{"utterance_id": "1686-142278-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0052.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.44, "milton northern"]], "response_mode": "offline", "answer": "[{\"word\":\"milton northern\",\"start\":0.67,\"end\":1.44}]"} {"utterance_id": "1701-141760-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0029.flac", "task_type": "lexical_detection", "query": "count", "prompt": "When the utterance is complete, report all instances of the word \"count\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.76, "count"]], "response_mode": "offline", "answer": "[{\"word\":\"count\",\"start\":1.34,\"end\":1.76}]"} {"utterance_id": "700-122867-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0041.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.89, "afternoon"]], "response_mode": "offline", "answer": "[{\"word\":\"afternoon\",\"start\":1.41,\"end\":1.89}]"} {"utterance_id": "6123-59186-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0010.flac", "task_type": "lexical_detection", "query": "poetry", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"poetry\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.78, "poetry"]], "response_mode": "offline", "answer": "[{\"word\":\"poetry\",\"start\":1.18,\"end\":1.78}]"} {"utterance_id": "4323-55228-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0013.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.23, 0.7, "one"], [5.07, 5.47, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.23,\"end\":0.7},{\"word\":\"two\",\"start\":5.07,\"end\":5.47}]"} {"utterance_id": "6455-67803-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0005.flac", "task_type": "semantic_counting", "query": "clock_time", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a time of day: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.67, 0.91, "one"], [3.04, 3.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.67,\"end\":0.91},{\"word\":\"two\",\"start\":3.04,\"end\":3.28}]"} {"utterance_id": "1686-142278-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0056.flac", "task_type": "lexical_detection", "query": "told", "prompt": "Process the full audio offline. Find every instance of the word \"told\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.34, "told"]], "response_mode": "offline", "answer": "[{\"word\":\"told\",\"start\":1.04,\"end\":1.34}]"} {"utterance_id": "700-122866-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0006.flac", "task_type": "lexical_detection", "query": "sloanes", "prompt": "Once the audio has finished, identify every instance of the word \"sloanes\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.85, 1.32, "sloanes"], [4.96, 5.38, "sloanes"], [7.81, 8.38, "sloanes"]], "response_mode": "offline", "answer": "[{\"word\":\"sloanes\",\"start\":0.85,\"end\":1.32},{\"word\":\"sloanes\",\"start\":4.96,\"end\":5.38},{\"word\":\"sloanes\",\"start\":7.81,\"end\":8.38}]"} {"utterance_id": "5849-50873-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0005.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Process the full audio offline. For every occurrence of a body part, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.42, 2.72, "one"], [4.57, 4.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.42,\"end\":2.72},{\"word\":\"two\",\"start\":4.57,\"end\":4.92}]"} {"utterance_id": "1585-131718-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0004.flac", "task_type": "lexical_detection", "query": "national", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"national\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.43, 1.87, "national"]], "response_mode": "offline", "answer": "[{\"word\":\"national\",\"start\":1.43,\"end\":1.87}]"} {"utterance_id": "6455-67803-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.84, "lucia"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia\",\"start\":0.5,\"end\":0.84}]"} {"utterance_id": "1630-73710-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.82, 10.24, "bat"]], "response_mode": "offline", "answer": "[{\"word\":\"bat\",\"start\":9.82,\"end\":10.24}]"} {"utterance_id": "7601-175351-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0000.flac", "task_type": "lexical_detection", "query": "revolt", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"revolt\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 2.15, "revolt"]], "response_mode": "offline", "answer": "[{\"word\":\"revolt\",\"start\":1.26,\"end\":2.15}]"} {"utterance_id": "4570-14911-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0017.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.71, 3.21, "sixteen"]], "response_mode": "offline", "answer": "[{\"word\":\"sixteen\",\"start\":2.71,\"end\":3.21}]"} {"utterance_id": "4153-186223-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0006.flac", "task_type": "lexical_detection", "query": "she'd", "prompt": "Once the audio has finished, identify every instance of the word \"she'd\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.88, 2.13, "she'd"]], "response_mode": "offline", "answer": "[{\"word\":\"she'd\",\"start\":1.88,\"end\":2.13}]"} {"utterance_id": "6455-67804-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0040.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.32, 2.55, "eye"]], "response_mode": "offline", "answer": "[{\"word\":\"eye\",\"start\":2.32,\"end\":2.55}]"} {"utterance_id": "1701-141760-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0009.flac", "task_type": "lexical_detection", "query": "friend", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"friend\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.54, 2.05, "friend"]], "response_mode": "offline", "answer": "[{\"word\":\"friend\",\"start\":1.54,\"end\":2.05}]"} {"utterance_id": "8254-115543-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0035.flac", "task_type": "lexical_counting", "query": "around", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"around\" (word, start, end). If none, reply [].", "labels": [[9.62, 9.94, "one"], [14.07, 14.35, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.62,\"end\":9.94},{\"word\":\"two\",\"start\":14.07,\"end\":14.35}]"} {"utterance_id": "7697-245712-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0012.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.87, 7.55, "plants"]], "response_mode": "offline", "answer": "[{\"word\":\"plants\",\"start\":6.87,\"end\":7.55}]"} {"utterance_id": "116-288045-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0022.flac", "task_type": "lexical_detection", "query": "aside", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"aside\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.25, 4.78, "aside"]], "response_mode": "offline", "answer": "[{\"word\":\"aside\",\"start\":4.25,\"end\":4.78}]"} {"utterance_id": "4570-56594-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0009.flac", "task_type": "lexical_detection", "query": "mouth", "prompt": "Process the full audio offline. Find every instance of the word \"mouth\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.56, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":1.06,\"end\":1.56}]"} {"utterance_id": "1651-136854-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0005.flac", "task_type": "lexical_detection", "query": "distinct", "prompt": "When the utterance is complete, report all instances of the word \"distinct\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.45, "distinct"]], "response_mode": "offline", "answer": "[{\"word\":\"distinct\",\"start\":1.9,\"end\":2.45}]"} {"utterance_id": "5849-50963-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0007.flac", "task_type": "lexical_detection", "query": "value", "prompt": "After the audio ends, extract all occurrences of the word \"value\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.3, "value"]], "response_mode": "offline", "answer": "[{\"word\":\"value\",\"start\":0.73,\"end\":1.3}]"} {"utterance_id": "1255-90413-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 1.08, "jocelyn"]], "response_mode": "offline", "answer": "[{\"word\":\"jocelyn\",\"start\":0.68,\"end\":1.08}]"} {"utterance_id": "6455-66379-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0017.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a plant name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.3, 0.49, "wood"]], "response_mode": "offline", "answer": "[{\"word\":\"wood\",\"start\":0.3,\"end\":0.49}]"} {"utterance_id": "6267-53049-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0001.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.62, 8.24, "sister"]], "response_mode": "offline", "answer": "[{\"word\":\"sister\",\"start\":7.62,\"end\":8.24}]"} {"utterance_id": "4572-112383-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0008.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.72, 6.35, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":5.72,\"end\":6.35}]"} {"utterance_id": "4323-18416-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.39, 2.82, "pickering"]], "response_mode": "offline", "answer": "[{\"word\":\"pickering\",\"start\":2.39,\"end\":2.82}]"} {"utterance_id": "4831-25894-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.74, "tessa"], [2.69, 3.13, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":0.4,\"end\":0.74},{\"word\":\"tommo\",\"start\":2.69,\"end\":3.13}]"} {"utterance_id": "7697-245712-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0016.flac", "task_type": "lexical_counting", "query": "nature", "prompt": "After listening to the whole recording, output timed count words for the word \"nature\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.46, 2.93, "one"], [5.31, 5.85, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.46,\"end\":2.93},{\"word\":\"two\",\"start\":5.31,\"end\":5.85}]"} {"utterance_id": "3660-6517-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0010.flac", "task_type": "lexical_counting", "query": "carcases", "prompt": "After listening to the whole recording, output timed count words for the word \"carcases\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[3.0, 3.57, "one"], [4.58, 5.