Xianfeng Zeng

neural machine translation (2) large language model (2) model ensemble (2) knowledge distillation (2) machine translation (2) deep neural network (1) model training (1) data selection (1) data leakage (1) instruction fine-tuning (1) layer normalization (1) data filtering (1) gradient normalization (1) back translation (1) machine translation evaluation (1) n-gram metric (1) transformer model (1) reference diversity (1) sequence generation (1) overestimation problem (1)