Dataset Viewer
Auto-converted to Parquet Duplicate
audio
dict
rich_caption
stringlengths
1.04k
4.97k
direction
stringclasses
3 values
category
stringclasses
3 values
example_id
stringlengths
41
164
dataset
stringclasses
1 value
source_utt_id
stringlengths
17
137
sample_rate
int32
16k
16k
channels
int32
1
1
audio_format
stringclasses
1 value
start_time
float64
0
20
end_time
float64
10
30
{ "bytes": [ 79, 103, 103, 83, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 156, 180, 119, 25, 0, 0, 0, 0, 120, 210, 164, 102, 1, 19, 79, 112, 117, 115, 72, 101, 97, 100, 1, 1, ...
The audio opens in a quiet, acoustically dry environment, likely a small office or studio with minimal reverberation. A faint, continuous electronic hum and low-frequency hiss are present, indicating the use of consumer-grade recording equipment. A single adult male, speaking in a General American English accent, begin...
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_015tu2cn_0001
clotho_train
clotho_analysis_015tu2cn_0001
16,000
1
opus
10
20
{ "bytes": [ 79, 103, 103, 83, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 148, 239, 67, 94, 0, 0, 0, 0, 166, 139, 7, 79, 1, 19, 79, 112, 117, 115, 72, 101, 97, 100, 1, 1, 56, ...
The audio clip begins in a quiet, acoustically treated room with a subtle, low-frequency electronic hum and a faint high-frequency hiss, characteristic of a studio environment. A male voice, positioned close to the microphone and speaking with a calm, neutral tone and a General American accent, says, "for the air." His...
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_015tu2cn_0002
clotho_train
clotho_analysis_015tu2cn_0002
16,000
1
opus
20
29.19
{ "bytes": [ 79, 103, 103, 83, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 140, 63, 145, 101, 0, 0, 0, 0, 128, 5, 181, 205, 1, 19, 79, 112, 117, 115, 72, 101, 97, 100, 1, 1, 56...
The audio begins with a male voice, amplified through a PA system, addressing a group of people in a large, reverberant indoor space. His tone is energetic and urgent, and he asks, “…last few, please. You’re all good.” The echo of his words reveals the size and hard surfaces of the venue, such as a gymnasium or hall. F...
und
speech
clotho_train/speech/und/clotho_analysis_01ay4ee5_0000
clotho_train
clotho_analysis_01ay4ee5_0000
16,000
1
opus
0
10
{ "bytes": [ 79, 103, 103, 83, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 132, 132, 220, 114, 0, 0, 0, 0, 196, 156, 86, 154, 1, 19, 79, 112, 117, 115, 72, 101, 97, 100, 1, 1, ...
The audio begins suddenly with a burst of loud, overlapping male voices, creating an energetic and chaotic atmosphere characteristic of a group in a large, reverberant indoor space such as a gymnasium or hall. At least three distinct male voices are heard, with the most prominent one positioned slightly left of center,...
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_01ay4ee5_0002
clotho_train
clotho_analysis_01ay4ee5_0002
16,000
1
opus
20
23.23
{"bytes":"T2dnUwACAAAAAAAAAAB8nlZVAAAAANJeEnABE09wdXNIZWFkAQE4AYA+AAAAAABPZ2dTAAAAAAAAAAAAAHyeVlUBAA(...TRUNCATED)
"The audio clip begins in a quiet indoor setting, marked by a persistent low-frequency hum and subtl(...TRUNCATED)
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_4bagpnkr_0000
clotho_train
clotho_analysis_4bagpnkr_0000
16,000
1
opus
0
10
{"bytes":"T2dnUwACAAAAAAAAAAB0bWR/AAAAADERXC8BE09wdXNIZWFkAQE4AYA+AAAAAABPZ2dTAAAAAAAAAAAAAHRtZH8BAA(...TRUNCATED)
"The audio begins with a faint, low-frequency thump and a subtle rustling sound, likely the result o(...TRUNCATED)
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_4bagpnkr_0001
clotho_train
clotho_analysis_4bagpnkr_0001
16,000
1
opus
10
20
{"bytes":"T2dnUwACAAAAAAAAAABs0R5uAAAAACOvyvoBE09wdXNIZWFkAQE4AYA+AAAAAABPZ2dTAAAAAAAAAAAAAGzRHm4BAA(...TRUNCATED)
"The audio clip opens with the continuous, low-level murmur of a bustling indoor environment, charac(...TRUNCATED)
und
speech
clotho_train/speech/und/clotho_analysis_5fqi2qcs_0002
clotho_train
clotho_analysis_5fqi2qcs_0002
16,000
1
opus
20
23.73
{"bytes":"T2dnUwACAAAAAAAAAABkqtJgAAAAAGK903MBE09wdXNIZWFkAQE4AYA+AAAAAABPZ2dTAAAAAAAAAAAAAGSq0mABAA(...TRUNCATED)
"The audio clip begins with the overwhelming energy of a large crowd gathered in a vast indoor stadi(...TRUNCATED)
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_5r1yc8d6_0000
clotho_train
clotho_analysis_5r1yc8d6_0000
16,000
1
opus
0
10
{"bytes":"T2dnUwACAAAAAAAAAABc2IBgAAAAAEnxvHIBE09wdXNIZWFkAQE4AYA+AAAAAABPZ2dTAAAAAAAAAAAAAFzYgGABAA(...TRUNCATED)
"The audio begins abruptly with a solitary male voice, his tone urgent and strained as he shouts, (...TRUNCATED)
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_5uiy8z7m_0000
clotho_train
clotho_analysis_5uiy8z7m_0000
16,000
1
opus
0
10
{"bytes":"T2dnUwACAAAAAAAAAABUO14XAAAAAOk0AMQBE09wdXNIZWFkAQE4AYA+AAAAAABPZ2dTAAAAAAAAAAAAAFQ7XhcBAA(...TRUNCATED)
"The audio opens within a vast, reverberant indoor space, most likely a church or auditorium, where (...TRUNCATED)
und,gen
speech
clotho_train/speech/und,gen/clotho_analysis_5uiy8z7m_0001
clotho_train
clotho_analysis_5uiy8z7m_0001
16,000
1
opus
10
20
End of preview. Expand in Data Studio

