphs ( from AshAuryn.Expressions.2, AshAuryn.Pose_Tools.2 and AshAuryn.Sexpressions.2, AshAuryn.AshAuryn's_Female_Genitalia_Collection.2, AshAuryn.AshAuryn's_Assorted_Expression_Pack_with_75_Morphs.7, AshAuryn.AshAuryn's_Tears_and_Pain_Pack.3)
klphgz - poses ( from Crouching, Standing, Lying & Prone, Sitting packs)
Jackaroo - morphs ( from Jackaroo.JarModularExpressions.2)
kemenate - morphs (kemenate.Morphs.10)
geesp0t - spankings vr hand collision detection (Spankings Plugin Mod v3.6 on reddit)

MacGruber, VR Adult Fun, Acid Bubbles & other plugin makers whose work inspired me to make some of the features here and plugins in general.

ElevenLabs.io AI voices used often in my videos.
OobaBooga's text-generation webui & AUTOMATIC1111's stable-diffusion-webui


FILES

SPQR.SPQRAlive.44.rar

Files

Comments

thomas d

related to the question above for appearances, I converted a lot of my appearances to the old vam looks. But I do not see a lot of them. Is there a limit to how many it reads? Also, maybe in the future a search ability.

Hallow2

im using desktop vr, how to i open the alive ui? cant seem to find it anywhere, ive added alive to the atom and aeternum to the session, and enabled the security settings mentioned

SPQRAeternum
>>111318498

There's a limit introduced in of the latest version, v42 or v43 i think. It should be 300. People with many looks had problems with loading the UI so I had to add that. I'll likely won't go in that direction, sorry. Myself I very much dislike the software feel to vam and the last thing I want to do is lean towards file management, especially with how quite a few people use vam with terabytes of content. To me that stuff is "wide as an ocean, deep as puddle" kind of thing, you don't even open or care about 99% of that stuff and will likely never even load or interact with it ever again. It's just there because there's nothing much you can do in vam. I'd rather try to build stuff to do

thomas d

I gave this a quick try this morning, and the Chat did not work. I selected a android and tried to interact, and it gave blank lines in the chat screen and did nothing. Not sure if it is on my side or not, as there is a lot of new stuff. I guess you have not had the chance to set it up yet to work? I liked the button to open and close the watch UI completely. It also over wrote my rive script and settings, luckily I had a recent back up. In 44, I did add my rive script and put it in the mainmenu.html file and restarted, but still got the same blank lines and the chat did not work. Maybe I did something wrong, and if I have time I will try again over the weekend. I went back to 43 as chat works for me there and I like that feature a lot.

Jeevz

Spqr going above and beyond as usual, had some excellent conversations with my android girls already!

Jeevz
>>111350914

You need to set up oobabooga as listed in the post to make it work. I've had several chats with several different androids already, works great.

Anonymous

The new android features are amazing! Unfortunately, the AI cam doesn't work for me. I have A1111 running with the --api flag, and the URL in the Services menu is set. Whenever I click 'Capture' or 'Preview' in the AI cam menu, I get this exception in the message log: err: System.NullReferenceException: Object reference not set to an instance of an object at SPQR.FaeScreenshot.stafxbac (Int32 quality) [0x00000] in :0 at SPQR.FaeScreenshot.Planfweab () [0x00000] in :0 at SPQR.FaeScreenshot.DAZMorphekdg () [0x00000] in :0 at SPQR.Aeternum.Update () [0x00000] in :0

SPQRAeternum
>>111362061

Were you on desktop, vr or desktop+vr? the latter might not work. Does it always return that error? Like if you reload the plugin, does it fix it? Also, can you try taking a normal vam screenshot first and then check if works after that? It might be something weird about the screenshot camera, I'm not entirely sure how vam handles that. If you have MacGruber's supershot active, that might not be compatible as it works with the screenshot camera also

thomas d

I tried again, but the person does not speak, I can say turn left, and the person turns left, but she says nothing. I added my rivescript file and when I say hello there, I do see it in the chat, but I she does not reply. When I test my voice in services, it works. I believe I turned on all I am supposed to. even when I type in a message and select send, I then see what I said in the chat, but the person says nothing and see no text response in the chat. Not sure what is going on. I am using the Local Text to speech & speech recognition. If I go back to v43 it works.

chris r

Hey you frackin coding GENUIS, SPQR, from this threadhttps://hub.virtamate.com/threads/alive.17856/post-114863 where you shared this video https://www.youtube.com/watch?v=P1IcaBn3ej0 is it possible to run this (or anything) on VAM to increase photo realism? Thanks!