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.0,\"end\":3.57},{\"word\":\"two\",\"start\":4.58,\"end\":5.13}]"} {"utterance_id": "4323-13259-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0008.flac", "task_type": "lexical_detection", "query": "conventional", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"conventional\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.83, "conventional"]], "response_mode": "offline", "answer": "[{\"word\":\"conventional\",\"start\":2.11,\"end\":2.83}]"} {"utterance_id": "3660-6517-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0017.flac", "task_type": "lexical_detection", "query": "poop", "prompt": "Process the full audio offline. Find every instance of the word \"poop\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.55, "poop"]], "response_mode": "offline", "answer": "[{\"word\":\"poop\",\"start\":2.22,\"end\":2.55}]"} {"utterance_id": "4831-18525-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0025.flac", "task_type": "lexical_detection", "query": "whisper", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"whisper\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.3, 2.7, "whisper"]], "response_mode": "offline", "answer": "[{\"word\":\"whisper\",\"start\":2.3,\"end\":2.7}]"} {"utterance_id": "700-122868-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0016.flac", "task_type": "lexical_detection", "query": "chance", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"chance\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.42, "chance"]], "response_mode": "offline", "answer": "[{\"word\":\"chance\",\"start\":0.95,\"end\":1.42}]"} {"utterance_id": "8254-115543-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0034.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.63, "neck"]], "response_mode": "offline", "answer": "[{\"word\":\"neck\",\"start\":1.38,\"end\":1.63}]"} {"utterance_id": "1686-142278-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0008.flac", "task_type": "lexical_detection", "query": "margaret", "prompt": "After listening to the whole recording, output every instance of the word \"margaret\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.08, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.5,\"end\":1.08}]"} {"utterance_id": "3915-98647-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0033.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[15.33, 15.94, "coolies"]], "response_mode": "offline", "answer": "[{\"word\":\"coolies\",\"start\":15.33,\"end\":15.94}]"} {"utterance_id": "4515-11057-0109", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0109.flac", "task_type": "lexical_detection", "query": "pretend", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"pretend\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.35, 5.78, "pretend"]], "response_mode": "offline", "answer": "[{\"word\":\"pretend\",\"start\":5.35,\"end\":5.78}]"} {"utterance_id": "4572-112375-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0005.flac", "task_type": "lexical_detection", "query": "path", "prompt": "Process the full audio offline. Find every instance of the word \"path\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.56, 5.01, "path"]], "response_mode": "offline", "answer": "[{\"word\":\"path\",\"start\":4.56,\"end\":5.01}]"} {"utterance_id": "6123-59150-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0019.flac", "task_type": "lexical_counting", "query": "grotesques", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"grotesques\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.48, 2.22, "one"], [3.74, 4.5, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.48,\"end\":2.22},{\"word\":\"two\",\"start\":3.74,\"end\":4.5}]"} {"utterance_id": "6455-67803-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0015.flac", "task_type": "lexical_detection", "query": "little", "prompt": "After the audio ends, extract all occurrences of the word \"little\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 1.88, "little"]], "response_mode": "offline", "answer": "[{\"word\":\"little\",\"start\":1.55,\"end\":1.88}]"} {"utterance_id": "1650-167613-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0044.flac", "task_type": "lexical_detection", "query": "success", "prompt": "When the utterance is complete, report all instances of the word \"success\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 2.1, "success"]], "response_mode": "offline", "answer": "[{\"word\":\"success\",\"start\":1.53,\"end\":2.1}]"} {"utterance_id": "4323-55228-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0004.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.8, 5.11, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":4.8,\"end\":5.11}]"} {"utterance_id": "1630-96099-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0017.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.21, 0.49, "one"], [4.4, 4.77, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.21,\"end\":0.49},{\"word\":\"two\",\"start\":4.4,\"end\":4.77}]"} {"utterance_id": "2506-11278-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0025.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 2.88, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":2.48,\"end\":2.88}]"} {"utterance_id": "1686-142278-0092", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0092.flac", "task_type": "lexical_detection", "query": "sadly", "prompt": "After the audio ends, extract all occurrences of the word \"sadly\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.29, 2.68, "sadly"]], "response_mode": "offline", "answer": "[{\"word\":\"sadly\",\"start\":2.29,\"end\":2.68}]"} {"utterance_id": "116-288045-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.93, 4.56, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":3.93,\"end\":4.56}]"} {"utterance_id": "1651-136854-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[5.98, 6.38, "teacher"]], "response_mode": "offline", "answer": "[{\"word\":\"teacher\",\"start\":5.98,\"end\":6.38}]"} {"utterance_id": "4515-11057-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0057.flac", "task_type": "lexical_detection", "query": "looked", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"looked\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.08, 2.38, "looked"]], "response_mode": "offline", "answer": "[{\"word\":\"looked\",\"start\":2.08,\"end\":2.38}]"} {"utterance_id": "6123-59186-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0009.flac", "task_type": "lexical_detection", "query": "catering", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"catering\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.03, "catering"]], "response_mode": "offline", "answer": "[{\"word\":\"catering\",\"start\":0.52,\"end\":1.03}]"} {"utterance_id": "6267-53049-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0024.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 3.41, "one"], [5.83, 6.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.79,\"end\":3.41},{\"word\":\"two\",\"start\":5.83,\"end\":6.66}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an organization name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.54, 3.23, "garrison"]], "response_mode": "offline", "answer": "[{\"word\":\"garrison\",\"start\":2.54,\"end\":3.23}]"} {"utterance_id": "6841-88291-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0017.flac", "task_type": "lexical_detection", "query": "promptly", "prompt": "After the audio ends, extract all occurrences of the word \"promptly\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.96, "promptly"]], "response_mode": "offline", "answer": "[{\"word\":\"promptly\",\"start\":0.54,\"end\":0.96}]"} {"utterance_id": "1630-102884-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.63, 6.23, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":5.63,\"end\":6.23}]"} {"utterance_id": "7641-96684-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0006.flac", "task_type": "lexical_detection", "query": "happiness", "prompt": "Once the audio has finished, identify every instance of the word \"happiness\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.84, 1.35, "happiness"]], "response_mode": "offline", "answer": "[{\"word\":\"happiness\",\"start\":0.84,\"end\":1.35}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "lexical_counting", "query": "land", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"land\" with word/start/end. If none, reply [].", "labels": [[1.41, 1.99, "one"], [8.01, 8.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.41,\"end\":1.99},{\"word\":\"two\",\"start\":8.01,\"end\":8.56}]"} {"utterance_id": "1686-142278-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0028.flac", "task_type": "lexical_detection", "query": "terribly", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"terribly\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.49, 3.96, "terribly"]], "response_mode": "offline", "answer": "[{\"word\":\"terribly\",\"start\":3.49,\"end\":3.96}]"} {"utterance_id": "116-288045-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0020.flac", "task_type": "lexical_detection", "query": "son", "prompt": "After the audio ends, extract all occurrences of the word \"son\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 1.74, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":1.51,\"end\":1.74}]"} {"utterance_id": "2506-11278-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0026.flac", "task_type": "lexical_detection", "query": "seventeen", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"seventeen\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.36, 0.86, "seventeen"]], "response_mode": "offline", "answer": "[{\"word\":\"seventeen\",\"start\":0.36,\"end\":0.86}]"} {"utterance_id": "4323-18416-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0033.