Numerical Analysis
Core Concepts
- Floating-point precision
- Numerical stability
- Gradient clipping
- Learning rate scheduling
Applications in Large Models
Preventing Gradient Explosion / Vanishing
- Large models are deep and computationally intensive, making numerical stability particularly critical.
BFloat16 / FP16 Training
- Understanding how different floating-point precisions affect model training.
Optimizer Selection
- Some optimizers are numerically more stable.
贡献者
这篇文章有帮助吗？
最近更新
Involution Hell© 2026 byCommunityunderCC BY-NC-SA 4.0
