sherpa VS GPT-4-Unlimited-Tools

I use Sherpa to run 7B and 13B LLaMA/Alpaca models on my phone: https://github.com/Bip-Rep/sherpa

If you are interested I made an app that runs on mobile phone that runs llama.cpp or vicuna. There is also a Mac and a windows version as it is built with flutter. https://github.com/Bip-Rep/sherpa

I just wanted to share that i was able to build an APK with recompiled llama as a shared c++ library. You can try it on my GitHub: https://github.com/Bip-Rep/sherpa

Let’s assume the singularity doesn’t happen, then what do we have left? I’ve been working with prompt engineering as of late focused on GPT4 it wasn’t that hard to make something like the other post the so called (SmartGPT) I’ve already worked on making inputs that include commands, systems and agents that act simultaneously to achieve an specific goal. These tools I’ve worked alone are extremely powerful to increase productivity and reasoning for any type of entity (take this into account), the only limits they currently hold is the lack of access to plugins, let me emphasize on this quickly, the current available plug-in models are running on ChatGPT 3.5 not 4, the power that these alone holds as it is is greatly understated from market forecast, precise analysis of resources and creation of advanced documents, these have already been started to be replicated by others in open source modality the current barrier for integration are costs to run the GPT4 API and the development of specific software focused for organizations (for reference I’ll leave this GitHub link that a friend developed).

I also made a web app that gives GPT-4 access to any commands you want, and you can add new ones on the fly: https://github.com/d3n7/GPT-4-Unlimited-Tools

Project link: https://github.com/d3n7/GPT-4-Unlimited-Tools