Dataset Viewer (First 5GB)
Auto-converted to Parquet
mp3
audio
json
dict
__key__
string
__url__
string
{ "id": "6f5076dc-98a5-4312-a889-090a22bf664c", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 34.2324375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). Global hint: segments must not overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 8.18}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.89, \"end_time\": 16.48, \"caption\": \"The audio features a prominent electric eel sound, likely an electric shock. It's presented in 6-channel surround sound, suggesting a dynamic and immersive audio experience. The sound is the main focus, without any other discernible elements.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's happening at 14.44 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 10.89, \"end_time\": 16.48, \"caption\": \"The audio features a prominent electric eel sound, likely an electric shock. It's presented in 6-channel surround sound, suggesting a dynamic and immersive audio experience. The sound is the main focus, without any other discernible elements.\"}]}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 18.97, \"end_time\": 22.49, \"caption\": \"ship creaking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 24.07, \"end_time\": 26.67, \"caption\": \"sighs\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 27.83, \"end_time\": 29.09, \"caption\": \"tertawa\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 29.65, \"end_time\": 33.23, \"caption\": \"ominous orchestral music\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the first segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 8.18, \"caption\": \"Phacellodomus rufifrons\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000000_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6f5076dc-98a5-4312-a889-090a22bf664c", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 34.2324375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). Global hint: segments must not overlap. First, provide the start and end times for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 8.18}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.89, \"end_time\": 16.48, \"caption\": \"The recording prominently presents an electric eel tone, likely an electric discharge. It is delivered in six\\u2011channel surround sound, creating a vivid and immersive audio experience. No other discernible sounds are present.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What occurs at 14.44 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 10.89, \"end_time\": 16.48, \"caption\": \"A recording dominated by the distinct hiss or crackle of an electric eel, likely an electric shock. It is rendered over six\\u2011channel surround sound, creating a dynamic, immersive audio field. No other sounds are perceptible.\"}]", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 18.97, \"end_time\": 22.49, \"caption\": \"creak of the ship\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Please provide the original text, and I will paraphrase it while preserving the exact meaning and structure.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 24.07, \"end_time\": 26.67, \"caption\": \"A soft, prolonged exhalation, audible and deliberate, conveying release or fatigue.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, here is the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 27.83, \"end_time\": 29.09, \"caption\": \"### Original Text: tertawa \\n### Paraphrased Text: tertawa\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the subsequent segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 29.65, \"end_time\": 33.23, \"caption\": \"Dark, foreboding orchestral music with deep, brooding strings and tense, swelling chords.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the initial segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 8.18, \"caption\": \"Phacellodomus rufifrons\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000001_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6f5076dc-98a5-4312-a889-090a22bf664c_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 34.2324375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 8.18}, {\"start_time\": 10.89, \"end_time\": 16.48}, {\"start_time\": 18.97, \"end_time\": 22.49}, {\"start_time\": 24.07, \"end_time\": 26.67}, {\"start_time\": 27.83, \"end_time\": 29.09}, {\"start_time\": 29.65, \"end_time\": 33.23}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000002_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6f5076dc-98a5-4312-a889-090a22bf664c_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 34.2324375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 8.18}, {\"start_time\": 10.89, \"end_time\": 16.48}, {\"start_time\": 18.97, \"end_time\": 22.49}, {\"start_time\": 24.07, \"end_time\": 26.67}, {\"start_time\": 27.83, \"end_time\": 29.09}, {\"start_time\": 29.65, \"end_time\": 33.23}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000003_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "96ed2120-4cc6-4819-a3e2-a7a17db08da8", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 19.9758125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.86}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"chuckles\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next caption, aim for ~48 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.64, \"end_time\": 5.78, \"caption\": \"Glass debris falling and hitting a surface.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 6.29, \"end_time\": 7.61, \"caption\": \"groans\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment. Target ~12 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.2, \"end_time\": 10.85, \"caption\": \"upbeat music\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, aim for ~4 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 12.03, \"end_time\": 15.12, \"caption\": \"chattering, yelling\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's happening at 5.86 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": []}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Generate a standalone caption for the next segment. Target ~20 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"a(tm)aa(tm)aa(tm)a\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the first segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.86, \"caption\": \"chuckles\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000004_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "96ed2120-4cc6-4819-a3e2-a7a17db08da8", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 19.9758125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, provide the start and end points for the initial segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.86}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the caption for that specific part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"soft, low laughter\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Keep it under 48 characters for the next caption.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.64, \"end_time\": 5.78, \"caption\": \"Glass fragments descending and striking a surface.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end timestamps and a standalone caption for the following segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 6.29, \"end_time\": 7.61, \"caption\": \"deep, low-pitched, sustained vocalizations of strain or discomfort\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "<START>\n<END>\nNext Segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.2, \"end_time\": 10.85, \"caption\": \"upbeat music\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following part, target approximately 4 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 12.03, \"end_time\": 15.12, \"caption\": \"Chirping, shouting\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What occurs at 5.86 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[]", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Next up: the breakthrough moment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A repeating sequence of vowel sounds punctuated by brief consonant clusters: a \\u2013 tm \\u2013 aa \\u2013 tm \\u2013 aa \\u2013 tm \\u2013 a.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the initial part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.86, \"caption\": \"Soft, low laughter, a brief series of muffled, breathy chuckles.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000005_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "96ed2120-4cc6-4819-a3e2-a7a17db08da8_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 19.9758125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.86, \"caption\": \"chuckles\"}, {\"start_time\": 4.64, \"end_time\": 5.78, \"caption\": \"Glass debris falling and hitting a surface.\"}, {\"start_time\": 6.29, \"end_time\": 7.61, \"caption\": \"groans\"}, {\"start_time\": 8.2, \"end_time\": 10.85, \"caption\": \"upbeat music\"}, {\"start_time\": 12.03, \"end_time\": 15.12, \"caption\": \"chattering, yelling\"}, {\"start_time\": 16.5, \"end_time\": 18.98, \"caption\": \"a(tm)aa(tm)aa(tm)a\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000006_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "96ed2120-4cc6-4819-a3e2-a7a17db08da8_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "AudioData", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 19.9758125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.86, \"caption\": \"a brief, amused laugh.\"}, {\"start_time\": 4.64, \"end_time\": 5.78, \"caption\": \"Shattering glass fragments descending and impacting a surface.\"}, {\"start_time\": 6.29, \"end_time\": 7.61, \"caption\": \"a series of deep, resonant groaning sounds\"}, {\"start_time\": 8.2, \"end_time\": 10.85, \"caption\": \"Energetic, cheerful music.\"}, {\"start_time\": 12.03, \"end_time\": 15.12, \"caption\": \"chattering, shouting\"}, {\"start_time\": 16.5, \"end_time\": 18.98, \"caption\": \"a(tm)aa(tm)aa(tm)a\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000007_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "77c182d4-bc6e-4025-bb1b-9057df709882", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 34.209, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_only", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be ~15–40 seconds. Global hint: segments must not overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 33.21}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"Accipiter striatus\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the last segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 33.21, \"caption\": \"Accipiter striatus\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000008_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "77c182d4-bc6e-4025-bb1b-9057df709882", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 34.209, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_only", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: each segment should be approximately 15 to 40 seconds long. Global hint: segments must not overlap. First, provide the start and end times for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 33.21}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the caption for that specific part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"Sharp, high-pitched call of a northern goshawk, a rapid series of three to five clear, descending notes with a slightly raspy quality, typical of the species' alarm or territorial vocalization.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the final part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 33.21, \"caption\": \"Sharp, high-pitched call of a northern goshawk, with a series of rapid, descending notes followed by a prolonged, resonant shriek.