Bagpiper Pretraining Data

Bagpiper Pretraining Data is the public rich-captioned audio snapshot associated with Bagpiper, an open-ended audio language model that learns bidirectional mappings between audio and comprehensive text descriptions across speech, music, environmental sound, and mixtures.

The en metadata describes the primary rich-caption language. Source audio can contain speech or singing in other languages; it is not an English-only audio guarantee.

The repository contains 155,151,789 rows in 7,780 valid Parquet shards across 18 source-family directories, using 5.229 TB. It is intended for large-scale research workflows; select only the source families needed for your experiment rather than downloading the entire repository by default.

Repository snapshot

Source-family directory Rows Valid Parquet shards GB
audiocaps 35,206 2 1.217
audioset 1,346,522 68 47.973
clotho_aqa 4,193 1 0.130
clotho_train 29,434 15 0.905
emilia_en 15,664,702 784 528.702
fma 2,317,679 116 84.010
laion_audio_300m_part1 22,439,013 1,122 469.972
laion_audio_300m_part2 23,840,302 1,193 527.090
laion_audio_300m_part3 24,424,316 1,222 541.761
laion_audio_300m_part4 18,983,198 950 352.998
laion_captioned_ai_music_snippets 2,448,073 123 150.037
laion_in_the_wild_sound_events 294,199 15 7.071
owsm_v4_caption 20,225,474 1,012 1,570.800
wavcaps 2,051,781 103 67.056
yodas_auto 14,251,524 713 516.842
yodas_manual 3,495,090 175 128.507
youtube_8m_arkive 2,327,341 117 165.080
yt8m 973,742 49 68.900
Total 155,151,789 7,780 5,229.050

The clotho_train directory also contains 43 zero-row, schema-only .parquet.tmp files; they are temporary artifacts, are excluded from all counts above, and must not be loaded. These are physical repository statistics for revision 1997390aef538950729203ed4886b78b077d1a71. Directory names identify packing families and do not themselves establish ownership or a uniform license.

