Multi-head attention and MLP blocks simplify transformer architecture learning

Technical course breaks down transformer model construction with multi-head attention and MLP blocks.