Code and data of the Fine-Grained R2R Dataset proposed in paper Sub-Instruction Aware Vision-and-Language Navigation