dupa

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Medeksza

Pages: [1] 2 3 ... 82
1
Please try again and let me know if the issue is fixed, I optimized a database on the server which was running slow and causing logins to timeout. This should fix the "An unknown error occurred while connecting to the Zabaware server. Please check network connection, firewall settings and credentials." error

2
Ultra Hal 7.0 / Re: Online login sync missing.
« on: May 30, 2024, 08:31:15 pm »
The link at https://www.zabaware.com/download-beta.html is the most recent.

The error message in your screenshot is related to a wrong email/password, different from the error messages you posted earlier and also by other people which said "An unknown error occurred while trying to connect to Zabaware server".

So I believe your email or password may be wrong. Please try the "Forgot Password" link and change your password to something else. You can also test your email/password combo at www.zabaware.com by clicking on the "Sign In" link in the upper left. If you can sign in there, you should be able to sign in with the Ultra Hal app as well.

3
Ultra Hal 7.0 / Re: Online login sync missing.
« on: May 30, 2024, 03:54:02 pm »
Please try again and let me know if the issue is fixed, I optimized a database on the server which was running slow and causing logins to timeout.

4
Thanks Robert.... 8)

You need change the pricing on the Main website now that credits are cheaper... It still says $30 for 50,000 credits, or has it gone up since I last bought credits????

So it used to be 1 GPT-3 Currie query used up 10 credits per response (or 1 credit if Hal's cloud base brain can respond without involving the GPT API at all). Now it is set that one MythoMax (or LLAMA2) query uses 5 credits (half of the prior 10). So I didn't update the pricing, but it uses less credits per response so a credit basically goes twice as far as before.

I am using a 13 billion parameter version of MythoMax, but could offer customers the option to use a 70 billion parameter version for more credits per response if people want more brain power from the AI. But I find the 13 billion version seems to be good enough for general conversation and more cost effective.

5
Sorry for my slow reply. I just made another update to the brain/LLM model. It is now connected to a Llama 2 variant called MythoMax, which is fine-tuned for role playing characters, so should be a better match for Hal. I also have the repetition_penalty feature working, so some of the repetitive behavior should be gone now.

It costs me about half as much as the older GPT3 Curie did, so I updated the credit system to reflect that. GPT3 Curie responses used to take 10 Zabaware credits. I updated the MythoMax model to use 5 instead, so credits should go further now.

6
Thanks for the feedback. I made a few more minor tweaks to improve it a bit, but it still has an issue with repetitiveness. The llama2 model has a setting called "repetition_penalty" that should help alleviate this problem, but it seems whenever I try to use it, it crashes. Not sure why yet, will try to work on it to figure it out.

7
OpenAI shut off their GPT-3 Curie and GPT-Davinci models today that was the basis of Hal's neural net brain. OpenAI wants everyone to switch to the ChatGPT 3.5 or 4.0 models and doesn't want to support the older GPT3 models anymore. While I think these are both ChatGPT 3.5 and 4.0 are highly intelligent and very useful models, they don't work well as a basis for an entertainment chatbot like Ultra Hal. They refuse to take on the persona of a fictional character and are very highly censored.

So I switched models on the backend today to use Meta AI's Llama 2 model, the 13 billion parameter version (the mid tier). It is very comparable to ChatGPT 3.5 in terms of intelligence and a huge bump up in intelligence from the GPT-3 Curie model Hal was using. But more importantly, it is able to take on fictional personas, and doesn't force high levels of censorship (other than the material it was finetuned on). It provides a very good backend to build Ultra Hal personalities on.

If you have neural network credits with Hal, please give it a try, you should notice a change in personality and quality, hopefully for the better.

There are a lot more tweaks I can do to further improve it to take advantage of features the new model has, like an increase in content window size which will mean I can improve Hal's memory recall and learning ability a bit. I'll try to make the tweaks over the next few weeks, but please give me feedback along the way if you can.

8
Ultra Hal 7.0 / Re: Ultra Hal 7 for Android Beta
« on: May 15, 2023, 09:18:38 pm »
I then entered the same password I did with the HAL desktop.  Whoala!  Didn't work.  Full stop.
What exactly happens when you say it doesn't work. What is the message you receive?

Let me know if there is possibly an error I may have done through records?  FYI - I did do a password reset on the Forum.  But perhaps that does not effect HAL program?
The forum is separate from the Hal login. You can use the "forgot password" link in the app itself or https://www.zabaware.com/products/store/comersus_optForgotPasswordForm.asp if you want to try resetting the password associated with your Hal account.

