Show HN: I built a tiny LLM to demystify how language models work

Show HN: I built a tiny LLM to demystify how language models work
544 by armanified | 59 comments on Hacker News.
Built a ~9M param LLM from scratch to understand how they actually work. Vanilla transformer, 60K synthetic conversations, ~130 lines of PyTorch. Trains in 5 min on a free Colab T4. The fish thinks the meaning of life is food. Fork it and swap the personality for your own character.

Why Switzerland has 25 Gbit internet and America doesn't

Why Switzerland has 25 Gbit internet and America doesn't
515 by sschueller | 382 comments on Hacker News.

Caveman: Why use many token when few token do trick

Caveman: Why use many token when few token do trick
534 by tosh | 270 comments on Hacker News.

German implementation of eIDAS will require an Apple/Google account to function

German implementation of eIDAS will require an Apple/Google account to function
497 by DyslexicAtheist | 463 comments on Hacker News.

The threat is comfortable drift toward not understanding what you're doing

The threat is comfortable drift toward not understanding what you're doing
552 by zaikunzhang | 396 comments on Hacker News.