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[1.84, 2.14, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":1.84,\"end\":2.14}]"} {"utterance_id": "6841-88291-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0007.flac", "task_type": "lexical_detection", "query": "wide", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wide\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.43, "wide"]], "response_mode": "offline", "answer": "[{\"word\":\"wide\",\"start\":1.09,\"end\":1.43}]"} {"utterance_id": "6841-88291-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0029.flac", "task_type": "lexical_detection", "query": "imagine", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"imagine\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 0.76, "imagine"]], "response_mode": "offline", "answer": "[{\"word\":\"imagine\",\"start\":0.35,\"end\":0.76}]"} {"utterance_id": "4831-18525-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0010.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.03, "third"]], "response_mode": "offline", "answer": "[{\"word\":\"third\",\"start\":0.79,\"end\":1.03}]"} {"utterance_id": "8254-84205-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0043.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.75, 1.94, "ned"]], "response_mode": "offline", "answer": "[{\"word\":\"ned\",\"start\":1.75,\"end\":1.94}]"} {"utterance_id": "6123-59186-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0001.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.29, "midnight"]], "response_mode": "offline", "answer": "[{\"word\":\"midnight\",\"start\":4.82,\"end\":5.29}]"} {"utterance_id": "6455-67804-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.41, "lucia"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia\",\"start\":1.05,\"end\":1.41}]"} {"utterance_id": "1650-173552-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0006.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[13.7, 14.42, "moscow"]], "response_mode": "offline", "answer": "[{\"word\":\"moscow\",\"start\":13.7,\"end\":14.42}]"} {"utterance_id": "8288-274162-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0005.flac", "task_type": "lexical_detection", "query": "understand", "prompt": "After listening to the whole recording, output every instance of the word \"understand\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.26, "understand"]], "response_mode": "offline", "answer": "[{\"word\":\"understand\",\"start\":0.69,\"end\":1.26}]"} {"utterance_id": "1585-157660-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0013.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.15, 6.92, "bull pups"]], "response_mode": "offline", "answer": "[{\"word\":\"bull pups\",\"start\":6.15,\"end\":6.92}]"} {"utterance_id": "5849-50873-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0023.flac", "task_type": "lexical_detection", "query": "lighted", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"lighted\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.17, 3.53, "lighted"]], "response_mode": "offline", "answer": "[{\"word\":\"lighted\",\"start\":3.17,\"end\":3.53}]"} {"utterance_id": "4323-13259-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.14, 7.54, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":7.14,\"end\":7.54}]"} {"utterance_id": "8254-84205-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.34, 1.76, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":1.34,\"end\":1.76}]"} {"utterance_id": "1701-141759-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0018.flac", "task_type": "lexical_detection", "query": "write", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"write\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.19, "write"]], "response_mode": "offline", "answer": "[{\"word\":\"write\",\"start\":0.95,\"end\":1.19}]"} {"utterance_id": "1255-90407-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0004.flac", "task_type": "lexical_detection", "query": "least", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"least\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.67, 3.05, "least"]], "response_mode": "offline", "answer": "[{\"word\":\"least\",\"start\":2.67,\"end\":3.05}]"} {"utterance_id": "6455-67803-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0024.flac", "task_type": "lexical_counting", "query": "angry", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"angry\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.91, 1.29, "one"], [2.93, 3.31, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.91,\"end\":1.29},{\"word\":\"two\",\"start\":2.93,\"end\":3.31}]"} {"utterance_id": "7601-101619-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.67, 4.1, "painters"], [17.41, 17.99, "painter's"]], "response_mode": "offline", "answer": "[{\"word\":\"painters\",\"start\":3.67,\"end\":4.1},{\"word\":\"painter's\",\"start\":17.41,\"end\":17.99}]"} {"utterance_id": "4570-14911-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0007.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.61, 14.12, "peterborough"]], "response_mode": "offline", "answer": "[{\"word\":\"peterborough\",\"start\":13.61,\"end\":14.12}]"} {"utterance_id": "1651-136854-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0011.flac", "task_type": "lexical_detection", "query": "fancy", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"fancy\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.62, 1.09, "fancy"]], "response_mode": "offline", "answer": "[{\"word\":\"fancy\",\"start\":0.62,\"end\":1.09}]"} {"utterance_id": "8254-84205-0072", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0072.flac", "task_type": "lexical_counting", "query": "they'd", "prompt": "After the audio ends, tally every instance of the word \"they'd\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.25, 0.39, "one"], [4.0, 4.16, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.25,\"end\":0.39},{\"word\":\"two\",\"start\":4.0,\"end\":4.16}]"} {"utterance_id": "4323-55228-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0049.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.73, "aunt hannah"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt hannah\",\"start\":0.31,\"end\":0.73}]"} {"utterance_id": "1686-142278-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0052.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.34, 6.73, "daughter's"]], "response_mode": "offline", "answer": "[{\"word\":\"daughter's\",\"start\":6.34,\"end\":6.73}]"} {"utterance_id": "6841-88294-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0029.flac", "task_type": "lexical_detection", "query": "picked", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"picked\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.29, 1.6, "picked"]], "response_mode": "offline", "answer": "[{\"word\":\"picked\",\"start\":1.29,\"end\":1.6}]"} {"utterance_id": "1701-141760-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0002.flac", "task_type": "lexical_detection", "query": "regiments", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"regiments\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 1.14, "regiments"]], "response_mode": "offline", "answer": "[{\"word\":\"regiments\",\"start\":0.64,\"end\":1.14}]"} {"utterance_id": "1651-136854-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0007.flac", "task_type": "lexical_counting", "query": "time", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"time\" with word/start/end. If none, reply [].", "labels": [[5.67, 6.33, "one"], [11.36, 11.85, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.67,\"end\":6.33},{\"word\":\"two\",\"start\":11.36,\"end\":11.85}]"} {"utterance_id": "6455-67803-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.52, 2.87, "lucia"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia\",\"start\":2.52,\"end\":2.87}]"} {"utterance_id": "4153-186222-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0026.flac", "task_type": "lexical_detection", "query": "silas", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"silas\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.15, "silas"]], "response_mode": "offline", "answer": "[{\"word\":\"silas\",\"start\":0.52,\"end\":1.15}]"} {"utterance_id": "3660-172182-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.64, 0.96, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":0.64,\"end\":0.96}]"} {"utterance_id": "4153-186222-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0032.flac", "task_type": "lexical_detection", "query": "equally", "prompt": "After listening to the whole recording, output every instance of the word \"equally\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.35, "equally"]], "response_mode": "offline", "answer": "[{\"word\":\"equally\",\"start\":0.9,\"end\":1.35}]"} {"utterance_id": "4831-18525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0000.flac", "task_type": "lexical_detection", "query": "music", "prompt": "After the audio ends, extract all occurrences of the word \"music\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.97, 6.46, "music"]], "response_mode": "offline", "answer": "[{\"word\":\"music\",\"start\":5.97,\"end\":6.46}]"} {"utterance_id": "6467-94831-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0006.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.65, 1.02, "fifteen"]], "response_mode": "offline", "answer": "[{\"word\":\"fifteen\",\"start\":0.65,\"end\":1.02}]"} {"utterance_id": "4831-25894-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0015.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After listening to the whole recording, output every instance of a money or currency term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.61, "dollar"], [2.62, 2.9, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"dollar\",\"start\":1.18,\"end\":1.61},{\"word\":\"money\",\"start\":2.62,\"end\":2.9}]"} {"utterance_id": "4515-11057-0088", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0088.flac", "task_type": "lexical_detection", "query": "end", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"end\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.55, 3.97, "end"]], "response_mode": "offline", "answer": "[{\"word\":\"end\",\"start\":3.55,\"end\":3.97}]"} {"utterance_id": "4323-18416-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0028.