Training and Finetuning Sparse Embedding Models with Sentence Transformers