Mixed precision training uses FP16 for most computations, FP32 master weights for accumulation, and loss scaling to enable accurate training of large DNNs with halved memory usage.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
GWT projects gradients into wavelet subspaces to compress optimizer states for memory-efficient LLM training while claiming performance parity with full-rank updates.
citing papers explorer
-
Mixed Precision Training
Mixed precision training uses FP16 for most computations, FP32 master weights for accumulation, and loss scaling to enable accurate training of large DNNs with halved memory usage.
-
GWT: Scalable Optimizer State Compression for Large Language Model Training
GWT projects gradients into wavelet subspaces to compress optimizer states for memory-efficient LLM training while claiming performance parity with full-rank updates.