This dataset card shows details about the Yapdo conversational speech corpus by Liva AI (YC S25). This dataset card details information for 30,000+ hours of recordings from 8,000+ speakers across 17 languages, with the rest of the hours still undergoing QA (estimated 50k total). The source audio is natively recorded with separate speaker channels; the samples here are presented as combined conversations.

The strength of this dataset is its naturalness. Recorded among friends "in-the-wild," it preserves the spontaneity of real dialogue and supports the development of better conversational AI. It also reflects friendly interactions across cultures and captures realistic turn-taking dynamics, which are essential for training models that sound natural. We note one limitation is in the acoustic quality; however, a potential solution during collaboration is to collect pairwise data with better/studio-quality equipment to train a model that can enhance the data.

Dataset Configs

Each config maps to a single directory of audio files with an accompanying metadata.jsonl.

Group	Configs	Description
Languages	`languages_ar`, `languages_arz`, `languages_bn`, `languages_es`, `languages_fr`, `languages_gu`, `languages_ha`, `languages_hi`, `languages_ig`, `languages_it`, `languages_kn`, `languages_pa`, `languages_pcm`, `languages_si`, `languages_sw`, `languages_ta`, `languages_te`, `languages_tl`, `languages_ur`, `languages_yo`	A subset of our multilingual speech organized by language code. Metadata: `nisqa`
Accents	`accents_Arabic_influenced_English`, `accents_East_African_English`, `accents_Filipino_English`, `accents_General_American`, `accents_Indian_English`, `accents_Nigerian_English`, `accents_South_African_English`, `accents_West_African_English`	Accent-stratified English speech across 8 accent groups. Metadata: `nisqa`
Full-duplex	`full_duplex_backchanneling`, `full_duplex_interruption_detection`, `full_duplex_pause_handling`, `full_duplex_turn_taking`	Recordings with full-duplex benchmark characteristics (backchanneling, interruption, pause handling, turn-taking)
Paralinguistics	`paralinguistics`	Paralinguistic elements (laughing, singing, elongated words, etc.). For the sake of displaying these elements, we identified emotion-rich segments using Empathic-Insight-Voice-Small since we observed that emotion-rich segments were likely to contain such elements. Metadata: `emotion`, `peak_score`, `avg_mos`
Random 50	`random_50`	50 completely random samples from our QA'd pool. Metadata: `nisqa`, `language`, `accent`, `speech_ratio`
Gaming	`gaming_assistant`, `gaming_companion`	Gaming conversations: helper-type conversations vs. companion. Metadata: `game`, `summary`

Dataset Overview


Total audio	31,592 hours
Unique conversations	31,822
Unique speakers	9,779
Languages	~17
Speakers per conversation	2–13 (avg 2.7)
Conversation duration	19s – 24.4 hrs (avg ~60 min)
Code-switching	~25% of conversations
Speech type	Spontaneous, unscripted, multi-party conversations
Quality score (NISQA)	1.8 - 4.8 (avg 2.5)
Common topics	Video games, daily life (jobs, school, relationships, earning money)

Languages

Language labels for each conversation were reviewed by a native human speaker.

Monolingual Conversations

17 languages with over 10 hours of monolingual conversation data. The below includes an estimation of the number of hours.

Language	Code	Conversations	Hours
English	en	15,044	12,950.8
Egyptian Arabic	arz	1,695	1,617.4
Spanish	es	1,085	1,529.4
Swahili	sw	1,412	908.4
Nigerian Pidgin	pcm	818	769.2
Hindi	hi	858	508.0
Arabic	ar	598	481.5
Tagalog	tl	166	251.2
Tamil	ta	145	172.8
Hausa	ha	223	163.4
Yoruba	yo	200	161.3
Italian	it	261	150.4
French	fr	32	49.0
Igbo	ig	36	21.1
Telugu	te	18	19.5
Cebuano	ceb	12	14.0
Kannada	kn	17	13.9

Code-Switching Conversations

28 language combinations with over 10 hours of code-switching data, spanning roughly 25% of all conversations. The below includes an estimation of the groups and hours.

Language Group	Conversations	Hours
English + Nigerian Pidgin	4,607	5,667.1
English + Tagalog	1,589	2,430.9
Cebuano + English + Tagalog	766	1,164.5
English + Swahili	480	549.1
English + Yoruba	264	274.1
English + Hausa	191	261.3
English + Nigerian Pidgin + Yoruba	96	132.2
Arabic + Egyptian Arabic	102	107.5
English + Hindi	156	104.6
Hausa + Swahili	74	97.0
English + Hiligaynon + Tagalog	55	87.8
English + Hausa + Swahili	40	74.8
English + Tamil	59	68.8
Nigerian Pidgin + Yoruba	71	67.5
Hindi + Urdu	43	53.9
English + Igbo + Nigerian Pidgin	22	43.7
English + Hausa + Nigerian Pidgin	24	35.8
English + Spanish	26	34.3
English + Igbo	31	30.2
Arabic + English	31	30.0
Egyptian Arabic + English	30	30.0
English + Telugu	27	28.7
English + Nigerian Pidgin + Swahili	14	26.0
Igbo + Nigerian Pidgin	21	23.3
Cebuano + English + Hiligaynon + Tagalog	12	22.9
Nigerian Pidgin + Swahili	16	19.9
Hausa + Nigerian Pidgin	12	12.6
English + Hindi + Urdu	12	10.7

Accents

These labels are for English only and were obtained from the city that participants self-reported they were from. The below includes an estimation of the accent groups and hours.

Accent Group	Hours
Nigerian English	15,946.6
General American	778.4
Indian English	1,030.4
East African English	905.2
Filipino English	549.1
West African English	348.5
Arabic-influenced English	141.1
South African English	104.7

Labels

Language labels were assigned at the speaker-track level by native speakers who reviewed each individual track within a conversation. A single conversation may carry multiple language labels when speakers use different languages. Accent labels are derived from each speaker's self-reported city of origin, providing a natural geographic proxy for dialect and accent variation.

Technical Analysis


Sample rate	48 kHz
Bit depth	16-bit PCM
File format	WAV
Mean SNR	~33 dB
Median RMS	-26 dBFS
Average speech ratio	0.35
Spectral centroid	~0.66 kHz
Frequency content	3.3 kHz (averaged over 10–30 second clips)

Combined vs. Separated Audio

Each sample in this dataset is a combined mix of all speakers. The parent Yapdo corpus stores each speaker on a separate, time-aligned track. Here's what that difference sounds like — a Telugu conversation with 2 speakers:

Combined (all speakers mixed)

Speaker 1 (isolated track)

Speaker 2 (isolated track)

Audio Artifacts

Source audio passes through Opus VoIP pipeline.

Artifact	Prevalence
Dropouts / packet loss	98.6%
Bandwidth ceiling (< 4 kHz)	97.2%
Clicks / pops	93.6%
Mains hum (50/60 Hz)	82.4%
Silence / dead air	34.6%
Frame repetition	18.2%
Echo	15.2%
Low signal level	5.8%
Onset transients	5.2%
Clipping	0.6%

Downloads last month: 57

Total file size:

1.06 GB