large-scale training
This practice involves training models on extensive datasets typically requiring significant computational resources. Effective large-scale training strategies are essential for developing robust models that can generalize well across diverse scenarios.
- ConTextTab: A Semantics-Aware Tabular In-Context Learner
- LabUtopia: High-Fidelity Simulation and Hierarchical Benchmark for Scientific Embodied Agents
- ThermalGen: Style-Disentangled Flow-Based Generative Models for RGB-to-Thermal Image Translation
- Understanding Differential Transformer Unchains Pretrained Self-Attentions
- Unified Scaling Laws for Compressed Representations