Tim Tsz-Kit Lau | Kolar ML Labhttps://mkolar.coffeejunkies.org/authors/tim-tsz-kit-lau/Tim Tsz-Kit LauHugoBlox Kit (https://hugoblox.com)en-ushttps://mkolar.coffeejunkies.org/media/authors/tim-tsz-kit-lau_hu_55f2c44fe37a7ea8.jpgTim Tsz-Kit Lauhttps://mkolar.coffeejunkies.org/authors/tim-tsz-kit-lau/AdAdaGrad: Adaptive Batch Size Schemes for Adaptive Gradient Methodshttps://mkolar.coffeejunkies.org/publication/lau-2026-adadagrad/Thu, 01 Jan 2026 00:00:00 +0000https://mkolar.coffeejunkies.org/publication/lau-2026-adadagrad/Adaptive Batch Size Schedules for Distributed Training of Language Models with Data and Model Parallelismhttps://mkolar.coffeejunkies.org/publication/lau-2025-adaptive/Wed, 01 Jan 2025 00:00:00 +0000https://mkolar.coffeejunkies.org/publication/lau-2025-adaptive/Communication-Efficient Adaptive Batch Size Strategies for Distributed Local Gradient Methodshttps://mkolar.coffeejunkies.org/publication/lau-2024-communication/Mon, 01 Jan 2024 00:00:00 +0000https://mkolar.coffeejunkies.org/publication/lau-2024-communication/