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.18, "sister's"]], "response_mode": "offline", "answer": "[{\"word\":\"sister's\",\"start\":0.82,\"end\":1.18}]"} {"utterance_id": "6467-94831-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0027.flac", "task_type": "lexical_detection", "query": "shot", "prompt": "Process the full audio offline. Find every instance of the word \"shot\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.37, "shot"]], "response_mode": "offline", "answer": "[{\"word\":\"shot\",\"start\":2.08,\"end\":2.37}]"} {"utterance_id": "6599-38591-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0004.flac", "task_type": "lexical_detection", "query": "spanish", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"spanish\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.41, "spanish"], [4.48, 4.95, "spanish"]], "response_mode": "offline", "answer": "[{\"word\":\"spanish\",\"start\":2.93,\"end\":3.41},{\"word\":\"spanish\",\"start\":4.48,\"end\":4.95}]"} {"utterance_id": "1686-142278-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0058.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.97, 4.7, "bracy common"]], "response_mode": "offline", "answer": "[{\"word\":\"bracy common\",\"start\":3.97,\"end\":4.7}]"} {"utterance_id": "7601-175351-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0023.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.3, 1.91, "hatteras"]], "response_mode": "offline", "answer": "[{\"word\":\"hatteras\",\"start\":1.3,\"end\":1.91}]"} {"utterance_id": "700-122866-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 2.04, "geraldine's"]], "response_mode": "offline", "answer": "[{\"word\":\"geraldine's\",\"start\":1.4,\"end\":2.04}]"} {"utterance_id": "1701-141759-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.1, 8.12, "anna mikhaylovna"]], "response_mode": "offline", "answer": "[{\"word\":\"anna mikhaylovna\",\"start\":7.1,\"end\":8.12}]"} {"utterance_id": "6123-59150-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.7, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.54,\"end\":0.7}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 2.12, "thucydides"]], "response_mode": "offline", "answer": "[{\"word\":\"thucydides\",\"start\":1.1,\"end\":2.12}]"} {"utterance_id": "7697-105815-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.09, 2.54, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":2.09,\"end\":2.54}]"} {"utterance_id": "4515-11057-0110", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0110.flac", "task_type": "lexical_detection", "query": "yesterday", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"yesterday\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 1.03, "yesterday"]], "response_mode": "offline", "answer": "[{\"word\":\"yesterday\",\"start\":0.46,\"end\":1.03}]"} {"utterance_id": "7697-245712-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0004.flac", "task_type": "lexical_detection", "query": "proved", "prompt": "Once the audio has finished, identify every instance of the word \"proved\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.44, 1.8, "proved"]], "response_mode": "offline", "answer": "[{\"word\":\"proved\",\"start\":1.44,\"end\":1.8}]"} {"utterance_id": "6455-66379-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0015.flac", "task_type": "lexical_detection", "query": "six", "prompt": "After the audio ends, extract all occurrences of the word \"six\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.64, "six"]], "response_mode": "offline", "answer": "[{\"word\":\"six\",\"start\":2.27,\"end\":2.64}]"} {"utterance_id": "4831-29134-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0007.flac", "task_type": "lexical_detection", "query": "higher", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"higher\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.6, "higher"]], "response_mode": "offline", "answer": "[{\"word\":\"higher\",\"start\":1.06,\"end\":1.6}]"} {"utterance_id": "8288-274150-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0002.flac", "task_type": "lexical_detection", "query": "appearance", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"appearance\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.15, 10.81, "appearance"]], "response_mode": "offline", "answer": "[{\"word\":\"appearance\",\"start\":10.15,\"end\":10.81}]"} {"utterance_id": "6123-59150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0003.flac", "task_type": "lexical_detection", "query": "destruction", "prompt": "After listening to the whole recording, output every instance of the word \"destruction\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.73, 3.35, "destruction"]], "response_mode": "offline", "answer": "[{\"word\":\"destruction\",\"start\":2.73,\"end\":3.35}]"} {"utterance_id": "700-122868-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0007.flac", "task_type": "lexical_detection", "query": "romantic", "prompt": "Once the audio has finished, identify every instance of the word \"romantic\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.46, 2.99, "romantic"]], "response_mode": "offline", "answer": "[{\"word\":\"romantic\",\"start\":2.46,\"end\":2.99}]"} {"utterance_id": "6267-53049-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0020.flac", "task_type": "lexical_detection", "query": "admiration", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"admiration\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.21, 5.93, "admiration"]], "response_mode": "offline", "answer": "[{\"word\":\"admiration\",\"start\":5.21,\"end\":5.93}]"} {"utterance_id": "1686-142278-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.97, "mister lennox"]], "response_mode": "offline", "answer": "[{\"word\":\"mister lennox\",\"start\":0.26,\"end\":0.97}]"} {"utterance_id": "6841-88294-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0064.flac", "task_type": "lexical_counting", "query": "why", "prompt": "After listening to the whole recording, output timed count words for the word \"why\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.52, 0.77, "one"], [1.85, 2.12, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.52,\"end\":0.77},{\"word\":\"two\",\"start\":1.85,\"end\":2.12}]"} {"utterance_id": "1650-157641-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0011.flac", "task_type": "lexical_detection", "query": "france", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"france\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.65, 4.02, "france"]], "response_mode": "offline", "answer": "[{\"word\":\"france\",\"start\":3.65,\"end\":4.02}]"} {"utterance_id": "1255-90407-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0001.flac", "task_type": "lexical_detection", "query": "rain", "prompt": "When the utterance is complete, report all instances of the word \"rain\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.15, "rain"]], "response_mode": "offline", "answer": "[{\"word\":\"rain\",\"start\":0.86,\"end\":1.15}]"} {"utterance_id": "6267-65525-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0054.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.23, 3.88, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":3.23,\"end\":3.88}]"} {"utterance_id": "8254-115543-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0002.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Listen to the full audio. Count every occurrence of a profession. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.3, "one"], [3.57, 4.06, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.92,\"end\":1.3},{\"word\":\"two\",\"start\":3.57,\"end\":4.06}]"} {"utterance_id": "5849-50964-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.15, 2.44, "nellie"]], "response_mode": "offline", "answer": "[{\"word\":\"nellie\",\"start\":2.15,\"end\":2.44}]"} {"utterance_id": "4323-18416-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0026.flac", "task_type": "lexical_detection", "query": "writing", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"writing\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.09, "writing"]], "response_mode": "offline", "answer": "[{\"word\":\"writing\",\"start\":0.8,\"end\":1.09}]"} {"utterance_id": "1686-142278-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0017.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.87, 3.1, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":2.87,\"end\":3.1}]"} {"utterance_id": "6455-67804-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.34, 0.89, "octavia"], [2.45, 3.11, "miss belinda"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":0.34,\"end\":0.89},{\"word\":\"miss belinda\",\"start\":2.45,\"end\":3.11}]"} {"utterance_id": "5849-50873-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.22, "mister prince"]], "response_mode": "offline", "answer": "[{\"word\":\"mister prince\",\"start\":0.5,\"end\":1.22}]"} {"utterance_id": "7697-105815-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0021.flac", "task_type": "lexical_detection", "query": "hill", "prompt": "Process the full audio offline. Find every instance of the word \"hill\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.46, 5.67, "hill"]], "response_mode": "offline", "answer": "[{\"word\":\"hill\",\"start\":5.46,\"end\":5.67}]"} {"utterance_id": "6599-38590-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0002.flac", "task_type": "lexical_counting", "query": "say", "prompt": "After listening to the whole recording, output timed count words for the word \"say\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[4.06, 4.28, "one"], [5.52, 5.