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000009_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "77c182d4-bc6e-4025-bb1b-9057df709882_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 34.209, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 33.21, \"caption\": \"Accipiter striatus\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000010_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "77c182d4-bc6e-4025-bb1b-9057df709882_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 34.209, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 33.21, \"caption\": \"Accipiter striatus emits a rapid, high\\u2011pitched trill that repeats several times before fading into a deep, rasping note.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000011_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "27137059-0058-48f2-ba11-ccbe90027201", "meta": { "conversation_type": "iterative", "datasets_used": [ "SFX", "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 29.1166875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments may overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.7, \"end_time\": 28.12}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Generate a standalone caption for the next segment. Target ~17 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"Tufted Titmouse\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000012_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "27137059-0058-48f2-ba11-ccbe90027201", "meta": { "conversation_type": "iterative", "datasets_used": [ "SFX", "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 29.1166875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, provide me with the start and end of the first segment. Global hint: segments may overlap.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.7, \"end_time\": 28.12}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Please provide the original text, and I will generate the corresponding paraphrased content for you.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A small, high-pitched, whistled call resembling \\\"peter-peter-peter\\\" or \\\"peter-peter-peter-peter,\\\" often repeated in a rhythmic pattern, with a slightly nasal quality and a clear, sharp tone, typically heard in woodland or suburban areas during spring and summer.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000013_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "27137059-0058-48f2-ba11-ccbe90027201_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "SFX", "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 29.1166875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.7, \"end_time\": 28.12}, {\"start_time\": 5.45, \"end_time\": 26.49}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000014_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "27137059-0058-48f2-ba11-ccbe90027201_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "SFX", "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 29.1166875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.7, \"end_time\": 28.12}, {\"start_time\": 5.45, \"end_time\": 26.49}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000015_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "b3c54e3c-e013-4505-ac79-3536ff0d3126", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series", "SFX" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.6485625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). Global hint: segments must not overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.45}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.05, \"end_time\": 6.53, \"caption\": \"horn honking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 7.63, \"end_time\": 9.72, \"caption\": \"grunts\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.65, \"end_time\": 12.53, \"caption\": \"a(tm)aa(tm)aa(tm)a\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's happening at 18.61 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 15.62, \"end_time\": 21.65, \"caption\": \"The audio contains a prolonged, loud hissing sound, consistent with air escaping from an air brake system. The hiss is continuous and punctuated by occasional, softer puffs of air. The sound suggests a large vehicle, likely a truck, and the intensity indicates a significant air leak.\"}]}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.37, \"end_time\": 14.93, \"caption\": \"grunts\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 15.62, \"end_time\": 21.65, \"caption\": \"The audio contains a prolonged, loud hissing sound, consistent with air escaping from an air brake system. The hiss is continuous and punctuated by occasional, softer puffs of air. The sound suggests a large vehicle, likely a truck, and the intensity indicates a significant air leak.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000016_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "b3c54e3c-e013-4505-ac79-3536ff0d3126", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series", "SFX" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.6485625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). Global hint: segments must not overlap. First, provide the start and end times for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.45}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.05, \"end_time\": 6.53, \"caption\": \"the honking of a horn\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, here is the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 7.63, \"end_time\": 9.72, \"caption\": \"brief, low\\u2011pitched, guttural vocal bursts\\u00a0\\u00a0\\n\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, please provide the original text you'd like me to paraphrase.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.65, \"end_time\": 12.53, \"caption\": \"A series of rapid, repeated short sounds, each resembling a brief \\\"a\\\" or \\\"ah\\\" noise, occurring in quick succession with consistent timing.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What occurs at 18.61 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 15.62, \"end_time\": 21.65, \"caption\": \"The audio includes an extended, loud hiss that is typical of air escaping from an air\\u2011brake system. The hiss runs continuously, with occasional softer puffs of air interrupting it. The sound indicates a large vehicle\\u2014most likely a truck\\u2014and the steady intensity points to a substantial air leak.\"}]", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.37, \"end_time\": 14.93, \"caption\": \"low, guttural vocalizations, rhythmic and exertion-based\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 15.62, \"end_time\": 21.65, \"caption\": \"The audio features a sustained, high-volume hissing noise, characteristic of air discharging from an air brake system. The hiss remains constant, interrupted at intervals by fainter bursts of air. The sound implies a substantial vehicle, most probably a truck, with the loudness indicating a considerable air leak.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000017_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "b3c54e3c-e013-4505-ac79-3536ff0d3126_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series", "SFX" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.6485625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.45, \"caption\": \"knocklng\"}, {\"start_time\": 4.05, \"end_time\": 6.53, \"caption\": \"horn honking\"}, {\"start_time\": 7.63, \"end_time\": 9.72, \"caption\": \"grunts\"}, {\"start_time\": 10.65, \"end_time\": 12.53, \"caption\": \"a(tm)aa(tm)aa(tm)a\"}, {\"start_time\": 13.37, \"end_time\": 14.93, \"caption\": \"grunts\"}, {\"start_time\": 15.62, \"end_time\": 21.65, \"caption\": \"The audio contains a prolonged, loud hissing sound, consistent with air escaping from an air brake system. The hiss is continuous and punctuated by occasional, softer puffs of air. The sound suggests a large vehicle, likely a truck, and the intensity indicates a significant air leak.\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000018_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "b3c54e3c-e013-4505-ac79-3536ff0d3126_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series", "SFX" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.6485625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.45, \"caption\": \"knocking\"}, {\"start_time\": 4.05, \"end_time\": 6.53, \"caption\": \"A car horn emits a sharp, loud honking sound.\"}, {\"start_time\": 7.63, \"end_time\": 9.72, \"caption\": \"Brief, low\\u2011pitched grunting sounds.\"}, {\"start_time\": 10.65, \"end_time\": 12.53, \"caption\": \"a(tm)aa(tm)aa(tm)a\"}, {\"start_time\": 13.37, \"end_time\": 14.93, \"caption\": \"low, guttural exclamations emitted in brief bursts.\"}, {\"start_time\": 15.62, \"end_time\": 21.65, \"caption\": \"The audio features a sustained, high-volume hissing noise, characteristic of air releasing from an air brake system. The hiss remains constant, occasionally interrupted by fainter bursts of air. The sound implies a substantial vehicle, most likely a truck, with the volume and persistence indicating a major air leak.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000019_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "13ccdbb8-fa79-4a52-bbac-d31144ffc7f1", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 13.3779375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be ~5–15 seconds. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 12.38}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000020_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "13ccdbb8-fa79-4a52-bbac-d31144ffc7f1", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 13.3779375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: Segments should be around 5 to 15 seconds. First, provide me with the start and end of the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 12.38}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000021_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "13ccdbb8-fa79-4a52-bbac-d31144ffc7f1_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 13.3779375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 12.38}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000022_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "13ccdbb8-fa79-4a52-bbac-d31144ffc7f1_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 13.3779375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments along with their start times and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 12.38}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000023_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9c6ba2e8-a5a1-418d-8f1b-f865b1abc11b", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.469375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.28, \"caption\": \"audience laughing\"}, {\"start_time\": 3.81, \"end_time\": 5.96, \"caption\": \"thunder roars\"}, {\"start_time\": 6.92, \"end_time\": 10.89, \"caption\": \"low rumbling\"}, {\"start_time\": 12.66, \"end_time\": 14.67, \"caption\": \"door slides open\"}, {\"start_time\": 15.56, \"end_time\": 17.25, \"caption\": \"alejandro speaking\"}, {\"start_time\": 18.0, \"end_time\": 19.47, \"caption\": \"patty grunts\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000024_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9c6ba2e8-a5a1-418d-8f1b-f865b1abc11b", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.469375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments including start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.28, \"caption\": \"People in the audience laughing.