4 ms·
In classic ML, you never evaluste against data that was in the training set. In LLMs, everything is the training set. Doesn't this seem wrong?
by brcmthrowaway 1y ago
In classic ML, you never evaluste against data that was in the training set. In LLMs, everything is the training set. Doesn't this seem wrong?