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.06,\"end\":4.28},{\"word\":\"two\",\"start\":5.52,\"end\":5.92}]"} {"utterance_id": "1585-157660-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.49, 0.88, "william"], [7.29, 7.75, "jack"]], "response_mode": "offline", "answer": "[{\"word\":\"william\",\"start\":0.49,\"end\":0.88},{\"word\":\"jack\",\"start\":7.29,\"end\":7.75}]"} {"utterance_id": "1650-167613-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0024.flac", "task_type": "lexical_detection", "query": "lose", "prompt": "After listening to the whole recording, output every instance of the word \"lose\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.13, "lose"]], "response_mode": "offline", "answer": "[{\"word\":\"lose\",\"start\":0.82,\"end\":1.13}]"} {"utterance_id": "4153-185072-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.05, 3.54, "officer"]], "response_mode": "offline", "answer": "[{\"word\":\"officer\",\"start\":3.05,\"end\":3.54}]"} {"utterance_id": "6123-59150-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0006.flac", "task_type": "lexical_detection", "query": "less", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"less\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.93, 4.19, "less"]], "response_mode": "offline", "answer": "[{\"word\":\"less\",\"start\":3.93,\"end\":4.19}]"} {"utterance_id": "2506-11278-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0000.flac", "task_type": "lexical_detection", "query": "snobbishness", "prompt": "After listening to the whole recording, output every instance of the word \"snobbishness\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.79, "snobbishness"]], "response_mode": "offline", "answer": "[{\"word\":\"snobbishness\",\"start\":2.06,\"end\":2.79}]"} {"utterance_id": "5849-50962-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[4.08, 4.45, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":4.08,\"end\":4.45}]"} {"utterance_id": "8254-84205-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0042.flac", "task_type": "lexical_detection", "query": "boldly", "prompt": "After the audio ends, extract all occurrences of the word \"boldly\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.31, 1.67, "boldly"]], "response_mode": "offline", "answer": "[{\"word\":\"boldly\",\"start\":1.31,\"end\":1.67}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "lexical_detection", "query": "lofty", "prompt": "Process the full audio offline. Find every instance of the word \"lofty\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.1, "lofty"]], "response_mode": "offline", "answer": "[{\"word\":\"lofty\",\"start\":0.54,\"end\":1.1}]"} {"utterance_id": "1650-167613-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.24, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":0.4,\"end\":1.24}]"} {"utterance_id": "1686-142278-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0079.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.91, 2.33, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.91,\"end\":2.33}]"} {"utterance_id": "8288-274162-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0023.flac", "task_type": "lexical_detection", "query": "son", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"son\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.0, 2.6, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":2.0,\"end\":2.6}]"} {"utterance_id": "4831-29134-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0006.flac", "task_type": "lexical_detection", "query": "soul", "prompt": "After the audio ends, extract all occurrences of the word \"soul\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.78, "soul"]], "response_mode": "offline", "answer": "[{\"word\":\"soul\",\"start\":0.44,\"end\":0.78}]"} {"utterance_id": "3915-98647-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0002.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Once the audio has finished, list each occurrence of a family relation as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.12, 2.47, "one"], [5.58, 6.0, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.12,\"end\":2.47},{\"word\":\"two\",\"start\":5.58,\"end\":6.0}]"} {"utterance_id": "3660-172182-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.37, 2.77, "arthur"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur\",\"start\":2.37,\"end\":2.77}]"} {"utterance_id": "6123-59150-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0024.flac", "task_type": "lexical_detection", "query": "author", "prompt": "Process the full audio offline. Find every instance of the word \"author\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.71, "author"]], "response_mode": "offline", "answer": "[{\"word\":\"author\",\"start\":3.22,\"end\":3.71}]"} {"utterance_id": "6841-88291-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0031.flac", "task_type": "lexical_detection", "query": "teeth", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"teeth\" with fields word, start, end (seconds). If none, reply [].", "labels": [[7.37, 7.66, "teeth"]], "response_mode": "offline", "answer": "[{\"word\":\"teeth\",\"start\":7.37,\"end\":7.66}]"} {"utterance_id": "4831-25894-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.81, 1.17, "ranza"], [4.51, 4.87, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"ranza\",\"start\":0.81,\"end\":1.17},{\"word\":\"tessa\",\"start\":4.51,\"end\":4.87}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "lexical_counting", "query": "tessa", "prompt": "When the utterance is complete, report how the word \"tessa\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.91, 2.24, "one"], [11.81, 12.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.91,\"end\":2.24},{\"word\":\"two\",\"start\":11.81,\"end\":12.28}]"} {"utterance_id": "1686-142278-0081", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0081.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.85, 1.25, "milton"], [9.94, 10.46, "helstone"]], "response_mode": "offline", "answer": "[{\"word\":\"milton\",\"start\":0.85,\"end\":1.25},{\"word\":\"helstone\",\"start\":9.94,\"end\":10.46}]"} {"utterance_id": "1255-90407-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0021.flac", "task_type": "lexical_detection", "query": "pitied", "prompt": "Process the full audio offline. Find every instance of the word \"pitied\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.03, "pitied"]], "response_mode": "offline", "answer": "[{\"word\":\"pitied\",\"start\":0.72,\"end\":1.03}]"} {"utterance_id": "1255-90413-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0010.flac", "task_type": "lexical_detection", "query": "twisted", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"twisted\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.54, 6.83, "twisted"]], "response_mode": "offline", "answer": "[{\"word\":\"twisted\",\"start\":6.54,\"end\":6.83}]"} {"utterance_id": "1585-131718-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0046.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.04, 4.65, "sixty five"]], "response_mode": "offline", "answer": "[{\"word\":\"sixty five\",\"start\":4.04,\"end\":4.65}]"} {"utterance_id": "4153-186223-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 3.21, "aunt miss merrick sir"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt miss merrick sir\",\"start\":1.25,\"end\":3.21}]"} {"utterance_id": "7641-96670-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0019.flac", "task_type": "lexical_detection", "query": "crown", "prompt": "After listening to the whole recording, output every instance of the word \"crown\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.31, "crown"]], "response_mode": "offline", "answer": "[{\"word\":\"crown\",\"start\":2.98,\"end\":3.31}]"} {"utterance_id": "6123-59186-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0040.flac", "task_type": "lexical_detection", "query": "blame", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"blame\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.96, 6.31, "blame"]], "response_mode": "offline", "answer": "[{\"word\":\"blame\",\"start\":5.96,\"end\":6.31}]"} {"utterance_id": "5543-27761-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0036.flac", "task_type": "lexical_detection", "query": "else", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"else\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.03, 2.41, "else"]], "response_mode": "offline", "answer": "[{\"word\":\"else\",\"start\":2.03,\"end\":2.41}]"} {"utterance_id": "3915-98647-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0020.flac", "task_type": "lexical_detection", "query": "chin", "prompt": "After the audio ends, extract all occurrences of the word \"chin\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.22, 5.65, "chin"]], "response_mode": "offline", "answer": "[{\"word\":\"chin\",\"start\":5.22,\"end\":5.65}]"} {"utterance_id": "5849-50873-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0021.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[1.22, 1.66, "one"], [3.56, 4.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.22,\"end\":1.66},{\"word\":\"two\",\"start\":3.56,\"end\":4.49}]"} {"utterance_id": "3915-98647-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0023.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.3, 4.69, "tea"]], "response_mode": "offline", "answer": "[{\"word\":\"tea\",\"start\":4.3,\"end\":4.69}]"} {"utterance_id": "4515-11057-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0047.flac", "task_type": "lexical_detection", "query": "fallen", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fallen\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.52, "fallen"]], "response_mode": "offline", "answer": "[{\"word\":\"fallen\",\"start\":2.06,\"end\":2.52}]"} {"utterance_id": "8288-274150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.8, 4.16, "painter"]], "response_mode": "offline", "answer": "[{\"word\":\"painter\",\"start\":3.8,\"end\":4.16}]"} {"utterance_id": "1686-142278-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0031.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 3.82, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.6,\"end\":3.82}]"} {"utterance_id": "3663-172005-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0003.