Transcribed output misses entire sentences

I'm experiencing an issue whereby the transcription output skips entire sentences that I've spoken roughly 30 to 40 percent of the time.