A comprehensive evaluation of large Language models on benchmark biomedical text processing tasks | Synapse