Get the latest tech news
Chinchilla Scaling: A Replication Attempt
Hoffmann et al. (2022) propose three methods for estimating a compute-optimal scaling law. We attempt to replicate their third estimation procedure, which involves fitting a parametric loss function to a reconstruction of data from their plots. We find that the reported estimates are inconsistent with their first two estimation methods, fail at fitting the extracted data, and report implausibly narrow confidence intervals--intervals this narrow would require over 600,000 experiments, while they likely only ran fewer than 500. In contrast, our rederivation of the scaling law using the third approach yields results that are compatible with the findings from the first two estimation procedures described by Hoffmann et al.
View a PDF of the paper titled Chinchilla Scaling: A replication attempt, by Tamay Besiroglu and 3 other authors We attempt to replicate their third estimation procedure, which involves fitting a parametric loss function to a reconstruction of data from their plots. In contrast, our rederivation of the scaling law using the third approach yields results that are compatible with the findings from the first two estimation procedures described by Hoffmann et al. From: Tamay Besiroglu [ view email][v1] Mon, 15 Apr 2024 19:19:56 UTC (947 KB)
Or read this on Hacker News