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "When the utterance is complete, report how an animal name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 1.0, "one"], [3.91, 4.29, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.57,\"end\":1.0},{\"word\":\"two\",\"start\":3.91,\"end\":4.29}]"} {"utterance_id": "3660-172182-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0015.flac", "task_type": "lexical_detection", "query": "guenever", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"guenever\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.13, 3.6, "guenever"]], "response_mode": "offline", "answer": "[{\"word\":\"guenever\",\"start\":3.13,\"end\":3.6}]"} {"utterance_id": "4831-29134-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0010.flac", "task_type": "semantic_detection", "query": "date", "prompt": "After the audio ends, extract all occurrences of a date. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.18, 5.47, "sunday"], [7.1, 7.45, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":5.18,\"end\":5.47},{\"word\":\"sunday\",\"start\":7.1,\"end\":7.45}]"} {"utterance_id": "116-288048-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.72, 5.37, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":4.72,\"end\":5.37}]"} {"utterance_id": "5543-27761-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0066.flac", "task_type": "lexical_detection", "query": "weakness", "prompt": "Process the full audio offline. Find every instance of the word \"weakness\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.71, 3.36, "weakness"]], "response_mode": "offline", "answer": "[{\"word\":\"weakness\",\"start\":2.71,\"end\":3.36}]"} {"utterance_id": "8254-84205-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0050.flac", "task_type": "lexical_detection", "query": "indians", "prompt": "After the audio ends, extract all occurrences of the word \"indians\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.84, "indians"]], "response_mode": "offline", "answer": "[{\"word\":\"indians\",\"start\":0.48,\"end\":0.84}]"} {"utterance_id": "700-122866-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0018.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.86, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":1.45,\"end\":1.86}]"} {"utterance_id": "4570-24733-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0001.flac", "task_type": "lexical_detection", "query": "conceive", "prompt": "Once the audio has finished, identify every instance of the word \"conceive\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.45, 3.4, "conceive"]], "response_mode": "offline", "answer": "[{\"word\":\"conceive\",\"start\":2.45,\"end\":3.4}]"} {"utterance_id": "4572-64670-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.7, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":1.3,\"end\":1.7}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "lexical_detection", "query": "rider", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"rider\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.89, 3.19, "rider"]], "response_mode": "offline", "answer": "[{\"word\":\"rider\",\"start\":2.89,\"end\":3.19}]"} {"utterance_id": "5543-27761-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0035.flac", "task_type": "lexical_detection", "query": "i'll", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"i'll\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.68, "i'll"]], "response_mode": "offline", "answer": "[{\"word\":\"i'll\",\"start\":0.53,\"end\":0.68}]"} {"utterance_id": "5849-50873-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0039.flac", "task_type": "lexical_detection", "query": "rhode", "prompt": "When the utterance is complete, report all instances of the word \"rhode\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.41, 5.53, "rhode"]], "response_mode": "offline", "answer": "[{\"word\":\"rhode\",\"start\":5.41,\"end\":5.53}]"} {"utterance_id": "6467-94831-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0039.flac", "task_type": "lexical_detection", "query": "donkey", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"donkey\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.42, 2.02, "donkey"]], "response_mode": "offline", "answer": "[{\"word\":\"donkey\",\"start\":1.42,\"end\":2.02}]"} {"utterance_id": "1255-90407-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0008.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.04, "father's"]], "response_mode": "offline", "answer": "[{\"word\":\"father's\",\"start\":0.6,\"end\":1.04}]"} {"utterance_id": "1686-142278-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0028.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.97, 5.37, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":4.97,\"end\":5.37}]"} {"utterance_id": "4323-55228-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0001.flac", "task_type": "lexical_detection", "query": "gasped", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"gasped\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.79, 3.1, "gasped"]], "response_mode": "offline", "answer": "[{\"word\":\"gasped\",\"start\":2.79,\"end\":3.1}]"} {"utterance_id": "1650-167613-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.34, 0.67, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":0.34,\"end\":0.67}]"} {"utterance_id": "6455-66379-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0000.flac", "task_type": "lexical_detection", "query": "wood", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"wood\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 2.14, "wood"]], "response_mode": "offline", "answer": "[{\"word\":\"wood\",\"start\":1.81,\"end\":2.14}]"} {"utterance_id": "4153-186222-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0035.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After listening to the whole recording, output every instance of a money or currency term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.75, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":2.35,\"end\":2.75}]"} {"utterance_id": "4515-11057-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0033.flac", "task_type": "lexical_detection", "query": "sometimes", "prompt": "After the audio ends, extract all occurrences of the word \"sometimes\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.85, "sometimes"]], "response_mode": "offline", "answer": "[{\"word\":\"sometimes\",\"start\":0.23,\"end\":0.85}]"} {"utterance_id": "6267-65525-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0014.flac", "task_type": "lexical_detection", "query": "answering", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"answering\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[14.29, 14.83, "answering"]], "response_mode": "offline", "answer": "[{\"word\":\"answering\",\"start\":14.29,\"end\":14.83}]"} {"utterance_id": "6455-67803-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0025.flac", "task_type": "lexical_detection", "query": "theobald", "prompt": "After the audio ends, extract all occurrences of the word \"theobald\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.18, "theobald"]], "response_mode": "offline", "answer": "[{\"word\":\"theobald\",\"start\":2.61,\"end\":3.18}]"} {"utterance_id": "2506-11278-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0034.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.36, "hearts"]], "response_mode": "offline", "answer": "[{\"word\":\"hearts\",\"start\":2.75,\"end\":3.36}]"} {"utterance_id": "6841-88294-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0054.flac", "task_type": "lexical_detection", "query": "true", "prompt": "When the utterance is complete, report all instances of the word \"true\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.46, "true"]], "response_mode": "offline", "answer": "[{\"word\":\"true\",\"start\":1.19,\"end\":1.46}]"} {"utterance_id": "4515-11057-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0060.flac", "task_type": "lexical_detection", "query": "consent", "prompt": "After listening to the whole recording, output every instance of the word \"consent\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 1.25, "consent"]], "response_mode": "offline", "answer": "[{\"word\":\"consent\",\"start\":0.62,\"end\":1.25}]"} {"utterance_id": "700-122867-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0037.flac", "task_type": "lexical_detection", "query": "stacy", "prompt": "Once the audio has finished, identify every instance of the word \"stacy\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.84, 5.39, "stacy"]], "response_mode": "offline", "answer": "[{\"word\":\"stacy\",\"start\":4.84,\"end\":5.39}]"} {"utterance_id": "3663-172528-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0016.flac", "task_type": "lexical_detection", "query": "recovery", "prompt": "Once the audio has finished, identify every instance of the word \"recovery\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.53, 2.26, "recovery"]], "response_mode": "offline", "answer": "[{\"word\":\"recovery\",\"start\":1.53,\"end\":2.26}]"} {"utterance_id": "2506-11267-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0003.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.63, "ostend"]], "response_mode": "offline", "answer": "[{\"word\":\"ostend\",\"start\":1.15,\"end\":1.63}]"} {"utterance_id": "3663-172528-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0054.