i'm so confused on the terminal & how you get yours to look like (venv) + ~/Desktop cd ai I'm running mine in Users>User disk but python3 -m venv venv command doesnt work
🎯 Key points for quick navigation: 00:00 *Un nouveau projet open source, le cadre de l'ordinateur auto-opérant, permet aux modèles multimodaux de contrôler votre ordinateur, éliminant ainsi les étapes manuelles.* 00:38 *Le projet exploite l'API Vision de GPT-4, promettant des améliorations rapides avec des contributions supplémentaires.* 01:49 *Python 3 est requis, et une clé API d'OpenAI est nécessaire pour l'utilisation du projet.* 03:32 *Les actions sur l'ordinateur sont entièrement automatisées après avoir donné une instruction vocale, démontrant le potentiel du projet malgré quelques bugs actuels.* Made with HARPA AI
Vision model is currently only supported on GPT 4. Also - this uses their API, not to be confused with ChatGPT. Both GPT 3.5 and GPT 4 are billed on token usage from their API. Cheers! 🥂
On Linux it will not run on python 3.8. And then I upgraded the venv defaulted to 3.11. Which then broke another dependency. Finally got it to work 3.9.
I'd be interested to see this working without a required API key. I think it should not be too difficult to work around, since even Llava GPT-vision can work without one.
A better idea is... 1 - get list of ai agents 2 - get a list of good things from all of them 3 - make new ai agents that combine all the good things from all the ai agent frameworks into one.
I've been wanting to do something like this for ages. This is so cool!!
Yes, I thought this was a very neat idea!! 💡
i'm so confused on the terminal & how you get yours to look like (venv) + ~/Desktop cd ai
I'm running mine in Users>User disk but python3 -m venv venv command doesnt work
yeah how does this work lol
can I use it with Huggingface or any other LLM?
🎯 Key points for quick navigation:
00:00 *Un nouveau projet open source, le cadre de l'ordinateur auto-opérant, permet aux modèles multimodaux de contrôler votre ordinateur, éliminant ainsi les étapes manuelles.*
00:38 *Le projet exploite l'API Vision de GPT-4, promettant des améliorations rapides avec des contributions supplémentaires.*
01:49 *Python 3 est requis, et une clé API d'OpenAI est nécessaire pour l'utilisation du projet.*
03:32 *Les actions sur l'ordinateur sont entièrement automatisées après avoir donné une instruction vocale, démontrant le potentiel du projet malgré quelques bugs actuels.*
Made with HARPA AI
🙏
It is possible to use GPT 3.5, which is free?
Vision model is currently only supported on GPT 4. Also - this uses their API, not to be confused with ChatGPT. Both GPT 3.5 and GPT 4 are billed on token usage from their API. Cheers! 🥂
@@DevelopersDigest Yes, I was wrong, but I've noticed that 3.5 is way more cheap than 4.0. Cheers!
NO
can you use it with lava model ?
Great question! I would have to take a look 👀
@@DevelopersDigest 4 months later, did you take a look?
@@DevelopersDigest 7 Months later did you take a look?
On Linux it will not run on python 3.8. And then I upgraded the venv defaulted to 3.11. Which then broke another dependency. Finally got it to work 3.9.
Thank you for leaving this comment! I am sure other Linux users will be glad to see this 🙂🙏
I'd be interested to see this working without a required API key. I think it should not be too difficult to work around, since even Llava GPT-vision can work without one.
You resumed why i hate python so much, the dependencies is a nightmare x)
only mac uses?
No it works on windows and linux also.
Input: "Improve upon your own source code" 🤔
I did that a while back on another early "agi" app. Wouldn't recommend it. It's cute until you don't recognize your own computer. 😂 full wipe
A better idea is...
1 - get list of ai agents
2 - get a list of good things from all of them
3 - make new ai agents that combine all the good things from all the ai agent frameworks into one.
I want this to apply for the endless amount of jobs while I sleep that i put in for at indeed and never hear from
Anyone got this working on windows ?
Yes
We need windows 12 with ai