I'm MS student in Computer Science at McGill and Mila, advised by Prof. David Ifeoluwa Adelani. I post-train LLMs, mostly multilingual models and RL for LLMs. The part I care about is what a model actually learns about a language, and whether that still holds once vision is in the loop. I release datasets and tools in the open. Before this I was an ML engineer at EAII and vella.ai.
In the past, I used to host research talks for the Cohere Labs community, with speakers including Zhijing Jin, Adaku Uchendu, Jack Morris, Niklas Muennighoff, Xingyao Wang, and others.