\"}, {\"start_time\": 3.81, \"end_time\": 5.96, \"caption\": \"Thunder rumbles loudly.\"}, {\"start_time\": 6.92, \"end_time\": 10.89, \"caption\": \"Low, continuous rumbling sound\"}, {\"start_time\": 12.66, \"end_time\": 14.67, \"caption\": \"The door slides open, producing a smooth, sliding sound.\"}, {\"start_time\": 15.56, \"end_time\": 17.25, \"caption\": \"Alejandro's voice is heard as he speaks.\"}, {\"start_time\": 18.0, \"end_time\": 19.47, \"caption\": \"Patty emits a brief, low\\u2011frequency grunt, the sound pressed close to its source.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000025_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9c6ba2e8-a5a1-418d-8f1b-f865b1abc11b_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.469375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.28}, {\"start_time\": 3.81, \"end_time\": 5.96}, {\"start_time\": 6.92, \"end_time\": 10.89}, {\"start_time\": 12.66, \"end_time\": 14.67}, {\"start_time\": 15.56, \"end_time\": 17.25}, {\"start_time\": 18.0, \"end_time\": 19.47}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000026_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9c6ba2e8-a5a1-418d-8f1b-f865b1abc11b_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.469375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.28}, {\"start_time\": 3.81, \"end_time\": 5.96}, {\"start_time\": 6.92, \"end_time\": 10.89}, {\"start_time\": 12.66, \"end_time\": 14.67}, {\"start_time\": 15.56, \"end_time\": 17.25}, {\"start_time\": 18.0, \"end_time\": 19.47}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000027_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "fa7171fb-1156-4705-814d-714afa2e3fdf", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "SFX", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 24.9360625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.69}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, it must not overlap the previous one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.39, \"end_time\": 5.43, \"caption\": \"The audio contains a series of rapid, sharp sounds, resembling crashes, impacts, and explosions. There's also a sense of fire and rubble, suggesting a destructive event. The sounds are quick and intense, creating a chaotic and alarming atmosphere.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.89, \"end_time\": 8.4, \"caption\": \"light music\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment. Target ~9 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 9.53, \"end_time\": 12.45, \"caption\": \"gunshots\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Generate a standalone caption for the next segment. Target ~4 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"screams) (tar squelches\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Generate a standalone caption for the next segment. Target ~8 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"abby] what you doing that for? [knocking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the first segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.69, \"caption\": \"background chatter\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000028_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "fa7171fb-1156-4705-814d-714afa2e3fdf", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "SFX", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 24.9360625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, provide the start and end points for the initial segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.69}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following part, it must not overlap with the prior one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.39, \"end_time\": 5.43, \"caption\": \"The recording includes a series of rapid, sharp noises that resemble crashes, impacts, and explosions. It also conveys a sense of fire and rubble, hinting at a destructive occurrence. These noises are quick and intense, generating a chaotic and uneasy atmosphere.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end timestamps and a standalone caption for the following segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.89, \"end_time\": 8.4, \"caption\": \"gentle, airy music with soft, melodic tones.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Start: 00:12 \nEnd: 00:18 \nCaption: Welcome back", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 9.53, \"end_time\": 12.45, \"caption\": \"a rapid succession of sharp, explosive bangs\\u2014a loud, staccato burst of high\\u2011frequency energy punctuated by metallic overtones and brief, echoing reverberations.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Echoes of the Night", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"screams) (tar squelches\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Next segment: A new chapter begins.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"Abby] asking, \\\"What are you doing that for?\\\" and a knock is heard.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the initial part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.69, \"caption\": \"Ambient conversation described as a low-level murmur of overlapping voices\\u2014brief bursts of laughter, casual comments, and intermittent clinks of cups\\u2014all blending into a subtle, constant backdrop of everyday chatter.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000029_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "fa7171fb-1156-4705-814d-714afa2e3fdf_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "SFX", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 24.9360625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.69, \"caption\": \"background chatter\"}, {\"start_time\": 4.39, \"end_time\": 5.43, \"caption\": \"The audio contains a series of rapid, sharp sounds, resembling crashes, impacts, and explosions. There's also a sense of fire and rubble, suggesting a destructive event. The sounds are quick and intense, creating a chaotic and alarming atmosphere.\"}, {\"start_time\": 5.89, \"end_time\": 8.4, \"caption\": \"light music\"}, {\"start_time\": 9.53, \"end_time\": 12.45, \"caption\": \"gunshots\"}, {\"start_time\": 13.75, \"end_time\": 18.6, \"caption\": \"screams) (tar squelches\"}, {\"start_time\": 20.76, \"end_time\": 23.94, \"caption\": \"abby] what you doing that for? [knocking\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000030_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "fa7171fb-1156-4705-814d-714afa2e3fdf_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "SFX", "sdh_movies", "sdh_movies", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 24.9360625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Please provide the original text, and I will paraphrase it for you.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.69, \"caption\": \"background chatter\"}, {\"start_time\": 4.39, \"end_time\": 5.43, \"caption\": \"The audio features a rapid sequence of sharp, abrupt sounds akin to crashes, impacts, and explosions. A distinct impression of fire and collapsing debris conveys a destructive occurrence. The sounds are swift and intense, generating a chaotic and unsettling auditory environment.\"}, {\"start_time\": 5.89, \"end_time\": 8.4, \"caption\": \"Soft, gentle musical tones playing in the background.\"}, {\"start_time\": 9.53, \"end_time\": 12.45, \"caption\": \"Reports of gunfire, sharp and abrupt, with a sudden, explosive crack followed by a brief echo.\"}, {\"start_time\": 13.75, \"end_time\": 18.6, \"caption\": \"screams) (tar squelches\"}, {\"start_time\": 20.76, \"end_time\": 23.94, \"caption\": \"Abby: Why are you doing that? [knocking sound]\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000031_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "c56172ed-a27c-42b5-82c6-de039c3c102f", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 16.1174375, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 5.83}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.77, \"end_time\": 7.82, \"caption\": \"rhythmic drumming\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 6.69, \"end_time\": 7.78, \"caption\": \"I hear the sound of a coffee pot being inserted into a machine. It's a brief, distinct sound, likely involving a gentle thud or click as the pot is placed. The sound is short and sharp, indicating a quick action.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 7.38, \"end_time\": 15.12, \"caption\": \"I hear a soft, sustained synth pad sound. It's a gentle, atmospheric tone, likely with a slow attack and release. The sound is calm and peaceful, creating a spacious and ambient feeling. It seems to be the primary element in the audio, providing a foundational texture.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 12.25, \"end_time\": 13.64, \"caption\": \"cash register clicking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.12, \"end_time\": 14.34, \"caption\": \"grunts\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the middle segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 7.38, \"end_time\": 15.12, \"caption\": \"I hear a soft, sustained synth pad sound. It's a gentle, atmospheric tone, likely with a slow attack and release. The sound is calm and peaceful, creating a spacious and ambient feeling. It seems to be the primary element in the audio, providing a foundational texture.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000032_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "c56172ed-a27c-42b5-82c6-de039c3c102f", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 16.1174375, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be brief, lasting 1 to 8 seconds. First, provide the start and end times for the initial segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 5.83}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, please provide the original text you'd like me to paraphrase.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.77, \"end_time\": 7.82, \"caption\": \"regular drumbeat\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Please provide the original text, and I will paraphrase it while preserving the exact meaning and structure.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 6.69, \"end_time\": 7.78, \"caption\": \"I hear the sound of a coffee pot being inserted into a machine. It's a brief, distinct sound, likely involving a gentle thud or click as the pot is placed. The sound is short and sharp, indicating a quick action.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 7.38, \"end_time\": 15.12, \"caption\": \"It\\u2019s a softly sustained synth pad, a gentle, atmospheric tone that probably develops with a slow attack and release. The sound feels calm and peaceful, evoking a wide, ambient space. It appears to be the main element, providing the foundational texture of the audio.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, please provide the original text you'd like me to paraphrase.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 12.25, \"end_time\": 13.64, \"caption\": \"clicking of a cash register\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the subsequent segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.12, \"end_time\": 14.34, \"caption\": \"short, guttural grunts.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the middle portion?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 7.38, \"end_time\": 15.12, \"caption\": \"I detect a soft, continuous synth pad. It's a calm, atmospheric tone, probably featuring a slow attack and release. The sound is tranquil and serene, evoking a wide, immersive ambiance. It appears to be the main component of the audio, establishing a base layer of texture.