This assignment was a paper implementation of VQGAN+CLIP, a natural language processing machine learning model that was developed to generate and modify images. The model utilizes a multimodal encoder to process a sequence of text that would then become the input of the decoder that generates the image. Like Bubble2Floor, the code was provided by my instructor and I was tasked with training the model with various text and image data sets. Shown are some of the results of the generated inputs after the model was trained.