RawAudio
Datasets
All datasets matching “RawAudio”raw_audio_with_audio_tokens_for_pretraining_just_tokensVoiceGiraffe-RawAudio
VoiceGiraffe — Raw Audio Repo
This is the gated audio companion to the public Jashin-Yeah/VoiceGiraffe benchmark. It contains the full-length source recordings referenced by the benchmark's test.jsonl.
⚠️ Access is gated: You must be logged in, fill in your information, and accept the terms above to request access.
Contents
audio/<youtube_id>.wav # full-length recordings, 16 kHz
Each file name is the audio id used in the benchmark's test.jsonl. To map a QA… See the full description on the dataset page: https://huggingface.co/datasets/Jashin-Yeah/VoiceGiraffe-RawAudio.raw_audioraw_audio_with_audio_tokens_for_pretrainingraw_audio_sliced_16khz
Armenian 7500-Hour Raw Audio Sliced (Gated Public)
This dataset contains ~7500 hours of unlabelled Armenian speech processed via Voice Activity Detection (VAD) and downsampled to 16 kHz (Mono, PCM_16).
Packaged into snappy-compressed Parquet shards of ~500 MB for direct compatibility with Hugging Face datasets and high-speed SSL pre-training (HuBERT, Wav2Vec2).
Access is Gated: Manual approval required.
rawAudioDataseto_of_SouthVietnam