Recognizing handwritten text is a challenging task due to the various writing styles and quality of the text. Deep learning approaches have shown promise in recent years, specifically CNN and RNN. In this study, we introduce a new approach to Handwritten Text Recognition (HTR) called the Hybrid CNN-Gated Recurrent Unit (GRU) method on the Washington Database. Our proposed model uses a CNN for feature extraction and a GRU for sequence modeling. The CNN extracts meaningful features from the input image, while the GRU models the temporal dependencies between the features to recognize the handwritten text. We tested our model on the Washington Database and compared it to several other state-of-the-art HTR models. The results demonstrate that both the CNN and GRU components are essential for achieving optimal HTR performance. In conclusion, our hybrid CNN-GRU approach provides a promising solution for HTR on the Washington Database and can be extended to other HTR tasks.