181 citations · 965 across the 49 of their papers we have counts for
Showing cs.CLShow all
2 papers · 1 filter
cs.CL2024
BabyLlama-2: Ensemble-Distilled Models Consistently Outperform Teachers With Limited Data
Jean-Loup Tastet, Inar Timiryasov
We present BabyLlama-2, a 345 million parameter model distillation-pretrained from two teachers on a 10 million word corpus for the BabyLM competition. On BLiMP and SuperGLUE bench…
cs.CL2023★ 2 cited
Baby Llama: knowledge distillation from an ensemble of teachers trained on a small dataset with no performance penalty
Inar Timiryasov, Jean-Loup Tastet
We present our submission to the BabyLM challenge, whose goal was to improve the sample efficiency of language models. We trained an ensemble consisting of a GPT-2 and small LLaMA…