sample diversity
Sample diversity refers to the variety and breadth of data examples used for training models. Diverse samples are critical for ensuring models generalize well across different scenarios and do not overfit to specific patterns.
- Align Your Flow: Scaling Continuous-Time Flow Map Distillation
- Counterfactual Evolution of Multimodal Datasets via Visual Programming
- GRIP: A Graph-Based Reasoning Instruction Producer
- Inference-Time Scaling for Flow Models via Stochastic Generation and Rollover Budget Forcing
- Off-policy Reinforcement Learning with Model-based Exploration Augmentation
- Señorita-2M: A High-Quality Instruction-based Dataset for General Video Editing by Video Specialists
- Solving Inverse Problems with FLAIR
- Towards Syn-to-Real IQA: A Novel Perspective on Reshaping Synthetic Data Distributions