Optimization and training dynamics · MLD-0025
Adam
An optimizer combining first- and second-moment gradient estimates with bias correction.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Train a neural language model.
Original caveat
Weight decay implemented as an L2 loss is not generally equivalent to AdamW.
Context and aliases
Level: Intermediate. Stages: Training · Theory.
Tags: ML · DL · SLM · LLM.
Aliases: Adaptive moment estimation.
Related concepts
Concepts linking here
Original references
- Deep Learning: online textbook ↗ · Consulted
- PyTorch optimization algorithms ↗ · Reference link
- PyTorch documentation ↗ · Reference link
- Original reference URL ↗