4 comments

[ 2.7 ms ] story [ 22.8 ms ] thread
not free
We are a small team and request a small fee to prevent AI-based bots from registering, which is more effective than using reCAPTCHA

As shown on our website, any amount will be refunded on launch

So, if you are not using LLM, where did you manage to get so much data to train your AI? :)
We use public datasets, but Umme's core is not based on LLMs. Similar to Google AlphaGo, our approach relies neither on LLMs nor neural nets.