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000033_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "c56172ed-a27c-42b5-82c6-de039c3c102f_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 16.1174375, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.95, \"end_time\": 5.83, \"caption\": \"ominous music\"}, {\"start_time\": 4.77, \"end_time\": 7.82, \"caption\": \"rhythmic drumming\"}, {\"start_time\": 6.69, \"end_time\": 7.78, \"caption\": \"I hear the sound of a coffee pot being inserted into a machine. It's a brief, distinct sound, likely involving a gentle thud or click as the pot is placed. The sound is short and sharp, indicating a quick action.\"}, {\"start_time\": 7.38, \"end_time\": 15.12, \"caption\": \"I hear a soft, sustained synth pad sound. It's a gentle, atmospheric tone, likely with a slow attack and release. The sound is calm and peaceful, creating a spacious and ambient feeling. It seems to be the primary element in the audio, providing a foundational texture.\"}, {\"start_time\": 12.25, \"end_time\": 13.64, \"caption\": \"cash register clicking\"}, {\"start_time\": 13.12, \"end_time\": 14.34, \"caption\": \"grunts\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000034_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "c56172ed-a27c-42b5-82c6-de039c3c102f_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 16.1174375, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.95, \"end_time\": 5.83, \"caption\": \"ominous music\"}, {\"start_time\": 4.77, \"end_time\": 7.82, \"caption\": \"steady, repeating drum beats\"}, {\"start_time\": 6.69, \"end_time\": 7.78, \"caption\": \"I hear a coffee pot being set into a machine, a clear, brief thump or click as it lands. The tone is short and sharp, reflecting a quick insertion.\"}, {\"start_time\": 7.38, \"end_time\": 15.12, \"caption\": \"I detect a soft, continuous synth pad. It's a mild, ambient tone, probably featuring a slow attack and release. The sound is tranquil and soothing, evoking a wide, open auditory space. It appears to be the main component of the audio, establishing a core textural layer.\"}, {\"start_time\": 12.25, \"end_time\": 13.64, \"caption\": \"cash register click sound\"}, {\"start_time\": 13.12, \"end_time\": 14.34, \"caption\": \"deep, guttural vocalizations made with effort or exertion\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000035_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "5c5f83a7-7e10-4d1d-8020-cccc28e7f73e", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 14.3435625, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 4.59}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's happening at 4.65 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.99, \"end_time\": 5.95, \"caption\": \"screaming\"}]}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the start and end time for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.99, \"end_time\": 5.95}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"screaming\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, it must not overlap the previous one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.95, \"end_time\": 12.81, \"caption\": \"The audio is a long, sustained, low-frequency drone with subtle, evolving textures. It sounds like a deep, resonating hum, possibly with a slight metallic or distorted quality. There are very few discernible melodic elements, and the overall effect is atmospheric and somewhat unsettling.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.27, \"end_time\": 12.91, \"caption\": \"papers rustling\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Generate a standalone caption for the next segment. Target ~1 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"crack\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the middle segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.95, \"end_time\": 12.81, \"caption\": \"The audio is a long, sustained, low-frequency drone with subtle, evolving textures. It sounds like a deep, resonating hum, possibly with a slight metallic or distorted quality. There are very few discernible melodic elements, and the overall effect is atmospheric and somewhat unsettling.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000036_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "5c5f83a7-7e10-4d1d-8020-cccc28e7f73e", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 14.3435625, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, provide the start and end points for the initial segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 4.59}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What occurs at 4.65 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.99, \"end_time\": 5.95, \"caption\": \"A loud, high\\u2011pitched, rapid vocalization that echoes with intense force.\"}]", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the start and end time for the upcoming segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.99, \"end_time\": 5.95}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the caption for that specific part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A high\\u2011pitched, piercing vocalization that sharply rises and falls, conveying intense fear or alarm.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following segment, it must not overlap the prior one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.95, \"end_time\": 12.81, \"caption\": \"A prolonged, low-pitched drone with faint, gradually shifting textures. It emits a deep, vibrating hum, potentially tinged with a metallic or slightly warped character. Minimal melodic features are present, creating an immersive, eerie atmosphere.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end timestamps and a standalone caption for the following segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.27, \"end_time\": 12.91, \"caption\": \"the soft rustling sound of papers.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Next", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"a sharp, brief burst of sound, like a sudden snap or break, producing a small but distinct crack.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the central part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.95, \"end_time\": 12.81, \"caption\": \"A prolonged, low-pitched drone with delicate, shifting textures. It emits a deep, vibrating hum, potentially carrying a faint metallic or slightly warped character. Minimal melodic features are present, creating an immersive, eerie atmosphere.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000037_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "5c5f83a7-7e10-4d1d-8020-cccc28e7f73e_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 14.3435625, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.95, \"end_time\": 4.59}, {\"start_time\": 3.99, \"end_time\": 5.95}, {\"start_time\": 5.22, \"end_time\": 6.38}, {\"start_time\": 5.95, \"end_time\": 12.81}, {\"start_time\": 10.27, \"end_time\": 12.91}, {\"start_time\": 11.94, \"end_time\": 13.34}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000038_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "5c5f83a7-7e10-4d1d-8020-cccc28e7f73e_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "SFX", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 14.3435625, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.95, \"end_time\": 4.59}, {\"start_time\": 3.99, \"end_time\": 5.95}, {\"start_time\": 5.22, \"end_time\": 6.38}, {\"start_time\": 5.95, \"end_time\": 12.81}, {\"start_time\": 10.27, \"end_time\": 12.91}, {\"start_time\": 11.94, \"end_time\": 13.34}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000039_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "71f48c07-899b-43dc-80f4-742c5140b005", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 28.2198125, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be ~5–15 seconds. Global hint: segments may overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.7, \"end_time\": 17.43}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"Tufted Titmouse\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the start and end time for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 16.47, \"end_time\": 27.22}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"Northern Parula\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000040_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "71f48c07-899b-43dc-80f4-742c5140b005", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 28.2198125, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: each segment should be approximately 5 to 15 seconds long. Global hint: segments can overlap. First, provide the start and end times for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.7, \"end_time\": 17.43}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the caption for that specific part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A small, high-pitched, whistled call, often described as \\\"peter-peter-peter,\\\" with a clear, slightly nasal tone, repeated in a rhythmic pattern, typically heard in wooded areas during spring and summer.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What are the start and end times for the upcoming segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 16.47, \"end_time\": 27.22}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the caption for that specific part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A high-pitched, rapid, and slightly buzzy bird call, beginning with a series of clear, thin notes that accelerate into a trill, ending abruptly.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000041_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "71f48c07-899b-43dc-80f4-742c5140b005_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 28.2198125, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.7, \"end_time\": 17.43, \"caption\": \"Tufted Titmouse\"}, {\"start_time\": 16.47, \"end_time\": 27.22, \"caption\": \"Northern Parula\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000042_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "71f48c07-899b-43dc-80f4-742c5140b005_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 28.2198125, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Please provide the original text, and I will paraphrase it for you.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.7, \"end_time\": 17.43, \"caption\": \"Tufted Titmouse\"}, {\"start_time\": 16.47, \"end_time\": 27.22, \"caption\": \"The Northern Parula\\u2019s call is a bright, high\\u2011pitch trill interspersed with a series of quick, crystal\\u2011clear metallic notes.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000043_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "546a8a3d-dbe6-4263-9067-7e72270295aa", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.86075, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 13.86, \"caption\": \"Cope's Gray Treefrog\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000044_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "546a8a3d-dbe6-4263-9067-7e72270295aa", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.86075, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "I’m sorry, but I can’t help with that.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 13.86, \"caption\": \"A high-pitched, repetitive trill, resembling a rapid series of short, clear notes, produced by a small amphibian in a tree, with a slightly metallic or ringing quality, lasting several seconds and repeating intermittently.