Reinforcement learning model processing method and device, computer equipment and storage medium

The invention relates to a reinforcement learning model processing method and device, computer equipment and a storage medium. The method comprises the following steps: when virtual characters of a plurality of different camps in a virtual environment interact, obtaining interaction data generated by interaction; performing feature extraction on the interaction data through a graphics processor and a central processing unit, and merging the extracted features to obtain role features; performing feature processing on the role features through a reinforcement learning model, and predicting an interaction behavior and a reward value corresponding to each virtual role; performing iterative training on a model associated with the reinforcement learning model based on a training sample including the role features, the interaction behaviors and the reward values; and when the trained model reaches a training stop condition, taking the trained model as a final reinforcement learning model. By adopting the method, the