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.52, "foot"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":1.04,\"end\":1.52}]"} {"utterance_id": "1255-74899-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0003.flac", "task_type": "lexical_detection", "query": "beast", "prompt": "After listening to the whole recording, output every instance of the word \"beast\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.85, 5.38, "beast"]], "response_mode": "offline", "answer": "[{\"word\":\"beast\",\"start\":4.85,\"end\":5.38}]"} {"utterance_id": "6123-59186-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0026.flac", "task_type": "lexical_detection", "query": "sway", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sway\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.83, 4.28, "sway"]], "response_mode": "offline", "answer": "[{\"word\":\"sway\",\"start\":3.83,\"end\":4.28}]"} {"utterance_id": "4323-13259-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0009.flac", "task_type": "lexical_detection", "query": "surface", "prompt": "Process the full audio offline. Find every instance of the word \"surface\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 2.0, "surface"]], "response_mode": "offline", "answer": "[{\"word\":\"surface\",\"start\":1.63,\"end\":2.0}]"} {"utterance_id": "6841-88291-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.77, 15.2, "cattle"]], "response_mode": "offline", "answer": "[{\"word\":\"cattle\",\"start\":14.77,\"end\":15.2}]"} {"utterance_id": "3915-57461-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.91, 5.24, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":4.91,\"end\":5.24}]"} {"utterance_id": "7697-245712-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0010.flac", "task_type": "lexical_detection", "query": "objection", "prompt": "After the audio ends, extract all occurrences of the word \"objection\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 1.14, "objection"]], "response_mode": "offline", "answer": "[{\"word\":\"objection\",\"start\":0.44,\"end\":1.14}]"} {"utterance_id": "8173-294714-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0016.flac", "task_type": "lexical_detection", "query": "debt", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"debt\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.92, 3.35, "debt"]], "response_mode": "offline", "answer": "[{\"word\":\"debt\",\"start\":2.92,\"end\":3.35}]"} {"utterance_id": "6123-59150-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0028.flac", "task_type": "lexical_detection", "query": "appeal", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"appeal\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.01, 2.47, "appeal"]], "response_mode": "offline", "answer": "[{\"word\":\"appeal\",\"start\":2.01,\"end\":2.47}]"} {"utterance_id": "6841-88291-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0030.flac", "task_type": "lexical_detection", "query": "released", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"released\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.27, 3.71, "released"]], "response_mode": "offline", "answer": "[{\"word\":\"released\",\"start\":3.27,\"end\":3.71}]"} {"utterance_id": "6123-59150-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0024.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.71, "author"]], "response_mode": "offline", "answer": "[{\"word\":\"author\",\"start\":3.22,\"end\":3.71}]"} {"utterance_id": "4515-11057-0105", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0105.flac", "task_type": "lexical_detection", "query": "bye", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"bye\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.56, 0.94, "bye"]], "response_mode": "offline", "answer": "[{\"word\":\"bye\",\"start\":0.56,\"end\":0.94}]"} {"utterance_id": "1630-96099-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0020.flac", "task_type": "lexical_detection", "query": "ass", "prompt": "When the utterance is complete, report all instances of the word \"ass\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.96, "ass"]], "response_mode": "offline", "answer": "[{\"word\":\"ass\",\"start\":1.59,\"end\":1.96}]"} {"utterance_id": "116-288045-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.42, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":0.81,\"end\":1.42}]"} {"utterance_id": "1630-102884-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0003.flac", "task_type": "lexical_detection", "query": "atmosphere", "prompt": "Once the audio has finished, identify every instance of the word \"atmosphere\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.47, 8.12, "atmosphere"]], "response_mode": "offline", "answer": "[{\"word\":\"atmosphere\",\"start\":7.47,\"end\":8.12}]"} {"utterance_id": "4153-186223-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0009.flac", "task_type": "lexical_detection", "query": "take", "prompt": "When the utterance is complete, report all instances of the word \"take\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.76, "take"]], "response_mode": "offline", "answer": "[{\"word\":\"take\",\"start\":1.48,\"end\":1.76}]"} {"utterance_id": "700-122866-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0003.flac", "task_type": "lexical_detection", "query": "sensible", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"sensible\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.5, 2.98, "sensible"]], "response_mode": "offline", "answer": "[{\"word\":\"sensible\",\"start\":2.5,\"end\":2.98}]"} {"utterance_id": "8254-84205-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0049.flac", "task_type": "lexical_detection", "query": "hunting", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"hunting\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.85, 2.2, "hunting"]], "response_mode": "offline", "answer": "[{\"word\":\"hunting\",\"start\":1.85,\"end\":2.2}]"} {"utterance_id": "7641-96252-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0022.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.2, 0.52, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.2,\"end\":0.52}]"} {"utterance_id": "4831-29134-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0003.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a profession with word/start/end. If none, reply [].", "labels": [[3.52, 3.98, "one"], [8.62, 9.4, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.52,\"end\":3.98},{\"word\":\"two\",\"start\":8.62,\"end\":9.4}]"} {"utterance_id": "1686-142278-0080", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0080.flac", "task_type": "lexical_detection", "query": "themselves", "prompt": "When the utterance is complete, report all instances of the word \"themselves\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.09, 3.64, "themselves"]], "response_mode": "offline", "answer": "[{\"word\":\"themselves\",\"start\":3.09,\"end\":3.64}]"} {"utterance_id": "6467-62797-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0000.flac", "task_type": "lexical_detection", "query": "curiosity", "prompt": "After listening to the whole recording, output every instance of the word \"curiosity\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.45, "curiosity"]], "response_mode": "offline", "answer": "[{\"word\":\"curiosity\",\"start\":0.65,\"end\":1.45}]"} {"utterance_id": "5849-50873-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0019.flac", "task_type": "lexical_detection", "query": "assembly", "prompt": "Process the full audio offline. Find every instance of the word \"assembly\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.72, 3.29, "assembly"]], "response_mode": "offline", "answer": "[{\"word\":\"assembly\",\"start\":2.72,\"end\":3.29}]"} {"utterance_id": "4323-55228-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0006.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.47, "one"], [3.08, 3.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.26,\"end\":0.47},{\"word\":\"two\",\"start\":3.08,\"end\":3.53}]"} {"utterance_id": "7641-96670-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.38, 0.65, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":0.38,\"end\":0.65}]"} {"utterance_id": "4831-18525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.09, 9.34, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":9.09,\"end\":9.34}]"} {"utterance_id": "5849-50873-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[3.49, 3.91, "charles"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":3.49,\"end\":3.91}]"} {"utterance_id": "7641-96670-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0004.flac", "task_type": "lexical_detection", "query": "jason", "prompt": "After the audio ends, extract all occurrences of the word \"jason\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.22, 0.64, "jason"]], "response_mode": "offline", "answer": "[{\"word\":\"jason\",\"start\":0.22,\"end\":0.64}]"} {"utterance_id": "3915-98647-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0011.flac", "task_type": "lexical_detection", "query": "owner", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"owner\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.05, "owner"]], "response_mode": "offline", "answer": "[{\"word\":\"owner\",\"start\":0.84,\"end\":1.05}]"} {"utterance_id": "1630-141772-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0005.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen to the full audio. After it finishes, find every instance of a disease name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.75, 11.1, "wound"]], "response_mode": "offline", "answer": "[{\"word\":\"wound\",\"start\":10.75,\"end\":11.1}]"} {"utterance_id": "8254-115543-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0035.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.07, 4.26, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":4.07,\"end\":4.26}]"} {"utterance_id": "6467-94831-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0043.flac", "task_type": "lexical_detection", "query": "master", "prompt": "When the utterance is complete, report all instances of the word \"master\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.55, "master"]], "response_mode": "offline", "answer": "[{\"word\":\"master\",\"start\":3.21,\"end\":3.55}]"} {"utterance_id": "4323-55228-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0045.flac", "task_type": "lexical_detection", "query": "billy", "prompt": "Once the audio has finished, identify every instance of the word \"billy\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.49, 0.98, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.49,\"end\":0.98}]"} {"utterance_id": "6599-38591-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0001.flac", "task_type": "lexical_detection", "query": "hath", "prompt": "When the utterance is complete, report all instances of the word \"hath\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.99, 6.16, "hath"]], "response_mode": "offline", "answer": "[{\"word\":\"hath\",\"start\":5.99,\"end\":6.16}]"} {"utterance_id": "3660-172183-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0000.