RadioTalk US Audio (Clean)

RadioTalk US Audio (Clean) Synthesized clean-speech audio for ~100k US air-traffic-control conversation scenarios. One row per turn, embedded 24 kHz mono PCM 16 WAV. This is the clean variant. A VHF-AM-channel-degraded variant is published as twangodev/radiotalk-us-audio-tada-noisy. Quick start from datasets import load dataset ds = load dataset(“twangodev/radiotalk-us-audio-tada-clean”, split=”train”, streaming=True) row = next(iter(ds)) print(row “text” … See the full description on the dataset page:

Source: Hugging Face Hub (twangodev/radiotalk-us-audio-tada-clean). Metadata imported from the dataset’s Hub tags.