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000045_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "546a8a3d-dbe6-4263-9067-7e72270295aa_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.86075, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 13.86, \"caption\": \"Cope's Gray Treefrog\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000046_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "546a8a3d-dbe6-4263-9067-7e72270295aa_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.86075, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide every segment along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 13.86, \"caption\": \"the soft, resonant call typical of a Cope's Gray Treefrog\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000047_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6085b951-d643-4c14-bab7-3e66c4259f94", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_movies", "SFX", "sdh_movies", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.9305, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). Global hint: segments must not overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.56}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next caption, aim for ~204 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.21, \"end_time\": 9.54, \"caption\": \"The audio contains a repetitive, distorted, and scratched vinyl sound. It's a looping sample, likely used for texture or a rhythmic element in a hip-hop track. The sound is noisy and lacks clear melodic content.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next caption, aim for ~1 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 11.92, \"end_time\": 13.16, \"caption\": \"exhales deeply\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment. Target ~8 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.71, \"end_time\": 14.93, \"caption\": \"door opens\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, it must not overlap the previous one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 15.47, \"end_time\": 19.05, \"caption\": \"spitfire engine rumbling\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 20.65, \"end_time\": 21.93, \"caption\": \"screaming\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000048_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6085b951-d643-4c14-bab7-3e66c4259f94", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_movies", "SFX", "sdh_movies", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.9305, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). Global hint: segments must not overlap. First, provide the start and end times for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.56}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following caption, strive for approximately 204 characters.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.21, \"end_time\": 9.54, \"caption\": \"The audio features a recurring, warped, and grainy vinyl noise. It's a repeated sample, probably employed for atmospheric texture or rhythmic pacing in a hip-hop composition. The sound is characterized by high noise levels and no discernible melody.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Short and sweet.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 11.92, \"end_time\": 13.16, \"caption\": \"takes a long, deep breath out\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Start: 00:12:34 \nEnd: 00:12:42 \nCaption: Welcome back", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.71, \"end_time\": 14.93, \"caption\": \"A door creaks open slowly, producing a low, sustained metallic scraping sound followed by a soft, abrupt click as it fully unlatches.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following segment, it must not overlap the prior one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 15.47, \"end_time\": 19.05, \"caption\": \"The rumbling sound of a Spitfire engine.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end timestamps and a standalone caption for the following segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 20.65, \"end_time\": 21.93, \"caption\": \"screaming\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000049_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6085b951-d643-4c14-bab7-3e66c4259f94_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_movies", "SFX", "sdh_movies", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.9305, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.56, \"caption\": \"sighs\"}, {\"start_time\": 4.21, \"end_time\": 9.54, \"caption\": \"The audio contains a repetitive, distorted, and scratched vinyl sound. It's a looping sample, likely used for texture or a rhythmic element in a hip-hop track. The sound is noisy and lacks clear melodic content.\"}, {\"start_time\": 11.92, \"end_time\": 13.16, \"caption\": \"exhales deeply\"}, {\"start_time\": 13.71, \"end_time\": 14.93, \"caption\": \"door opens\"}, {\"start_time\": 15.47, \"end_time\": 19.05, \"caption\": \"spitfire engine rumbling\"}, {\"start_time\": 20.65, \"end_time\": 21.93, \"caption\": \"screaming\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000050_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "6085b951-d643-4c14-bab7-3e66c4259f94_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_movies", "SFX", "sdh_movies", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 22.9305, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide every segment along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.56, \"caption\": \"breath escapes gently\"}, {\"start_time\": 4.21, \"end_time\": 9.54, \"caption\": \"The recording features a repetitive, distorted, and scratched vinyl noise. It's a looping sample, likely used for texture or rhythmic support in a hip\\u2011hop track, and the sound is noisy with no distinct melodic content.\"}, {\"start_time\": 11.92, \"end_time\": 13.16, \"caption\": \"takes a long, deep breath out\"}, {\"start_time\": 13.71, \"end_time\": 14.93, \"caption\": \"The sound of a door opening.\"}, {\"start_time\": 15.47, \"end_time\": 19.05, \"caption\": \"The rumbling sound of a Spitfire engine\"}, {\"start_time\": 20.65, \"end_time\": 21.93, \"caption\": \"an intense, high\\u2011frequency vocal outburst with piercing, shrill tones.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000051_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "275981df-3157-4b12-bb4f-52a169426dd3", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "sdh_series", "sdh_series", "sdh_movies", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 26.0588125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 4.69, \"caption\": \"brain groans\"}, {\"start_time\": 5.84, \"end_time\": 8.66, \"caption\": \"laughing\"}, {\"start_time\": 9.91, \"end_time\": 12.24, \"caption\": \"music playing\"}, {\"start_time\": 13.28, \"end_time\": 16.09, \"caption\": \"I hear the distinct sounds of glass breaking and falling, likely from a cardboard box. There are multiple impacts, creating a scattering of sharp, brittle sounds. The audio suggests a deliberate action of pouring or dropping glass debris, resulting in a cascade of shattering pieces.\"}, {\"start_time\": 17.35, \"end_time\": 20.76, \"caption\": \"car door opens, closes\"}, {\"start_time\": 22.29, \"end_time\": 25.06, \"caption\": \"gojodoeneun eumag\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000052_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "275981df-3157-4b12-bb4f-52a169426dd3", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "sdh_series", "sdh_series", "sdh_movies", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 26.0588125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments including start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 4.69, \"caption\": \"A low, deep groan emanates from the brain, as if the mind itself is straining or creaking under pressure.\"}, {\"start_time\": 5.84, \"end_time\": 8.66, \"caption\": \"laughing\"}, {\"start_time\": 9.91, \"end_time\": 12.24, \"caption\": \"A musical piece is being played.\"}, {\"start_time\": 13.28, \"end_time\": 16.09, \"caption\": \"I hear the clear, sharp sounds of glass breaking and falling, probably originating from a cardboard box. Multiple impacts produce a series of brittle, cracking noises. The audio indicates a deliberate act of pouring or dropping glass fragments, leading to a rapid sequence of shattering pieces.\"}, {\"start_time\": 17.35, \"end_time\": 20.76, \"caption\": \"The vehicle\\u2019s door swings open with a gentle hiss, then shuts with a crisp click.\"}, {\"start_time\": 22.29, \"end_time\": 25.06, \"caption\": \"The sound of a door creaking open slowly.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000053_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "275981df-3157-4b12-bb4f-52a169426dd3_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "sdh_movies", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 26.0588125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 4.69}, {\"start_time\": 5.84, \"end_time\": 8.66}, {\"start_time\": 9.91, \"end_time\": 12.24}, {\"start_time\": 13.28, \"end_time\": 16.09}, {\"start_time\": 17.35, \"end_time\": 20.76}, {\"start_time\": 22.29, \"end_time\": 25.06}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000054_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "275981df-3157-4b12-bb4f-52a169426dd3_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "sdh_movies", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 26.0588125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 4.69}, {\"start_time\": 5.84, \"end_time\": 8.66}, {\"start_time\": 9.91, \"end_time\": 12.24}, {\"start_time\": 13.28, \"end_time\": 16.09}, {\"start_time\": 17.35, \"end_time\": 20.76}, {\"start_time\": 22.29, \"end_time\": 25.06}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000055_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9cb9e2a4-054e-4b45-8270-7ddc3ac558c6", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 25.4349375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments must not overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.39}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's happening at 2.37 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.39, \"caption\": \"stammers\"}]}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, aim for ~2 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.96, \"end_time\": 7.18, \"caption\": \"audience laughs\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment. Target ~1 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.62, \"end_time\": 10.13, \"caption\": \"prisoners grunting\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, it must not overlap the previous one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.81, \"end_time\": 15.61, \"caption\": \"The audio is a guitar playing a distorted, repetitive riff. The riff consists of a series of rapid, plucked notes with a heavy distortion effect. The sound is consistent and somewhat monotonous, creating a drone-like effect. There are no discernible changes in melody or rhythm throughout the duration.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, aim for ~4 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 17.75, \"end_time\": 20.59, \"caption\": \"squawks\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment. Target ~4 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 21.85, \"end_time\": 24.43, \"caption\": \"energetic rock music\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the first segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.39, \"caption\": \"stammers\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000056_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9cb9e2a4-054e-4b45-8270-7ddc3ac558c6", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 25.4349375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments must not overlap. First, provide the start and end points for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.39}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's occurring at 2.37 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.39, \"caption\": \"stutters\"}]", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following part, target approximately 2 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.96, \"end_time\": 7.18, \"caption\": \"Audience chuckles and laughs.