Very Impressive!

[MentalGear]

Just tried out the 1.5b model via ollama and I'm super impressed by its speed and fluency!
I tried previously a 7b model which only generated about a token every 3 seconds on my testing system (intel mac), so this is a huge improvement.

Btw: Maybe consider adding a bluesky account - wanted to send some praise your way, but not supporting X/Twitter anymore due to obvious fascism.