6 citations · 6 across the 1 of their papers we have counts for
1 paper · 1 filter
Mouadh Yagoubi, Yasser Dahou, Billel Mokeddem +12
Existing benchmarks have proven effective for assessing the performance of fully trained large language models. However, we find striking differences in the early training stages o…