Dataset Viewer
Auto-converted to Parquet Duplicate
row_id
string
example_id
string
partition
string
source_subset
string
split
string
direction
string
messages
list
audio
audio
audio_sha256
string
audio_bytes
int64
audio_format
string
sample_rate
int32
channels
int16
num_frames
int64
duration_seconds
float64
metadata_json
string
source_revision
string
schema_version
string
part2_gen_v1_realistic:audiocaps_182_realistic
audiocaps_182_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
1a8da52794c3c29a44de43b2df973d679f7b4c25b50d6e5d73a79f9cc30fafb3
204,134
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":182,"persona":"a meditation app developer"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_185_realistic
audiocaps_185_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
c938017a9e521af59403c459bd66c4d30fa30f120617868b0d10814ee8a4a455
214,946
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":185,"persona":"a film maker working on a short film"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_188_realistic
audiocaps_188_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
c3cfc541b6d1f8f5208902f0d83e41b7ff5030abdc9b6b99638028d06d875dd6
200,075
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":188,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_190_realistic
audiocaps_190_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
304849d3bd640ed6f9d2e8d41dd3de59a80a9ada9cf34e09398a27c52cb43075
208,428
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":190,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_193_realistic
audiocaps_193_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
0ba44eaf4d0ff2d5a47076d7506f8721125fb3372e127a8d4f068fd4d590d9ab
167,906
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":193,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_195_realistic
audiocaps_195_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
44e081182a2587977ce4276a0bbe7164f9427afb719c65ef4eb8b7e4a6226778
220,422
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":195,"persona":"a meditation app developer"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_198_realistic
audiocaps_198_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
982c7e9aaf5b16414ec16a02ba5baed954ea582aa5880b5a0b2fd85cc78ebc16
172,814
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":198,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_204_realistic
audiocaps_204_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
55afca58a73a375d26c5257619d6fab57fd462cdbaed9c29ac5854f0c79a1122
96,609
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":204,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_206_realistic
audiocaps_206_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
f017f9afbce4aaa0d3655c718ea1fe22f3beab992bce960d7f2e8f26452c2a7a
174,289
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":206,"persona":"a DJ creating sound samples"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_221_realistic
audiocaps_221_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
45a0a4ae9a898572fd938dccf52eaaa5a51aa073927b46d5d349b46b6d985185
194,637
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":221,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_223_realistic
audiocaps_223_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
defccfa55d13093dc3f9eafec4934127710cfb6ea0f0088ab1a07f03438c0241
160,277
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":223,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_224_realistic
audiocaps_224_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
dabde24ccb76cbafcbb83628ae4d1cfd9b0085da4552b1659b94fe453ec0ca26
144,859
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":224,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_226_realistic
audiocaps_226_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
4881d5964fa5f55c5424faa3221020aa0b431025d86984f4c30b9b6aecd7716b
218,027
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":226,"persona":"a meditation app developer"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_229_realistic
audiocaps_229_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
0cb9721ba0de633b5b87dcb74bdc35458dcc2c94d3929eae522cbb3456c07056
157,820
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":229,"persona":"a game developer building an indie game"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_232_realistic
audiocaps_232_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
f50b672f7f6f5aef8310eecbfcc2b2bb04514f96d256d1564f4a6dad959549d1
254,234
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":232,"persona":"a VR/AR experience developer"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_234_realistic
audiocaps_234_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
178ae44acecf1ab49704a974ab896646db1601a282d502702e09fff4477bed43
172,636
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":234,"persona":"an advertising professional making commercials"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_236_realistic
audiocaps_236_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
4bf1ce92cee0bfc56ef7d411a104ee505c8e3343d5aef7dae489724df72ef3f8
184,937
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":236,"persona":"a meditation app developer"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_240_realistic
audiocaps_240_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
5ea964a844d12a7858cec12f26ec4741cbc2188a8a840a34ddd7542d4d68eeed
212,758
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":240,"persona":"a museum exhibit designer"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_243_realistic
audiocaps_243_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
31feed93048bbd84b6b22ff5975de1c16b2668fd7e9e657ac8bbef7b130991f5
172,123
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":243,"persona":"a sound designer for video games"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_247_realistic
audiocaps_247_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
59774626f493f94fc07b1e1e2f7e217827f87ac502b0804b7b607defe5b929ef
205,072
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":247,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_249_realistic
audiocaps_249_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
f0264b52673a13c7439e3aaba3a4102b38874da2be263caff2b4241fc9ee4f15
232,231
flac
16,000
1
159,840
9.99
{"dataset":"audiocaps","detail_level":"realistic","original_idx":249,"persona":"an advertising professional making commercials"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_251_realistic
audiocaps_251_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
c76097d9489a7e4a78dfb6e90c1a2b24dec2cf0f6755a1158014ebc85a699c64
126,252
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":251,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_254_realistic
audiocaps_254_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
0ba44eaf4d0ff2d5a47076d7506f8721125fb3372e127a8d4f068fd4d590d9ab
167,906
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":254,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_255_realistic
audiocaps_255_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
55afca58a73a375d26c5257619d6fab57fd462cdbaed9c29ac5854f0c79a1122
96,609
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":255,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_256_realistic
audiocaps_256_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
279388bbdf8799a14cac6d0faa68c5f9c864ff9c25034afbf868accb54df4eee
204,936
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":256,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_257_realistic
audiocaps_257_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
441399e92bad227f0a54099d70e983a431444b3ea8df97a9e1244e34b80ef403
160,295
flac
16,000
1
159,840
9.99
{"dataset":"audiocaps","detail_level":"realistic","original_idx":257,"persona":"a content creator making YouTube videos"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
part2_gen_v1_realistic:audiocaps_258_realistic
audiocaps_258_realistic
generation
part2_gen_v1_realistic
train
text_to_audio
[ { "role": "system", "modality": "text", "content": "You are a helpful assistant that generates audio based on user requests. You can create various types of audio including sound effects, music, speech, ambient sounds, and any combination of these. When given a request, first think through what the user...
1608c23a35254e8048668aee37942eda4c3d2d00fc0c519e662eb205641f7fee
200,508
flac
16,000
1
160,000
10
{"dataset":"audiocaps","detail_level":"realistic","original_idx":258,"persona":"a documentary filmmaker"}
37a5cf80ac3b989b61f7345b991ce7217627134f
bagpiper-self-contained-parquet-v1
End of preview. Expand in Data Studio

