Julia Gusak

Researcher@Inria·法国·OpenReview

研究方向

Efficient training · model parallelism · pipelining · checkpointing · rematerialization · offloading · data parallelism · Efficient inference · quantization · factorization · pruning · speed-up · compression · Tensor decompositions · Neural differential equations · neural ODE · Robustness · adversarial attacks · Speech synthesis · voice conversion