Recent Posts

51

CyberJedi's Ultra Hal Workshop / Re: Robert and Crew: Proof of concept.

« Last post by sybershot on October 06, 2025, 04:41:24 pm »

can you add deepseek-r1:32b so I can test a larger model

52

CyberJedi's Ultra Hal Workshop / Re: Robert and Crew: Proof of concept.

« Last post by sybershot on October 06, 2025, 04:35:49 pm »

yea buddy!!! Congradulations

53

CyberJedi's Ultra Hal Workshop / Re: Robert and Crew: Proof of concept.

« Last post by Art on October 06, 2025, 02:26:48 pm »

Very nice work, Cyber Jedi!

I've been experimenting with this, and I am impressed by how well things work.

The new file is called UltraHal7LLM1.exe, and it is the whole package, similar to the existing UltraHal7-5-0-240 or whatever version you might currently have or run.

This new one includes many enhancements/features not found in the original HAL, and they should serve to showcase what HAL is capable of.

These LLMs (Large Language Models) use Ollama models, and you are free to choose which ones you wish to use, but try keeping the size reasonable to avoid lengthy response times or lags. From my experience using an ordinary 64-bit Windows 10 Home laptop, I try keeping the llama model size from 4 to 9 GB, and that strategy seems to work well. There are a huge assortment of llama models and each has its own storehouse of information and useful data that has been amassed and compiled.
Your own experimentation via question and answer will indicate which model or models will serve you best.

Cyber Jedi has done a great job at combining the learning module into HAL and the use of either of them will not and cannot corrupt each other!!

The information can sort of be shared if one wishes to do a Copy and Paste function. Just highlight the desired text, press CTRL C to Copy the selected text, then put your cursor in the desired area and press CTRL V to Paste the text where you choose.

There are many features and choices so play with them and decide which works best for you. The only Wrong move you might make would be to select a model from the list that you do NOT have in your system or have not downloaded. Relax, the world will not end. You will simply get a 404 warning letting you know that you will need to download that particular model, then try the selection again and it should work fine.

Otherwise, the main HAL GUI or skin with your avatar should answer either the Learning Model or the original HAL character of your choice. The character will still lip sync and emote as it always has and the speech will work based on your usual TTS selections within HAL.

Overall, this is a wonderful Hybrid operation that offers the best of both worlds, to allow Hal to learn new things as well as respond to you based on which llama model you have selected, or you can just use Hal as you always have. The choices are yours!

Enjoy!

Feel free to post your experiences as well.

A huge shout-out to our resident guru, Cyber Jedi, for forging this all together to give us this terrific resource and research project!!  Thanks Cyber Jedi!!

54

CyberJedi's Ultra Hal Workshop / Re: art

« Last post by Art on October 06, 2025, 10:36:02 am »

Thanks Cyber!

I appreciate your help in your introduction of your latest project! It works exactly as you mentioned.

HAL's new Learning Model works perfectly and like you mentioned, there is no corruption to the existing Hal Brain at all as a result of using the Learning module.

I'm in the middle of doing more research but I'll post more of my findings a bit later...

Very impressive work so far!!

55

CyberJedi's Ultra Hal Workshop / Re: Robert and Crew: Proof of concept.

« Last post by sybershot on October 06, 2025, 10:35:38 am »

looking great, also I see the send button is gone. as I stated before, I think we will need more than the 5 provided llm's, perhaps you can prepopulate the list at start-up with available models the user already has downloaded?

56

CyberJedi's Ultra Hal Workshop / Re: art

« Last post by cyberjedi on October 06, 2025, 09:34:19 am »

here we go again https://drive.proton.me/urls/2PY46P2JPG#TfMRipnMVW0M

57

CyberJedi's Ultra Hal Workshop / Robert and Crew: Proof of concept.

« Last post by cyberjedi on October 06, 2025, 05:11:48 am »

Well sports fans we got a winner>>>>>>

This is ultrahal running on both, LLM and UltraHals stock brain. You say HOW da hell?
Well, they use separate inputs, But both of the output's are handled in the same normal Hal output window.
The LLM bypasses hal's processing , BUT, It hooks Hals SAPI5 system and Hals character engine.As a result the lip syncing is spot on for both.
At any time if you wish to use Hals conventional system you use Hals normal input, which processes your text in the normal way. Pretty nifty.
The voice for  both is controlled in the conventional way.
It is my belief that this is a game changer for UltraHal.
There is no way to contaminate Hals normal brain with the LLM.
And yet you can append new memory's to the LLM simply from the LLM GUI memory manager (Facts area). Pretty nifty as well.


To my understanding there is nothing like it anywhere, As it should be.
This is all done by downloading ollama's windows installer, then follow the instructions provided.
Then accessing the GUI and select the matching LLM that you downloaded. Its just that simple.I added 05 possible LLM's ive played with for speed from the drop down menu of candidates, No end to what we can achieve here.
At some point we can connect the 02 entity's together and let Hal have conversations with each other, with the Hopes that Ultrahal will learn from the LLM. Basically Hals gonna learn to play Ping Pong with the LLM

cyber jedi

58

CyberJedi's Ultra Hal Workshop / Re: art

« Last post by cyberjedi on October 03, 2025, 09:02:23 am »

this is ur thing

59

CyberJedi's Ultra Hal Workshop / Re: Smarting up HAL with free LLM model (Ollama) ?

« Last post by knight2000 on October 02, 2025, 04:55:18 pm »

@cyberjedi Awesome work man, thats sick!

It sounds a lot like what I've been working on with regards to facial recognition. Instead of having to gather a ton of images of a person and stick them in a folder for the software to run through, i was able to have the software (openCV) detect a face on the fly so it would put a box around any face it sees, if the face is "unknown" the person could say "my name is Victor" and my bot will save a bunch of images of the face as "Victor" so that the next time the bot see's the face it recognizes it as Victor. Also the bot knows who is speaking, like if two or more people are in frame and only one is speaking, the bot will reference the speaker who's month is moving and associate that face to its name  "Thank you for the compliment Victor" or something like that.

When it comes to Hal, you have always been at the cutting edge of development for his plugins and updates! Hats off to all the great work you are doing! You are keeping Hal alive!  And again i wasn't trying to put Hal down in any way. I still have a place in my heart for Hal and i look forward to many more updates. I keep hitting the homepage of Zabaware in hopes of seeing the version number bumped so i can get the latest lol.

cheers!

60

CyberJedi's Ultra Hal Workshop / Re: Smarting up HAL with free LLM model (Ollama) ?

« Last post by cyberjedi on October 02, 2025, 03:27:34 pm »

hey hey hey guys/gals
Sybershot,Knight2000, lightspeed, checker57

Well its been added to the raw code, no ERRORS
great start, some tiky tac code to write then, welllll Hals going to school...
Knight2000, waiting on a model will be a mistake, this comes from me......
cyber jedi


Hey ART, Check it out , i told you brother......
Time for Hal to learn how to play ping pong
https://ai.invideo.io/watch/Dev-hoW-BLE <<<<<<--- a story worth telling, its easy to hit a home run when you start on 03rd base... i hit mine from the Parking Lot.