VE VectorWorks Academy Optimizing Distributed Training: Understanding Data, Pipeline, and Tensor Parallelism Introduction