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Start: [Next segment begins] \nEnd: [End of segment] \nCaption: Transition", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.62, \"end_time\": 10.13, \"caption\": \"Prisoners grunting\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following segment, it must not overlap the prior one.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.81, \"end_time\": 15.61, \"caption\": \"A guitar produces a distorted, repeating riff characterized by fast, plucked notes enhanced with strong distortion. The sound remains steady and unvarying, generating a sustained, drone-like quality. There are no noticeable shifts in melody or rhythm over time.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following segment, target around 4 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 17.75, \"end_time\": 20.59, \"caption\": \"Sharp, high-pitched bird calls, repeated rapidly.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Start: [4-word caption] \nEnd: [4-word caption]", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 21.85, \"end_time\": 24.43, \"caption\": \"Upbeat rock music with dynamic rhythms and powerful guitar riffs.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the initial part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.39, \"caption\": \"The speaker hesitates, producing brief, repeated, and irregular vocal disruptions in speech, indicating a struggle with fluency.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000057_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9cb9e2a4-054e-4b45-8270-7ddc3ac558c6_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 25.4349375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.39}, {\"start_time\": 3.96, \"end_time\": 7.18}, {\"start_time\": 8.62, \"end_time\": 10.13}, {\"start_time\": 10.81, \"end_time\": 15.61}, {\"start_time\": 17.75, \"end_time\": 20.59}, {\"start_time\": 21.85, \"end_time\": 24.43}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000058_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "9cb9e2a4-054e-4b45-8270-7ddc3ac558c6_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "sdh_series", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 25.4349375, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.39}, {\"start_time\": 3.96, \"end_time\": 7.18}, {\"start_time\": 8.62, \"end_time\": 10.13}, {\"start_time\": 10.81, \"end_time\": 15.61}, {\"start_time\": 17.75, \"end_time\": 20.59}, {\"start_time\": 21.85, \"end_time\": 24.43}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000059_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "feb4a69a-27ad-4647-94da-d9b1450e3ffa", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 18.26625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.33}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"bed squeaking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the start and end time for the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.87, \"end_time\": 5.25}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that exact segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"tar man snarls\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me start/end for the next segment. Aim for ~4 seconds segments overall.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.87, \"end_time\": 9.22}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me start/end for the next segment. Aim for ~2 seconds segments overall.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.72, \"end_time\": 12.1}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide start/end and a standalone caption for the next segment. Target ~2 words.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 12.72, \"end_time\": 14.3, \"caption\": \"sniffles\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the next segment, aim for ~1 seconds.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 15.01, \"end_time\": 17.27, \"caption\": \"eunyuyi himjuneun sumsori\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the first segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.33, \"caption\": \"bed squeaking\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000060_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "feb4a69a-27ad-4647-94da-d9b1450e3ffa", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 18.26625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_plus_local", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: Segments should be relatively short, ranging from 1 to 8 seconds. First, please provide the start and end time for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.33}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the caption for that specific segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A creaking sound emanating from the bed, consistent and intermittent, likely due to movement or pressure on the frame or mattress.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the start and end time for the upcoming segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.87, \"end_time\": 5.25}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Provide the caption for that specific part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"caption\": \"A tar man is producing a snarling noise.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the start and end times for the next segment. Target around 4 seconds total.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.87, \"end_time\": 9.22}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the start and end times for the next segment. Target around 2 seconds total.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.72, \"end_time\": 12.1}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "<START>\n<END>\nNext Segment", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 12.72, \"end_time\": 14.3, \"caption\": \"soft, quick whisk of the nose in a brief, gentle sniff.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "For the following part, target approximately 1 second.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 15.01, \"end_time\": 17.27, \"caption\": \"a delicate, resonant hum\\n\\n\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the initial part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 3.33, \"caption\": \"A bed creaking under weight, with rhythmic, high-pitched squeaks accompanied by subtle groans from the frame.\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000061_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "feb4a69a-27ad-4647-94da-d9b1450e3ffa_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 18.26625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 3.33, \"caption\": \"bed squeaking\"}, {\"start_time\": 3.87, \"end_time\": 5.25, \"caption\": \"tar man snarls\"}, {\"start_time\": 5.87, \"end_time\": 9.22, \"caption\": \"inmates applauding and cheering\"}, {\"start_time\": 10.72, \"end_time\": 12.1, \"caption\": \"snarling\"}, {\"start_time\": 12.72, \"end_time\": 14.3, \"caption\": \"sniffles\"}, {\"start_time\": 15.01, \"end_time\": 17.27, \"caption\": \"eunyuyi himjuneun sumsori\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000062_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "feb4a69a-27ad-4647-94da-d9b1450e3ffa_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_movies", "sdh_series", "sdh_series", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 18.26625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 3.33, \"caption\": \"A bed creaking under weight, with intermittent, high-pitched squeaks as the frame shifts.\"}, {\"start_time\": 3.87, \"end_time\": 5.25, \"caption\": \"A tar man growls with a harsh, guttural snarl.\"}, {\"start_time\": 5.87, \"end_time\": 9.22, \"caption\": \"inmates clapping and shouting in approval\"}, {\"start_time\": 10.72, \"end_time\": 12.1, \"caption\": \"A low, guttural growl accompanied by sharp, aggressive canine teeth clacking, indicating intense hostility or warning.\"}, {\"start_time\": 12.72, \"end_time\": 14.3, \"caption\": \"A series of short, intermittent inhalations through the nose, often accompanied by a slight, soft rushing or bubbling sound, typically indicating nasal congestion or a mild respiratory irritation.\"}, {\"start_time\": 15.01, \"end_time\": 17.27, \"caption\": \"eunyuyi himjuneun sumsori\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000063_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a4f7c8cc-1076-46af-913a-8ba085530f65", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "AnimalSpeak", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 13.47975, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 4.0}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.61, \"end_time\": 5.72, \"caption\": \"cheers and applause\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.94, \"end_time\": 7.01, \"caption\": \"all laugh\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 6.24, \"end_time\": 9.24, \"caption\": \"I hear a male voice reading a story. He's speaking clearly and at a moderate pace. The audio appears to be an excerpt from a narrative, likely a dramatic or emotional one, centered around a character motivated by love to protect someone.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.13, \"end_time\": 12.48, \"caption\": \"Song Thrush\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.87, \"end_time\": 12.19, \"caption\": \"camgajadeuli nolranda\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the first segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 4.0, \"caption\": \"laughing\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000064_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a4f7c8cc-1076-46af-913a-8ba085530f65", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "AnimalSpeak", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 13.47975, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: Segments should be relatively short, ranging from 1 to 8 seconds. First, please provide the start and end time for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 4.0}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 3.61, \"end_time\": 5.72, \"caption\": \"an exuberant wave of clapping accompanied by vocal cheers.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, please provide the original text you'd like me to paraphrase.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 4.94, \"end_time\": 7.01, \"caption\": \"all laughter\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Please provide the original text, and I will paraphrase it for you.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 6.24, \"end_time\": 9.24, \"caption\": \"I hear a male speaker delivering a story with clear diction at a moderate speed. The recording sounds like an excerpt from a dramatic or emotionally intense narrative, featuring a character who is motivated by love to protect someone.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.13, \"end_time\": 12.48, \"caption\": \"A Song Thrush sings, its melodic notes cascading in clear, repeated phrases, each phrase slightly varied, creating a rich, rhythmic, and expressive birdcall.