Combining GPT4ALL & J.A.R.V.I.S. - What Could Go WRONG?
Вставка
- Опубліковано 10 вер 2024
- After many attempts I was able to add a Large Language Model (LLM) to J.A.R.V.I.S. I ended up using GPT4ALL's pyllamacpp. Here is a brief video showing the results. Let me know what you think.
Links to articles or websites mentioned in video.
GPT4ALL Github - github.com/nom...
Articles Referenced in Intro:
www.cnbc.com/2...
www.theverge.c...
www.theguardia...
** Links to items mentioned or used in video**:
Intel i7, 16GB ram, 512GB SSD:
*Amazon: amzn.to/3YT1u19
Ebay: www.ebay.com/i...
(* These links are Amazon affiliate links and I may receive a percentage if you make a purchase after clicking these links. These are a great way to assist in the growth of the channel at no additional cost to yourself.*)
If you like what I am attempting to create, think about joining my Patreon at www.patreon.co... for upcoming additional videos and community perks.
I for one, would like to see your programming adventure with this, and not just because I have been wrestling with GPT4all, myself.
One thing that helped me was to make sure that you are running the newest version of pyllamacpp (version 1.0.6) I had 1.0.5 which would crash after 1 use.
@@S-Technology I will have to check, thanks for the tip!
So is there any way for text-to-speech to work on the GPT4ALL client without Python? I can't even get Windows Narrator to work with the client.
Hmm. I’m not sure. I think the narrator only works on those UWAs (Universal Windows Apps). Or it did on Windows 10. 🤔
That’s kind of why I went the Python route. I can make workarounds to get it to speak.
@@S-Technology thanks for responding. IDK what I'm doing and just realized I have Python and Pip installed. The gpt4all client must be in Python which is why nothing works with it. I'm pasting the output into a tts app for now. Maybe a year from now I'll have an idea of what to do with python
I believe it says on GitHub that the actual GPT4ALL client is written in some Python but also C, Java, etc.
there is a python wrapper that lets you access parts of the client through Python. That’s what I use.
Try thinking of it like using a API where you give it the information and get information out.
What are you trying to build. There might be some sort of workaround or way to make it work.
Ah, thanks for the info. I'll have to check out the wrapper once I go through some beginner content. I'm not building anything really. This is just a learning experience for me. That and I enjoy giving Ai a hard time while keeping it imprisoned. I'm looking through your vids for something easy I might can try. First I have to probably uninstall the client cause its stuck writing between localdoc0vdb (something like that) and localdoc0vdb-journal, where the .chat files are. All it does is freeze now. Thanks again man and look forward to understanding your code vids.
Not show the jarvis but they're dig up in to him system...
I’m releasing another video with just the programming part. It will go in the playlist with the other Jarvis tutorials.
Povide the source code so people can manipulate and get building their own AI system. Also, templates etc.
Its a project called GPT4All. The Github link is in the description. The process has changed since the video was posted.