This repository contains the training and evaluation code accompanying the thesis "Learning with Less: Contrastive Weight Tying on the BabyLM Challenge" (Ino van de Wouw, VU Amsterdam, 2025). The project studies headless language models, models pretrained with Contrastive Weight Tying (CWT) (Godey et al., 2024) instead of a standard cross-entropy
glueslurmpytorchtransformercomputational-linguisticsgptbertblimphuggingfacepretrained-language-modelpytorch-lightningcontrastive-learningsample-efficient-rlllmbabylmsample-efficient-mlcontrastive-weight-tyingheadless-language-modelssample-efficient-learning
-
Updated
Jul 26, 2026 - Python