Post AcBZ5258NrSqKrssvw by prospero_san@www.minds.com
 (DIR) More posts by prospero_san@www.minds.com
 (DIR) Post #AcBZ5258NrSqKrssvw by prospero_san@www.minds.com
       2023-11-26T03:03:28+00:00
       
       1 likes, 0 repeats
       
       An interesting question. This paper estimates that roughly 9 trillion words of high-quality data are available for training. If large language models continue growing at recent rates, they’ll bump up against this limit around 2026.https://epochai.org/blog/will-we-run-out-of-ml-data-evidence-from-projecting-dataset