piyushgupta53

Wrote a blog on how RL can make a model to produce better watercolor paintings via reinforcement learning

hey everyone, Opus 4.6 and I built this simple, lightweight yet effective personal ai agent that just works.

This was a fun project and it's working surprising well. I've implemented some nice patterns for memory and background tasks. Have a look, play with it and share feedback if any.

hey everyone, Opus 4.6 and I built this simple, lightweight yet effective personal ai agent that just works.

This was a fun project and it's working surprising well. I've implemented some nice patterns for memory and background tasks. Have a look, play with it and share feedback if any.

thanks for pointing this out. I've added this in my to-dos.

thanks for the suggestion, appreciate it.

not yet. I'll be adding soon.

it was challenging for sure. Took me almost a month to get acquainted with the protocol, how bencoding works etc, build a mental model and then eventually writing code.

Magnet and DHT are yet to be added.

this is such a great read!