9
Ultra Hal 7.0 / Re: Ultra Hal 7 for Android Beta
« on: May 14, 2023, 09:17:05 pm »
What exactly happens on screen when you try to login? Any error messages? Did it ask you to verify your email with a code? If so, did that verification code work? Did the forgot password link from the app send you an email?

I'll try to figure out why its not working with your current account/email, but do you perhaps have another email you haven't used with Zabaware before you can test with to see if it will create a new account for you (and grant you 1000 credits)?

10
Ultra Hal 7.0 / Re: Ultra Hal 7 for Android Beta
« on: May 11, 2023, 10:36:15 pm »
Finally set your bot's name, your name, genders and a few other options. Currently only "GPT3 Curie" works as the neural network engine and only "Memory Bank A" works for the memory (which is the same memory as Ultra Hal 7 for Windows Cloud uses). I plan on implementing support for additional "memory banks" soon so you can create different bots with different memories.

After setting these options you'll get to the final screen where you can chat with your bot. Type in the box below or use the microphone for speech recognition. Tap the speaker icon at top if you want to hear the bot speak (or leave muted for silent conversation).

If you come across any bugs or have any feedback/suggestions for improvements, please let me know.

11
Ultra Hal 7.0 / Re: Ultra Hal 7 for Android Beta
« on: May 11, 2023, 10:32:27 pm »
If you allow Hal to record audio (optional permission) you can use your phone built in speech recognition to talk to him (not all phones support this)

After installing you can use the + button to create a new bot profile

First select an animated character, then a voice, then a background

12
Ultra Hal 7.0 / Ultra Hal 7 for Android Beta
« on: May 11, 2023, 10:20:01 pm »
I apologize I haven't been more visible/active here for quite a while. In the background I have been working very slowly these past few years on creating a new cross platform Hal that can run on Android and IOS devices. I am happy to announce though that I have have a preliminary beta version that can run on Android devices.

If you are interested in testing you can grab the installation APK from:
https://a80db4ff6ae0d3c0fb9a-8e71e26de9dcc976c118d235118a125a.ssl.cf2.rackcdn.com/ultra-hal-7-beta-release.apk

This is not yet in the Google Play store as it is too early still, so will require "sideloading". Most phones block sideloading by default as a security precaution. If you are comfortable with doing it, here are instructions on how to sideload apps for Android: https://www.xda-developers.com/how-to-sideload-install-android-app-apk/

If you use the cloud based Ultra Hal 7 for windows, you can use the same username and password to login to the mobile version of Hal and it will share the same learned info and brain. It uses the OpenAI GPT3 Curie engine. If you don't have credits for GPT3 Curie then the system will still automatically give you 1000 free credits upon installing the mobile app, so should be good for testing at least.

I developed this mobile app using a system called Flutter, so it should actually cross-compile to iPhone (IOS) and also Mac and Linux as well. But I need to actually buy a Mac myself before I can compile an iPhone and Mac app and those systems absolutely do not allow compiling from Windows or sideloading, so maybe in few months I'll have beta version in the official app stores.

I'll post some screenshots and more details in a bit.


13
Ultra Hal 7.0 / Re: Recover my account/activation
« on: January 05, 2023, 12:06:10 pm »
I am working on a bug that certain Hal 7.5 users who have had Hal for over 5 years are running into. I am wondering if you are running 7.5 or an older version?

If its 7.5 then I am working on a fix that should be ready in a few days. If its not 7.5 then please try downloading it as you may not be effected by the bug I'm working to fix.

14
General Discussion / Re: trouble reinstalling my hal
« on: December 09, 2022, 08:53:02 am »
I responded to your emails, but maybe you're not getting them for some reason (maybe check your junk mail?). But anyway, I'll post my reply here:

Hal shouldn't be requiring purchase at all, should be free for all, so definitely a bug.

Thanks for including the debuglog.txt. After looking at the log and the source code, it looks like this bug occurs if you originally installed Hal over 4 years ago and then reinstall. I'll try to fix it in the source code sometime, but let's try this work around first:

  • First exit out of all instances of Hal (use task manager if you have to). Then as administrator, run "regedit" and navigate to:
        HKEY_LOCAL_MACHINE\SOFTWARE\WOW6432Node\Zabaware\Ultra Hal Assistant 7.0\Reg
  • Find the "InstallDate" entry and edit the date to something more recent like "2022-12-7"
  • Restart Hal and hopefully it will work for you.

15
Sorry about that. A security update to the Zabaware database server a few days ago shut down a service used by Hal. I have fixed it and it appears to be running normally again. Let me know if you have any more issues.

Pages: [1] 2 3 ... 82