Hello 👋

💻 I am a research scientist working at NuMind on structured information exctraction with language models. I specifically work on model evaluations and multimodality (text, images, pdf...).
🎸 Before that, I worked on music generation topics at the Metacreation Lab and at Sorbonne University where I got my PhD. I notably worked on the tokenization of symbolic music, and created the MidiTok library which I still maintain to this day, and helped to build and release the GigaMIDI dataset.
🌴 I can be contacted on the plateforms linked above.

Research interests:

• Natural language generation (NLG)
• Discrete generative models
• Music generation