Yes. The segmentation of sentences is critical with a lot of downstream integrations you might want to use (translation works best within an aws environment when called sentence by sentence, and doing any sort of sentiment analysis or entity detection is also affected by sentence breaks)