Pretraining Objectives & Scaling Laws2 / 4
Pretraining Objectives & Scaling Laws
Given Chinchilla coefficients, return compute-optimal and for a budget (numerical optimization).
chinchilla_loss(N, D)Implement the function/class skeleton in the editor. Any correct approach is accepted.
import numpy as np
from scipy.optimize import minimize_scalar
def chinchilla_loss(N, D):
raise NotImplementedError
def optimal_ND(C, A=A, B=B, alpha=ALPHA, beta=BETA):
raise NotImplementedError
def optimal_ND_closed_form(C, A=A, B=B, alpha=ALPHA, beta=BETA):
raise NotImplementedError⌘/Ctrl↵to submit
No review yet
Submit your solution for a verdict, a score out of 100 and concrete feedback. Any correct approach passes.
Pretraining Objectives & Scaling Laws
Given Chinchilla coefficients, return compute-optimal and for a budget (numerical optimization).
chinchilla_loss(N, D)Implement the function/class skeleton in the editor. Any correct approach is accepted.