Senior Staff Engineer and Manager at Google who leads teams to build parameter efficient tuning techniques.
How media typically covers Ming Zhang
Referenced in coverage
Random masking of parameter updates in adaptive optimizers like RMSProp outperforms state-of-the-art optimizers; a new method called Magma reduces perplexity by 19% for 1B models compared to Adam.
“Ming Zhang is a co-author on Google's research on random masking for LLM optimization.”