SPQRAeternum
>>111430193

lol no code genius here, I broke some person management stuff unfortunately in v44. But v45 is almost ready I don't think stuff like that in that gta video is yet ready for real time use. They spent a lot of time to render/process it, maybe minutes for a frame. There's new stuff coming up right now and people create videos with AI like https://www.youtube.com/watch?v=t8fsOEzEnbk but they're slow to generate. Realtime use I think it's years away, but who knows. Last year almost nobody would have believed all the text and image ai going on right now, so things might move faster

chris r

Even prolific brilliant programmer's make mistakes because they're still human.......I think - thanks for the quick reply!

thomas d

Ok, I got the nvidia card and got oogabooga working. I was curious, so if I download a model, and change the different to a different android, it effects how the model answers and so forth correct? I am new to all this so, I was curious how much of the response is because of the Android settings, and how much the model. I mean guess the android setting just effect what the specific model is already capable of. I still seem to get a lot of loops, also, saying numbers after what she replies. Also, will start talking, and then reply to my comment. Or just reply the same way over and over. This happens a lot. I select new android, it works for a short time and then just repeats. I know this is all new and your still setting it up. So, far my best luck is with Yhyu13/Guanaco-gptq-4bit, trying different androids. I am going to try it with ravescript as well and see how that works out.

thomas d

I tried it with the local chat rivescript. It does work, however, if it is there in the rivescript, she starts to reply, but then cuts it off with what she is saying from oogabooga. even without a android it would cut off and say what oogabooga was going to say. I tried the Andreia Barns android, and it actually worked sometimes, but think its because the android has no reply. Or it seemed to work, if the was no reply in the rivescript , but oogabooga, had one. Anyway, I know your still working on this stuff. I think at one point, you said if it was found in the local it would use that and if not the oogabooga reply. that would work really well I think if its possible. This is all interesting stuff, Thank You!

Jeevz
>>111487501

If you got that 3060 with 12gb of vram i'd say try this model: https://huggingface.co/Monero/Pygmalion-Metharme-7b-4bit-TopScore I get reliable answers speeds from it, and it usually takes quite a long time before the AI loses track of the story/history and starts making mistakes before I have to clear chat history and start over. I have also played around with this model; https://huggingface.co/anon8231489123/vicuna-13b-GPTQ-4bit-128g But it requires allot more GPU resources and unless you run VaM in low settings mode it wont run. The answers from the vicuna 13b model are allot more longer, and descriptive compared to that 7b model, but, it's downside is that with a 3060 with only 12gb vram, it will slow down significantly after only a couple minutes of conversation due to reaching it's vram limit. I'd say download both models and give them a try. If you edit your webui.py file and change the line where it says "CMD_FLAGS =" to CMD_FLAGS = '--chat --model-menu --pre_layer 32 --no-stream --api' You'll get a quick menu when you start up oobabooga and you can choose which model to load. It helps greatly imo, when testing out multiple models. To answer your question regarding " if I download a model, and change the different to a different android, it effects how the model answers and so forth correct? " The model decides the type of language it will prefer (pygmalion models for example can output X-rated type of dialogue, where as other types of models are more limited in this, and thus will avoid such type of conversation, regardless of how you configure your android, you'd just reach a limitation of the model), however, what the android settings will control is the "personality" and choice of words (within it's limits, set by the model) of said android. The model is the brains, and the android settings are the persona. I hope this clears things up a bit Ps* I found that, starting up oobabooga, and loading in a model first, while your Vram is empty, because you have not started up VaM yet, HELPS, allot. As it seems to then, prioritize the AI, in computational power, over VaM, once you start having conversations with the androids.