Charlotte is currently pursuing a doctoral degree with ANITI and the MELODI team of IRIT as a part of a CIFRE fellowship. Her thesis, which directly contributes to the OpenLLM project, focuses on understanding how best to train multilingual models during the pretraining phase, looking at how different mixes of multilingual data during training impact the multilingual capabilities of a model. Charlotte completed her Master’s degree in computational linguistics at Jean Jaurès University in the fall of 2024. During her end-of-studies internship, she worked on Named Entities Recognition for materials and processes in unstructured textual data with large language models.