Team-LEMON/lemon-train-code download history

Team-LEMON/lemon-train-code is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 80 times (3 in the last 7 days), and 589 times in total. It ranks #133,464 among datasets by monthly downloads.

ZEST: Injecting Inductive Bias at Tokenizer level Domain-informed tokenization enables small models to match or exceed much larger baselines. We demonstrate this across two domains: proteins (Nature's code) and programming languages (Human code). Key Result A 70M-parame

Open Team-LEMON/lemon-train-code on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.