Loading selected Parquet families

Use explicit Parquet globs to keep data acquisition intentional:

from datasets import load_dataset

dataset = load_dataset(
    "parquet",
    data_files={
        "train": [
            "hf://datasets/espnet/Bagpiper_PreTrain_Data/audiocaps/*.parquet",
            "hf://datasets/espnet/Bagpiper_PreTrain_Data/clotho_train/*.parquet",
        ]
    },
    split="train",
    streaming=True,
)

print(dataset.features)
example = next(iter(dataset))

All audited valid shards use one shared schema:

Column Type Meaning
audio struct (bytes, path) Embedded encoded audio; path may be null
rich_caption string Machine-generated comprehensive caption
direction string Training eligibility: und, gen, or und,gen
category string Speech, music, or sound taxonomy
example_id string Stable packed example identifier
dataset string Source-family identifier
source_utt_id string Source utterance identifier
sample_rate int32 Audio sample rate
channels int32 Audio channel count
audio_format string Encoded audio format
start_time float64 Optional source-segment start offset
end_time float64 Optional source-segment end offset

direction denotes whether the pair is eligible for audio-to-text understanding, text-to-audio generation, or both; it does not imply duplicated physical rows. Inspect dataset.features before writing a consumer and ignore .parquet.tmp files.

The Hub datasets-server currently exposes only a default clotho_train view (29,434 rows), not the complete 18-family snapshot. Use the explicit glob approach above for the intended families.

Relationship to the paper

The Bagpiper paper describes a broader pretraining pipeline beginning from approximately 422 million raw audio-caption pairs, with clips capped at 30 seconds and captions generated by a Qwen3-Omni captioner. Its 600B figure is a training-token budget—300B text-to-audio, 150B audio-to-text, and 150B text-only—not the row count or byte size of this Hub snapshot.

This repository is a public artifact associated with that pipeline. Do not infer that its 18 directories are a complete one-to-one dump of every raw pair or every text-only source described in the paper.

Construction overview

Audio from heterogeneous speech, music, and sound collections is paired with rich captions generated by the paper's Qwen3-Omni-30B-A3B-Captioner. A Qwen3-32B classifier separates speech, music, and sound; text/audio quality and alignment filtering uses heuristic and LLM judgments, UTMOS for speech, AudioBox-Aesthetics for non-speech, CLAP alignment, Gumbel top-k sampling, and MinHash text deduplication. Rich captions can describe:

  • speech transcription, speaker attributes, language/accent, timing, and acoustic conditions;
  • instruments, genre, rhythm, melody, harmony, and musical structure;
  • environmental events, temporal ordering, spatial context, ambience, and recording quality.

Bagpiper uses these descriptions in both directions: audio-to-rich-caption for understanding and rich-caption-to-audio for generation, alongside text-only language-model data.

Limitations, provenance, and responsible use

  • Captions are machine-generated and may hallucinate content, timing, speaker attributes, or acoustic details.
  • Source families are heterogeneous and can include web-derived or transformed media, copyrighted speech/music, identifiable voices, and sensitive content.
  • This repository does not declare a blanket license. Users must review the terms, attribution requirements, privacy/consent constraints, and redistribution permissions of the underlying source represented by each row before use or further redistribution.
  • Directory-level public availability is not proof that every underlying asset is cleared for every commercial or biometric use.
  • The corpus is not exhaustively moderated for personal information, unsafe content, bias, or offensive language.
  • Generated captions and source media can inherit demographic, linguistic, geographic, and cultural biases.

For provenance corrections or takedown requests, use the repository community tab and provide the source-family directory plus a stable row identifier. Do not repost sensitive media in the report.

Related resources

Citation

@inproceedings{tian2026bagpiper,
  title={Bagpiper: Solving Open-Ended Audio Tasks via Rich Captions},
  author={Tian, Jinchuan and Wang, Haoran and Su, Bo-Hao and Huang, Chien-yu and Wang, Qingzheng and Shi, Jiatong and Chen, William and Gong, Xun and Arora, Siddhant and Li, Chin-Jou and Someki, Masao and Maekaku, Takashi and Goto, Keita and Shinohara, Yusuke and Sakuma, Jin and Yang, Chao-Han Huck and Watanabe, Shinji},
  booktitle={Third Conference on Language Modeling},
  year={2026},
  url={https://openreview.net/forum?id=FuHs64E3X6}
}
Downloads last month
14,958

Collection including espnet/Bagpiper_PreTrain_Data