About Me
I am a research scientist at Meta, in the Modern Recommendation Systems - GPU Efficiency team. I received my Ph.D. from University of Illinois Chicago, where I worked on efficient neural network inference and training. More specifically, I have experience in the following areas: early exit networks, pruning, sparsity, quantization, knowledge distillation, semantic segmentation, large language models, mixture of experts and streaming models.
In the past, I had the pleasure of writing about Formula 1 news for damalibayrak.com. Nowadays, I am passionate about running and a little bit of paragliding. I enjoy watching all sorts of sporting events, but in particular I like road bicycle racing. I think correspondence chess is the best type of chess. Beyond these, I try my best to read and travel as much as possible.
News
- (12/2024) I started working at Meta as a research scientist in the Modern Recommendation Systems - GPU Efficiency team!
- (10/2024) I defended my Ph.D. thesis! Read it here!
- (06/2024) Our paper Class-aware Initialization of Early Exits for Pre-training Large Language Models has been accepted to WANT@ICML 2024.
- (05/2024) I will be at Google in Mountain View, CA this summer as a research intern! I will work on Project Starline.
- (04/2024) Our paper Class Based Thresholding in Early Exit Semantic Segmentation Networks has been published in IEEE Signal Processing Letters.
- (11/2023) Passed my preliminary exam. I am now a Ph.D. candidate!
- (10/2023) Presented Dataset Pruning Using Early Exit Networks in the Cohere for AI - ML Efficiency group meeting.
- (06/2023) Coming to Honolulu, HI for ICML 2023! I will present Dataset Pruning Using Early Exit Networks.
- (02/2023) I will be interning at Apple in the Core ML Tools team in Seattle, WA this summer! Super exciting!