Will I lead a completed pretraining of a >=1B param language model before EOY 2024?
Basic
1
Ṁ80
Jan 1
87%
chance

Must be trained on at least 100B tokens, and start from random initialization. Distillation is okay only if it meets these requirements.

I'll cast an initial guess vote and then no longer participate further in this market

Get
Ṁ1,000
and
S1.00