Datasets for training Text models using deep learning for text classification, text generation, embeddings, NER, etc.