We now live in an era of generative artificial intelligence, where “Generative AI” is being used in almost all walks of life at least to some extent or the other. One of those is also in the domains of creativity, music, to be precise. This also has a great use case, not only in terms of pushing the limits of artificial intelligence as we know it, but also in terms of real world use cases such as for generating new samples of music or even for sparking creativity and act as inspiration. A few major problems exist in this field for any researchers, some of those being, the lack of a standard dataset and lack of standard evaluation metrics which is commonly used across various researches. This in turn prevents reproducible results and comparisons between the new developments in the field. Moreover, since music is such a subjective field, it’s difficult to evaluate if the generated music is actually good or not, hence more the need for standardization in terms of the evaluation criteria and metrics. Proper comparisons are necessary on same datasets and equal evaluation metrics to be able to understand and grasp the advancements that are being done in the field. Using the ideas of Natural Language Processing (NLP) techniques along with probabilistic and deep learning-based approaches, this study seeks to analyze music and it’s generation in an artificially augmented manner treating music as a pseudo-language, also referred to as symbolic language or symbolic music. Moreover, the study seeks to provide a baseline in terms of standardized dataset as well as metrics for any and all future studies in the same domain.