Bagpiper SFT Data

Release status: the validated Parquet release is being uploaded. The homepage and metadata may appear before every large shard is committed.

Bagpiper SFT Data is the supervised fine-tuning corpus for Bagpiper, an open-ended audio language model that understands and generates speech, music, environmental sound, and their mixtures through rich textual captions and planning.

The public release has exactly two configurations:

Configuration Direction Contents
generation natural-language request → planning, rich caption, audio Open-ended speech, music, sound, and mixed-audio generation sequences
understanding audio and request → rich caption, reasoning, answer Audio question answering, multiple-choice reasoning, and transcription sequences

The original source groups are recorded as provenance inside each row; they do not create additional public configurations.

Loading the data

Named Hugging Face configurations are deferred. Load a partition with an explicit Parquet glob; streaming is recommended for this 1.13 TB release:

from datasets import load_dataset

generation = load_dataset(
    "parquet",
    data_files={"train": "hf://datasets/espnet/Bagpiper_SFT_Data/generation/*.parquet"},
    split="train",
    streaming=True,
)
understanding = load_dataset(
    "parquet",
    data_files={"train": "hf://datasets/espnet/Bagpiper_SFT_Data/understanding/*.parquet"},
    split="train",
    streaming=True,
)

example = next(iter(generation))
audio_bytes = example["audio"]["bytes"]

Every row will be self-contained: the ordered text conversation and complete encoded audio bytes are stored in the same Parquet row. No external media download, filesystem path, or table join will be required. Data will be split across many deterministic Parquet shards targeting approximately 256 MiB each.

See Schema for the planned row format and Release status for the publication gates.

Data construction

Bagpiper SFT examples are synthesized from rich-captioned audio. Understanding sequences combine an input audio clip and task request with a rich caption, reasoning trace, and answer. Generation sequences reverse the direction: a request is expanded into planning and a rich caption before the target audio. The project uses large language models for request/reasoning simulation and quality filtering.

Release statistics

Partition Rows Shards Parquet bytes Embedded audio bytes
generation 1,474,011 1,944 363,645,830,327 411,563,826,060
understanding 1,193,006 4,860 768,716,849,416 1,251,578,476,222
Total 2,667,017 6,804 1,132,362,679,743 1,663,142,302,282
Partition source_subset Rows
generation part2_gen_v1_realistic 463,850
generation part2_gen_v1_imaginary 371,827
generation part3_gen_v1_realistic 95,134
generation part3_gen_v1_imaginary 72,508
generation part4_gen_v1_realistic 268,584
generation part4_gen_v1_imaginary 202,108
understanding airbench_train_v1 357,896
understanding mmau_train_v1 334,224
understanding asr_v2_inverse_200k 200,000
understanding audiobench_train_v1 300,886

Seven understanding rows sharing one header-only, zero-frame FoR_4308 WAV were removed after content audit. Intended reuse of valid audio across multiple rows is retained.

Known limitations

  • Captions, reasoning, and quality judgments are machine-generated and can contain hallucinations or biases.
  • The Bagpiper paper reports audio-fingerprint overlap between its SFT inputs and portions of AudioBench (12.8%) and AIR-Bench (3.2%). Results on those benchmarks should be interpreted with this disclosure.
  • Source audio has heterogeneous provenance and licensing. Configurations or examples without confirmed redistribution rights will be withheld rather than assigned an unsupported blanket license.
  • The corpus may include synthetic or transformed audio and should not be treated as verified human annotation.
  • The current understanding partition follows the checked-in recipe's v1 mixture. It is larger than the approximately 845k understanding sequences reported by the paper; paper-v2 alignment is deferred by owner decision.
  • The corpus is not exhaustively moderated for personal information, unsafe content, offensive language, or copyrighted text embedded in transcripts.

Related resources

Citation

Please cite the accepted Bagpiper paper when using this dataset:

@inproceedings{anonymous2026bagpiper,
  title={Bagpiper: Solving Open-Ended Audio Tasks via Rich Captions},
  author={Jinchuan Tian and Haoran Wang and Bo-Hao Su and Chien-yu Huang and
          Qingzheng Wang and Jiatong Shi and William Chen and Xun Gong and
          Siddhant Arora and Chin-Jou Li and Masao Someki and Takashi Maekaku and
          Keita Goto and Yusuke Shinohara and Jin Sakuma and
          Chao-Han Huck Yang and Shinji Watanabe},
  booktitle={Third Conference on Language Modeling},
  year={2026},
  url={https://openreview.net/forum?id=FuHs64E3X6}
}

Contact and takedown

Please use the repository community tab for data issues, provenance corrections, or takedown requests. Include the configuration and example_id; do not repost sensitive media in the report.

Downloads last month
43

Collection including espnet/Bagpiper_SFT_Data