4 ms·
Number of parameters aside, I am really surprised that we havent yet reached hundreds of TB of training data. Especially Chinese model only used less than 10 TB
by eunos 5y ago
Number of parameters aside, I am really surprised that we havent yet reached hundreds of TB of training data. Especially Chinese model only used less than 10 TB of data.