Yura Gorishniy
@YuraFiveTwo
I do research on deep learning for tabular data
TabM now has a Python package! TabM is a simple and powerful DL architecture for tabular data that efficiently imitates an ensemble of MLPs 🏆 TabM has been used in winning solutions on Kaggle, and performs well on TabReD -- a challenging benchmark! 💻 pip install tabm 👇Link
I'd like to share our new diffusion distillation method, SwD, which produces few-step generators with progressive resolution scaling over the diffusion process. On SD3.5, SwD matches the speed of two full-size steps but with much better quality. Demo & models are released. (1/9)
I am excited to announce that I will join the University of Zurich as an assistant professor in August this year! I am looking for PhD students and postdocs starting from the fall. My research is on optimization, federated learning, machine learning, privacy, and unlearning.
We're excited to share that the 4th Table Representation Learning (TRL) workshop will be held at ACL 2025 in Vienna 🇪🇺. We invite papers on all things LLMs/ML and Tabular Data! …ble-representation-learning.github.io/ACL2025/. In 2025, TRL@ACL being in Europe and in summer complements TRL@NeurIPS ☀️.
For PyTorch users interested in TabM, we provide: 📄 A minimal single-file PyTorch implementation 💻 A Jupyter notebook with a usage example See the updated documentation: github.com/yandex-researc…
What DL architecture to try on tabular data? TabM is our new answer. TabM is leading on the benchmarks, while being simple, practical, and scalable to large datasets. 🧵In tabular ML, there is one thing that most people trust... 1/10
Also, TabM is our first tabular DL research project based on Pixi. I highly recommend trying Pixi instead of conda/pip/poetry. Pixi makes it so easy to maintain and share reproducible development environments! @prefix_dev is the amazing team behind Pixi!
What DL architecture to try on tabular data? TabM is our new answer. TabM is leading on the benchmarks, while being simple, practical, and scalable to large datasets. 🧵In tabular ML, there is one thing that most people trust... 1/10