convergence guarantee
A convergence guarantee assures that as iterations of an algorithm proceed, the estimates or solutions it generates will approach a true value or optimal condition. This is key for ensuring that learning algorithms reliably find optimal solutions over time.
- AdaLRS: Loss-Guided Adaptive Learning Rate Search for Efficient Foundation Model Pretraining
- Breaking the Frozen Subspace: Importance Sampling for Low-Rank Optimization in LLM Pretraining
- Diffusion Federated Dataset
- Fast Non-Log-Concave Sampling under Nonconvex Equality and Inequality Constraints with Landing
- One for All: Universal Topological Primitive Transfer for Graph Structure Learning