I've downloaded iiab on ubuntu 24 and also ollama with gemma2 the two billion parameter version. i already have the api working in python for ollama. I needed to search the iiab library through an api search system that you have started building four years ago. I was wondering if i have your attention for the api calling in python. is it possible to share a snippet of code that i can use for the search api calling in iiab. you know the search system better than i do. after all you built it.
the reason I'm using iiab not google or the like, is for three reasons. if you're curious:
the internet is designed to be unwelcoming to bots in general as it is been used in malicious manner before LM became popular. so it is often blocked by security measures asking to prove that LM is a human.
the servers usually cap the speed limit of read to 3.5 megabytes a second as a standard protocol. the read speed on ssd is 500 megabytes a second. this allows for multiple searches for the intricacies of the topic and overall better understanding of it.
the internet is built for humans which are mainly relies on visuals and audio to understand which are not the main strong suites of LM. while the iiab library is mainly text based which are more suited for language models. this makes the google search less effective as the LM only process the text part which isn't enough to provide the full prospective of the matter being searched.
I hope i have your support. this system should be significantly more capable after the augmentation of the search ability. the idea is to have a small model and a large library and put them against a large model with an incompatible library and see which is better. if it turns out to be the small model can out perform the large model with the advantage of the library then this would completely change the way LM are built and will eventually lead to having local models that are very smart and very efficient as the power required to run a small local model is very little that i can have it run on a phone no problem.
I've downloaded iiab on ubuntu 24 and also ollama with gemma2 the two billion parameter version. i already have the api working in python for ollama. I needed to search the iiab library through an api search system that you have started building four years ago. I was wondering if i have your attention for the api calling in python. is it possible to share a snippet of code that i can use for the search api calling in iiab. you know the search system better than i do. after all you built it.
the reason I'm using iiab not google or the like, is for three reasons. if you're curious:
I hope i have your support. this system should be significantly more capable after the augmentation of the search ability. the idea is to have a small model and a large library and put them against a large model with an incompatible library and see which is better. if it turns out to be the small model can out perform the large model with the advantage of the library then this would completely change the way LM are built and will eventually lead to having local models that are very smart and very efficient as the power required to run a small local model is very little that i can have it run on a phone no problem.