datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
nigerian-pidgin-1.0
Language:
- Nigerian Pidgin English (West African Pidgin variant)
Dataset Description
Dataset Summary
The Nigerian Pidgin ASR dataset (v1.0) is the first publicly released speech-to-text corpus for Nigerian Pidgin English, a widely spoken lingua franca across Nigeria and West Africa. This dataset comprises over 3,000 audio recordings paired with sentence-level transcriptions, recorded by native speakers across different genders and age groups. It is tailored for… See the full description on the dataset page: https://huggingface.co/datasets/asr-nigerian-pidgin/nigerian-pidgin-1.0.Pidgin_ASR_Dataset_Combined
Naija-ASR-Corpus v1.0 (NAC-v1.0)
A Foundational Automatic Speech Recognition Corpus for Nigerian Pidgin (Naija, PCM)
📌 Dataset Summary
Naija-ASR-Corpus (NAC-v1.0) is a speech dataset derived from the Universal Dependencies Naija Spoken Corpus (UD_Naija-NSC).
The NAC Team processed the original long-form recordings by:
Segmenting the audio into sentence-level clips.
Transcribing/Aligning the text to create paired audio-text data suitable for ASR training.
This dataset… See the full description on the dataset page: https://huggingface.co/datasets/timniel/Pidgin_ASR_Dataset_Combined.Pidgin_ASR_Dataset_Combined
Naija-ASR-Corpus v1.0 (NAC-v1.0)
A Foundational Automatic Speech Recognition Corpus for Nigerian Pidgin (Naija, PCM)
📌 Dataset Summary
Naija-ASR-Corpus (NAC-v1.0) is a speech dataset derived from the Universal Dependencies Naija Spoken Corpus (UD_Naija-NSC).
The NAC Team processed the original long-form recordings by:
Segmenting the audio into sentence-level clips.
Transcribing/Aligning the text to create paired audio-text data suitable for ASR training.
This dataset… See the full description on the dataset page: https://huggingface.co/datasets/AnonXx/Pidgin_ASR_Dataset_Combined.pidgin-asr-combined
Pidgin ASR Combined
A unified Nigerian Pidgin English speech-to-text dataset that combines
publicly available Pidgin ASR sources into a single train / validation /
test setup with a consistent schema. Built for fine-tuning Whisper-family
models on Nigerian Pidgin (Naija, pcm).
~8.6 hours, 4,278 clips, 10 source speakers, 16 kHz mono WAV.
Used to train michaelodafe/whisper-pidgin-v1
(21.37% WER on the test split, beating the published Wav2Vec2-XLSR-53
baseline by 8.2 pp).… See the full description on the dataset page: https://huggingface.co/datasets/michaelodafe/pidgin-asr-combined.pidginData
Naija-ASR-Corpus v2.0 (NAC-v2.0)
A Foundational Automatic Speech Recognition Corpus for Nigerian Pidgin (Naija, PCM)
📌 Dataset Summary
Naija-ASR-Corpus (NAC-v2.0) is a speech dataset derived from the Universal
Dependencies Naija Spoken Corpus (UD_Naija-NSC).
The NAC Team processed the original long-form recordings by:
Segmenting the audio into sentence-level clips.
Aligning each clip to its transcript (text_ortho) from the CoNLL-U source.
Tagging each sample… See the full description on the dataset page: https://huggingface.co/datasets/voicedata/pidginData.nigerian-pidgin-speech
Nigerian Pidgin Audio + Text Dataset for Whisper Fine-tuning
Nigerian Pidgin speech dataset for Whisper fine-tuning
Dataset Summary
This dataset contains audio recordings and transcriptions in Nigerian Pidgin English, designed for fine-tuning speech recognition models, particularly OpenAI's Whisper.
Dataset Structure
Train Split: 65 samples
Test Split: 8 samples
Total Duration: 0.0 hours (estimated)
Average Duration: 2.4 seconds per sample
Sample Rate: 16kHz… See the full description on the dataset page: https://huggingface.co/datasets/Rexe/nigerian-pidgin-speech.PidginMed-ASR
PidginMed-ASR
PidginMed-ASR is a Nigerian Pidgin English medical speech benchmark for automatic speech recognition.
Dataset Structure
The dataset contains audio recordings paired with Nigerian Pidgin transcriptions and English source translations.
Column
Description
audio
Path to audio file
transcription
Nigerian Pidgin transcript
english_translation
Original English text
speaker_id
Anonymized speaker ID
duration
Audio duration
split
Train, validation… See the full description on the dataset page: https://huggingface.co/datasets/AnnonSubmit/PidginMed-ASR.