The complex flow characteristics of transverse jet in high-speed crossflow involve several separation regions and multiple shock waves, which make it difficult to capture and precisely predict the flow field state in real time merely by relying on traditional approaches. With the rapid advancement of deep learning technology, its powerful data processing capability offers a fast method for the prediction of the transverse jet flow field. Consequently, a prediction model based on deep learning is established, with the aim of obtaining the flow characteristics of a transverse jet under different freestream and jet conditions. This study segments the complex grid into several individual grids and trains them independently. The trained model can successfully establish the nonlinear mapping relationship between the transverse jet flow field and the input parameters. The prediction accuracy of the established model for the wall pressure under different conditions exceeds 99%, and the established model is also capable of reproducing structures such as shock waves and recirculation zones in the overall flow field, thereby achieving highly precise and efficient prediction of the jet structure and flow information. The results suggest that in contrast to the traditional numerical simulation, this deep learning model demonstrates greater efficiency in predicting the transverse jet flow field.