Study the paper
Scaling Laws for Neural Language Models
Lessons, visuals, quizzes, flashcards, and resources—organized in teaching order.
Optimal Allocation of Compute Budgets
Intersection point of scaling laws
Intersection point of scaling laws
The intersection point where the scaling laws suggest a potential contradiction or transition is characterized by a compute budget PF-Days, model size parameters, dataset size tokens, and loss nats/token.
Sources
S6.E8
C∗∼104PF-DaysN∗∼1012parameters,D∗∼1012tokens,L∗∼1.7nats/tokenformulae-sequencesimilar-tosuperscript𝐶superscript104PF-Daysformulae-sequencesimilar-tosuperscript𝑁superscript1012parametersformulae-sequencesimilar-tosuperscript𝐷superscript1012tokenssimilar-tosuperscript𝐿1.7nats/tokenC^{*}\sim 10^{4}\leavevmode\nobreak\ \mathrm{PF}{\text{-}}\mathrm{Days}\quad N^{*}\sim 10^{12}\leavevmode\nobreak\ \text{parameters},\quad D^{*}\sim 10^{12}\leavevmode\nobreak\ \text{tokens},\quad L^{*}\sim 1.7\leavevmode\nobreak\ \text{nats/token} (6.8)
C^{*}\sim 10^{4}\leavevmode\nobreak\ \mathrm{PF}{\text{-}}\mathrm{Days}\quad N^{*}\sim 10^{12}\leavevmode\nobreak\ \text{parameters},\quad D^{*}\sim 10^{12}\leavevmode\nobreak\ \text{tokens},\quad L^{*}\sim 1.7\leavevmode\nobreak\ \text{nats/token}| Measure | Value |
|---|
Sources
S6.E8
C∗∼104PF-DaysN∗∼1012parameters,D∗∼1012tokens,L∗∼1.7nats/tokenformulae-sequencesimilar-tosuperscript𝐶superscript104PF-Daysformulae-sequencesimilar-tosuperscript𝑁superscript1012parametersformulae-sequencesimilar-tosuperscript𝐷superscript1012tokenssimilar-tosuperscript𝐿1.7nats/tokenC^{*}\sim 10^{4}\leavevmode\nobreak\ \mathrm{PF}{\text{-}}\mathrm{Days}\quad N^{*}\sim 10^{12}\leavevmode\nobreak\ \text{parameters},\quad D^{*}\sim 10^{12}\leavevmode\nobreak\ \text{tokens},\quad L^{*}\sim 1.7\leavevmode\nobreak\ \text{nats/token} (6.8)
C^{*}\sim 10^{4}\leavevmode\nobreak\ \mathrm{PF}{\text{-}}\mathrm{Days}\quad N^{*}\sim 10^{12}\leavevmode\nobreak\ \text{parameters},\quad D^{*}\sim 10^{12}\leavevmode\nobreak\ \text{tokens},\quad L^{*}\sim 1.7\leavevmode\nobreak\ \text{nats/token}