Peng Qiao

deep reinforcement learning (1) transformer architecture (1) contrastive learning (1) metric learning (1) few-shot learning (1) attention mechanism (1) relation extraction (1) distributed learning (1) policy learning (1) video understanding (1) parallel computing (1) distributed training (1) partial differential equation (1) vision-language model (1) speculative decoding (1) linear attention (1) neural operator (1) parallel computation (1) model acceleration (1) asynchronous communication (1)