Data compression and high-performance calculation method in multi-round listening dialogue model

The invention relates to a data compression and high-performance calculation method in a multi-round listening dialogue model. The implementation stages of the method comprise a data preprocessing stage, a data vectorization stage, a vector aggregation stage, a model fitting stage and a parallel computing stage. The technology involved in the invention comprises the steps of pre-training a language model, a deep recurrent neural network, an attention mechanism and parallel calculation. According to the technical scheme provided by the invention, the language model has strong dialogue strategy constraint and enough knowledge breadth at the same time in limited calculation performance and development cycle. 本发明涉及一种多轮倾听对话模型中的数据压缩与高性能计算方法。该方法实现阶段包括：数据预处理阶段、数据向量化阶段、向量聚合阶段、模型拟合阶段、并行计算阶段。本发明涉及的技术包括：预训练语言模型、深度循环神经网络、注意力机制、并行计算。本发明中所提出的技术方案在有限的计算性能与开发周期内，使得语言模型同时具备强对话策略约束与足够的知识广度。