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.87, \"end_time\": 12.19, \"caption\": \"camgajadeuli nolranda\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the initial part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.95, \"end_time\": 4.0, \"caption\": \"a series of quick, rhythmic chuckles\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000065_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a4f7c8cc-1076-46af-913a-8ba085530f65_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "AnimalSpeak", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 13.47975, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.95, \"end_time\": 4.0}, {\"start_time\": 3.61, \"end_time\": 5.72}, {\"start_time\": 4.94, \"end_time\": 7.01}, {\"start_time\": 6.24, \"end_time\": 9.24}, {\"start_time\": 8.13, \"end_time\": 12.48}, {\"start_time\": 10.87, \"end_time\": 12.19}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000066_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a4f7c8cc-1076-46af-913a-8ba085530f65_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "sdh_series", "sdh_series", "SFX", "AnimalSpeak", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 13.47975, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.95, \"end_time\": 4.0}, {\"start_time\": 3.61, \"end_time\": 5.72}, {\"start_time\": 4.94, \"end_time\": 7.01}, {\"start_time\": 6.24, \"end_time\": 9.24}, {\"start_time\": 8.13, \"end_time\": 12.48}, {\"start_time\": 10.87, \"end_time\": 12.19}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000067_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "48174c3f-4e08-410d-9afb-2d39862a0edf", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 26.49575, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 25.5, \"caption\": \"Eastern Bluebird\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000068_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "48174c3f-4e08-410d-9afb-2d39862a0edf", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 26.49575, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments including start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 25.5, \"caption\": \"Eastern Bluebirds emit a crisp, bright, staccato song of quick, tonal notes that roll together in a short, continuous trill.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000069_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "48174c3f-4e08-410d-9afb-2d39862a0edf_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 26.49575, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 25.5}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000070_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "48174c3f-4e08-410d-9afb-2d39862a0edf_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 26.49575, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 25.5}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000071_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "099674b7-9778-4be9-ab25-9b8cb1e3bf7b", "meta": { "conversation_type": "full_list_no_captions", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 20.8760625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.7, \"end_time\": 19.88}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000072_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "099674b7-9778-4be9-ab25-9b8cb1e3bf7b", "meta": { "conversation_type": "full_list_no_captions", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 20.8760625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.7, \"end_time\": 19.88}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000073_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "099674b7-9778-4be9-ab25-9b8cb1e3bf7b_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 20.8760625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.7, \"end_time\": 19.88, \"caption\": \"Setophaga coronata auduboni\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000074_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "099674b7-9778-4be9-ab25-9b8cb1e3bf7b_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "long", "duration_hint": "segments should be ~15–40 seconds", "duration_s": 20.8760625, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.7, \"end_time\": 19.88, \"caption\": \"A high-pitched, thin, and slightly buzzy bird call, resembling a sharp \\\"see-see-see\\\" or \\\"tsee-tsee-tsee,\\\" with rapid, evenly spaced notes that are slightly descending in pitch, typical of the Audubon's warbler.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000075_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "8279a97f-053b-4e0d-98ba-ecc6fddd616c", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.3311875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 4.35}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.35, \"end_time\": 7.8, \"caption\": \"inmates laughing\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.9, \"end_time\": 9.9, \"caption\": \"chuckles\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What's happening at 12.9 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": []}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.35, \"end_time\": 12.31, \"caption\": \"camera whirring, clicking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.19, \"end_time\": 16.11, \"caption\": \"papers crinkling\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 17.41, \"end_time\": 19.33, \"caption\": \"brakes screech\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the last segment?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 17.41, \"end_time\": 19.33, \"caption\": \"brakes screech\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000076_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "8279a97f-053b-4e0d-98ba-ecc6fddd616c", "meta": { "conversation_type": "iterative", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.3311875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "global_only", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments should be relatively short (1–8 seconds). First, provide the start and end times for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 2.11, \"end_time\": 4.35}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, here is the next segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 5.35, \"end_time\": 7.8, \"caption\": \"Inmates laughing loudly, their voices overlapping in a chaotic, echoing burst of laughter.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 8.9, \"end_time\": 9.9, \"caption\": \"Soft, low laughter, barely above a whisper, with a gentle, rhythmic quality.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What occurs at 12.9 seconds?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[]", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Sure, please provide the original text you'd like me to paraphrase.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 10.35, \"end_time\": 12.31, \"caption\": \"camera whirring, shutter clicking\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 13.19, \"end_time\": 16.11, \"caption\": \"The sound of paper rustling or folding, with a crisp, dry crackling noise as sheets move against each other.\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "Give me the following part.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 17.41, \"end_time\": 19.33, \"caption\": \"brake screeches\"}", "type": "text" } ], "role": "assistant" }, { "content": [ { "text": "What was the final part?", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 17.41, \"end_time\": 19.33, \"caption\": \"Squealing brakes\"}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000077_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "8279a97f-053b-4e0d-98ba-ecc6fddd616c_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.3311875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.11, \"end_time\": 4.35}, {\"start_time\": 5.35, \"end_time\": 7.8}, {\"start_time\": 8.9, \"end_time\": 9.9}, {\"start_time\": 10.35, \"end_time\": 12.31}, {\"start_time\": 13.19, \"end_time\": 16.11}, {\"start_time\": 17.41, \"end_time\": 19.33}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000078_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "8279a97f-053b-4e0d-98ba-ecc6fddd616c_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_movies", "sdh_series", "sdh_movies", "sdh_series", "sdh_movies", "sdh_movies" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 20.3311875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.11, \"end_time\": 4.35}, {\"start_time\": 5.35, \"end_time\": 7.8}, {\"start_time\": 8.9, \"end_time\": 9.9}, {\"start_time\": 10.35, \"end_time\": 12.31}, {\"start_time\": 13.19, \"end_time\": 16.11}, {\"start_time\": 17.41, \"end_time\": 19.33}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000079_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "7a578e7b-1c65-4f16-89d7-578837fe9e7c", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.8805, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments must not overlap. First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 13.88}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000080_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "7a578e7b-1c65-4f16-89d7-578837fe9e7c", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.8805, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": true }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Global hint: segments must not overlap. First, provide the start and end points for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 13.88}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000081_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "7a578e7b-1c65-4f16-89d7-578837fe9e7c_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.8805, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 13.88}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000082_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "7a578e7b-1c65-4f16-89d7-578837fe9e7c_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.8805, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments along with their start and end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 13.88}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000083_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "0ff519d2-8941-43f1-9ba2-2db4f085cac5", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 32.04275, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.7, \"end_time\": 17.65, \"caption\": \"Common Reed Warbler\"}, {\"start_time\": 16.68, \"end_time\": 31.04, \"caption\": \"Call of a blue-throated barbet\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000084_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "0ff519d2-8941-43f1-9ba2-2db4f085cac5", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 32.04275, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide me with all segments along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.7, \"end_time\": 17.65, \"caption\": \"A high-pitched, melodious warbling call, characteristic of the Common Reed Warbler, with rapid, trilling notes and a slightly buzzy quality, often heard in reed beds and wetland habitats.\"}, {\"start_time\": 16.68, \"end_time\": 31.04, \"caption\": \"The booming, resonant croak of the blue\\u2011throated barbet echoes through the canopy, punctuated by sharp, staccato notes that ripple like a gravelly whistle under a steady pulse of deep, gravel\\u2011laden tones.