Introducing Madhuram-v0.6 - a 151M-parameter language model trained on just 600B tokens.
That's 3–10× less data than the models we put it up against. Here's how it stacks up on standard pretraining benchmarks 👇
My dad said "You’ll always be a boy to me and never a girl”
The table went silent.
My little cousin, slammed her fork down and said:
“No. She’s my aunt. And she’s pretty”
That one sentence did more for my soul than years of therapy ...
Sometimes kids get it better than adults ever will 🥺���️⚧️💕