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000085_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "0ff519d2-8941-43f1-9ba2-2db4f085cac5_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 32.04275, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 3.7, \"end_time\": 17.65, \"caption\": \"Common Reed Warbler\"}, {\"start_time\": 16.68, \"end_time\": 31.04, \"caption\": \"Call of a blue-throated barbet\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000086_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "0ff519d2-8941-43f1-9ba2-2db4f085cac5_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak", "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 32.04275, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 2, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide every segment along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 3.7, \"end_time\": 17.65, \"caption\": \"A birdcall consisting of a rapid, melodic series of short, high-pitched notes, slightly varied in pitch and rhythm, characteristic of the Common Reed Warbler's song.\"}, {\"start_time\": 16.68, \"end_time\": 31.04, \"caption\": \"A high-pitched, sharp call of a blue-throated barbet, characterized by a series of rapid, staccato notes with a slightly descending inflection, resembling a repeated \\\"kree-kree-kree\\\" or \\\"chik-chik-chik\\\" sound, often delivered in short bursts with a clear, ringing quality.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000087_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "1d6b3d5f-b638-436f-9636-41fa40c23fbc", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "macs" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 11.5125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 10.51, \"caption\": \"a car is moving when it raining\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000088_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "1d6b3d5f-b638-436f-9636-41fa40c23fbc", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "macs" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 11.5125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Please provide the original text, and I will paraphrase it for you.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 10.51, \"caption\": \"A car is moving while rain is falling.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000089_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "1d6b3d5f-b638-436f-9636-41fa40c23fbc_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "macs" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 11.5125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 10.51, \"caption\": \"a car is moving when it raining\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000090_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "1d6b3d5f-b638-436f-9636-41fa40c23fbc_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "macs" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 11.5125, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "I’m sorry, but I can’t help with that.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 0.51, \"end_time\": 10.51, \"caption\": \"A car is moving through rain, with the sound of tires rolling on wet pavement and the steady patter of raindrops hitting the vehicle's surface.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000091_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "88d669bb-4860-4395-aca6-39e0c04ce2a5", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "sdh_series", "AudioData", "AudioData", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 17.8921875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.95, \"end_time\": 5.13, \"caption\": \"sighing\"}, {\"start_time\": 4.32, \"end_time\": 10.45, \"caption\": \"A wine glass is struck once with a soft mallet, producing a clear, soft chime.\"}, {\"start_time\": 8.18, \"end_time\": 9.48, \"caption\": \"The audio contains the distinct sound of a handgun being fired. It's a close perspective, likely capturing the sharp crack of a single shot from a Thompson/Center Contender .22 LR pistol. The sound is crisp and clear, indicative of a single shot.\"}, {\"start_time\": 9.0, \"end_time\": 16.41, \"caption\": \"The audio contains the sound of machine gun bursts and bullet ricochets off dirt. There are multiple instances of bullets impacting and bouncing off a dirt surface, creating a dynamic and somewhat chaotic soundscape. The overall impression is one of intense gunfire and the immediate aftermath of bullets striking a ground.\"}, {\"start_time\": 13.66, \"end_time\": 16.01, \"caption\": \"sighs\"}, {\"start_time\": 15.14, \"end_time\": 16.89, \"caption\": \"rim shot\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000092_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "88d669bb-4860-4395-aca6-39e0c04ce2a5", "meta": { "conversation_type": "full_list_with_captions", "datasets_used": [ "sdh_series", "AudioData", "AudioData", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 17.8921875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide all segments including start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.95, \"end_time\": 5.13, \"caption\": \"A soft, drawn\\u2011out exhalation that usually conveys frustration, disappointment, or weariness.\"}, {\"start_time\": 4.32, \"end_time\": 10.45, \"caption\": \"A wine glass is struck once with a gentle mallet, producing a bright, soft ringing sound.\"}, {\"start_time\": 8.18, \"end_time\": 9.48, \"caption\": \"The audio captures the sharp, distinct crack of a single handgun discharge, recorded from a close proximity. The sound is crisp and clear, consistent with a single shot fired from a Thompson/Center Contender .22 LR pistol.\"}, {\"start_time\": 9.0, \"end_time\": 16.41, \"caption\": \"The audio features the sounds of machine gun bursts and bullets ricocheting off dirt, with multiple instances of bullets impacting and bouncing off a dirt surface, creating a dynamic and somewhat chaotic soundscape, giving an overall impression of intense gunfire and the immediate aftermath of bullets striking the ground.\"}, {\"start_time\": 13.66, \"end_time\": 16.01, \"caption\": \"A long, audible sigh\\u2014a slow exhalation of breath that resonates with a low-frequency tone.\"}, {\"start_time\": 15.14, \"end_time\": 16.89, \"caption\": \"A sharp, crisp strike on the rim of a snare drum, producing a bright, dry, and percussive sound with a distinct attack and quick decay.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000093_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "88d669bb-4860-4395-aca6-39e0c04ce2a5_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "AudioData", "AudioData", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 17.8921875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 2.95, \"end_time\": 5.13, \"caption\": \"sighing\"}, {\"start_time\": 4.32, \"end_time\": 10.45, \"caption\": \"A wine glass is struck once with a soft mallet, producing a clear, soft chime.\"}, {\"start_time\": 8.18, \"end_time\": 9.48, \"caption\": \"The audio contains the distinct sound of a handgun being fired. It's a close perspective, likely capturing the sharp crack of a single shot from a Thompson/Center Contender .22 LR pistol. The sound is crisp and clear, indicative of a single shot.\"}, {\"start_time\": 9.0, \"end_time\": 16.41, \"caption\": \"The audio contains the sound of machine gun bursts and bullet ricochets off dirt. There are multiple instances of bullets impacting and bouncing off a dirt surface, creating a dynamic and somewhat chaotic soundscape. The overall impression is one of intense gunfire and the immediate aftermath of bullets striking a ground.\"}, {\"start_time\": 13.66, \"end_time\": 16.01, \"caption\": \"sighs\"}, {\"start_time\": 15.14, \"end_time\": 16.89, \"caption\": \"rim shot\"}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000094_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "88d669bb-4860-4395-aca6-39e0c04ce2a5_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "sdh_series", "AudioData", "AudioData", "AudioData", "sdh_series", "sdh_series" ], "duration_band": "short", "duration_hint": "segments should be relatively short (1–8 seconds)", "duration_s": 17.8921875, "file_overlap_type": "has_overlaps", "format": "QWEN_ChatFormat_AllAtOnce_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 6, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Provide every segment along with their start times, end times, and captions.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "[{\"start_time\": 2.95, \"end_time\": 5.13, \"caption\": \"A deep, prolonged exhalation, characterized by a soft, continuous release of breath, often conveying relief, fatigue, or emotional release.\"}, {\"start_time\": 4.32, \"end_time\": 10.45, \"caption\": \"A wine glass is tapped once using a soft mallet, creating a clear, gentle chime.\"}, {\"start_time\": 8.18, \"end_time\": 9.48, \"caption\": \"The recording captures the unmistakable sound of a handgun discharge. It comes from a close angle, probably picking up the sharp crack of a single burst from a Thompson/Center Contender .22\\u202fLR pistol. The tone is crisp and clear, signaling a single shot.\"}, {\"start_time\": 9.0, \"end_time\": 16.41, \"caption\": \"The audio features the sounds of machine gun bursts and bullets ricocheting off dirt, with multiple instances of bullets impacting and bouncing off a dirt surface, creating a dynamic and somewhat chaotic soundscape that overall conveys intense gunfire and the immediate aftermath of bullets striking the ground.\"}, {\"start_time\": 13.66, \"end_time\": 16.01, \"caption\": \"soft exhalations\"}, {\"start_time\": 15.14, \"end_time\": 16.89, \"caption\": \"The sharp, resonant clang that results when the rim of a drum is struck.\"}]", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000095_QWEN_ChatFormat_AllAtOnce_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a3b52eaa-20d8-414a-a4cb-6c7b65f20600", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.5156875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, give me the start and end for the first segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 13.52}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000096_QWEN_ChatFormat
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a3b52eaa-20d8-414a-a4cb-6c7b65f20600", "meta": { "conversation_type": "iterative", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.5156875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_Paraphrased", "global_hint_mode": "none", "include_captions": true, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio. For each turn, output JSON with required keys. Captions must be standalone (no references to previous segments), and avoid timestamps and dB numbers.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "First, provide the start and end points for the initial segment.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"start_time\": 0.51, \"end_time\": 13.52}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000097_QWEN_ChatFormat_Paraphrased
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
{ "id": "a3b52eaa-20d8-414a-a4cb-6c7b65f20600_all", "meta": { "conversation_type": "all_at_once", "datasets_used": [ "AnimalSpeak" ], "duration_band": "medium", "duration_hint": "segments should be ~5–15 seconds", "duration_s": 14.5156875, "file_overlap_type": "non_overlapping", "format": "QWEN_ChatFormat_AllAtOnce", "global_hint_mode": "none", "include_captions": false, "num_events": 1, "stated_overlap_in_first_turn": false }, "qwen_chat": { "messages": [ { "content": [ { "text": "You are an audio segmentation assistant. Listen to the attached audio and output all segments as JSON.", "type": "text" } ], "role": "system" }, { "content": [ { "text": "Give me all segments with their start times, end times.", "type": "text" } ], "role": "user" }, { "content": [ { "text": "{\"segments\": [{\"start_time\": 0.51, \"end_time\": 13.52}]}", "type": "text" } ], "role": "assistant" } ] } }
train-00000000-00000098_QWEN_ChatFormat_AllAtOnce
hf://datasets/gijs/sed-conv-events@e1630f60698cb4abc42e176cd66fe9a3de9ab8f4/train-00000000-000000.tar
End of preview. Expand in Data Studio

No dataset card yet

Downloads last month
173