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.52, 0.96, "one"], [2.97, 3.3, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.52,\"end\":0.96},{\"word\":\"two\",\"start\":2.97,\"end\":3.3}]"} {"utterance_id": "4570-14911-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0016.flac", "task_type": "lexical_counting", "query": "smooth", "prompt": "Process the full audio offline. For every occurrence of the word \"smooth\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.74, "one"], [3.22, 3.97, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.27,\"end\":1.74},{\"word\":\"two\",\"start\":3.22,\"end\":3.97}]"} {"utterance_id": "3660-172182-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0018.flac", "task_type": "lexical_detection", "query": "guenever", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"guenever\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.5, 0.92, "guenever"]], "response_mode": "offline", "answer": "[{\"word\":\"guenever\",\"start\":0.5,\"end\":0.92}]"} {"utterance_id": "1585-131718-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0011.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After the audio ends, extract all occurrences of an organization name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.67, 5.03, "company"]], "response_mode": "offline", "answer": "[{\"word\":\"company\",\"start\":4.67,\"end\":5.03}]"} {"utterance_id": "4323-18416-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.98, 2.3, "charlotte"]], "response_mode": "offline", "answer": "[{\"word\":\"charlotte\",\"start\":1.98,\"end\":2.3}]"} {"utterance_id": "4831-25894-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0032.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.46, 9.86, "pigs"]], "response_mode": "offline", "answer": "[{\"word\":\"pigs\",\"start\":9.46,\"end\":9.86}]"} {"utterance_id": "2506-169427-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0001.flac", "task_type": "lexical_detection", "query": "reader", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"reader\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.45, "reader"]], "response_mode": "offline", "answer": "[{\"word\":\"reader\",\"start\":0.85,\"end\":1.45}]"} {"utterance_id": "116-288048-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0005.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Listen carefully to the entire clip, then report every instance of a date as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 2.04, "friday"]], "response_mode": "offline", "answer": "[{\"word\":\"friday\",\"start\":1.45,\"end\":2.04}]"} {"utterance_id": "5849-50963-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0010.flac", "task_type": "lexical_detection", "query": "lifetime", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"lifetime\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.98, 2.53, "lifetime"]], "response_mode": "offline", "answer": "[{\"word\":\"lifetime\",\"start\":1.98,\"end\":2.53}]"} {"utterance_id": "2506-13150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.66, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":0.84,\"end\":1.66}]"} {"utterance_id": "6455-67804-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0032.flac", "task_type": "lexical_detection", "query": "contrary", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"contrary\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.28, 1.79, "contrary"]], "response_mode": "offline", "answer": "[{\"word\":\"contrary\",\"start\":1.28,\"end\":1.79}]"} {"utterance_id": "6841-88294-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0040.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.17, 3.73, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.17,\"end\":3.73}]"} {"utterance_id": "1701-141760-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0044.flac", "task_type": "lexical_detection", "query": "allowance", "prompt": "When the utterance is complete, report all instances of the word \"allowance\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.53, 5.96, "allowance"]], "response_mode": "offline", "answer": "[{\"word\":\"allowance\",\"start\":5.53,\"end\":5.96}]"} {"utterance_id": "4570-56594-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0015.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.6, "salt"]], "response_mode": "offline", "answer": "[{\"word\":\"salt\",\"start\":2.26,\"end\":2.6}]"} {"utterance_id": "8288-274162-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0046.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.29, "carpenter"]], "response_mode": "offline", "answer": "[{\"word\":\"carpenter\",\"start\":0.6,\"end\":1.29}]"} {"utterance_id": "1686-142278-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 1.03, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.51,\"end\":1.03}]"} {"utterance_id": "6455-67804-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0027.flac", "task_type": "lexical_detection", "query": "don't", "prompt": "After listening to the whole recording, output every instance of the word \"don't\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.59, "don't"]], "response_mode": "offline", "answer": "[{\"word\":\"don't\",\"start\":2.39,\"end\":2.59}]"} {"utterance_id": "7641-96670-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0008.flac", "task_type": "lexical_detection", "query": "medea", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"medea\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.35, 4.84, "medea"]], "response_mode": "offline", "answer": "[{\"word\":\"medea\",\"start\":4.35,\"end\":4.84}]"} {"utterance_id": "8254-84205-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.74, 6.12, "pony"]], "response_mode": "offline", "answer": "[{\"word\":\"pony\",\"start\":5.74,\"end\":6.12}]"} {"utterance_id": "7641-96684-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0014.flac", "task_type": "lexical_detection", "query": "spoke", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"spoke\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.6, "spoke"]], "response_mode": "offline", "answer": "[{\"word\":\"spoke\",\"start\":1.27,\"end\":1.6}]"} {"utterance_id": "1630-73710-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0009.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.75, "pony"]], "response_mode": "offline", "answer": "[{\"word\":\"pony\",\"start\":1.18,\"end\":1.75}]"} {"utterance_id": "3660-172182-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0002.flac", "task_type": "lexical_detection", "query": "horn", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"horn\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.09, "horn"]], "response_mode": "offline", "answer": "[{\"word\":\"horn\",\"start\":1.77,\"end\":2.09}]"} {"utterance_id": "1255-90407-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0025.flac", "task_type": "lexical_detection", "query": "hungry", "prompt": "After the audio ends, extract all occurrences of the word \"hungry\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.79, "hungry"]], "response_mode": "offline", "answer": "[{\"word\":\"hungry\",\"start\":1.47,\"end\":1.79}]"} {"utterance_id": "4570-102353-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0000.flac", "task_type": "lexical_detection", "query": "accurate", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"accurate\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.96, 3.62, "accurate"]], "response_mode": "offline", "answer": "[{\"word\":\"accurate\",\"start\":2.96,\"end\":3.62}]"} {"utterance_id": "6599-38591-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0010.flac", "task_type": "lexical_detection", "query": "francis", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"francis\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.87, "francis"]], "response_mode": "offline", "answer": "[{\"word\":\"francis\",\"start\":0.4,\"end\":0.87}]"} {"utterance_id": "5543-27761-0094", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0094.flac", "task_type": "lexical_detection", "query": "confused", "prompt": "After listening to the whole recording, output every instance of the word \"confused\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 1.37, "confused"]], "response_mode": "offline", "answer": "[{\"word\":\"confused\",\"start\":0.66,\"end\":1.37}]"} {"utterance_id": "7641-96684-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0011.flac", "task_type": "lexical_detection", "query": "willingly", "prompt": "After listening to the whole recording, output every instance of the word \"willingly\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.51, "willingly"]], "response_mode": "offline", "answer": "[{\"word\":\"willingly\",\"start\":0.95,\"end\":1.51}]"} {"utterance_id": "1255-90407-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0015.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.54, 0.96, "one"], [5.08, 5.4, "two"], [6.83, 7.49, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.54,\"end\":0.96},{\"word\":\"two\",\"start\":5.08,\"end\":5.4},{\"word\":\"three\",\"start\":6.83,\"end\":7.49}]"} {"utterance_id": "4323-55228-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0028.flac", "task_type": "lexical_detection", "query": "stirred", "prompt": "After the audio ends, extract all occurrences of the word \"stirred\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.72, "stirred"]], "response_mode": "offline", "answer": "[{\"word\":\"stirred\",\"start\":1.37,\"end\":1.72}]"} {"utterance_id": "7641-96670-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0018.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.28, 3.62, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":3.28,\"end\":3.62}]"} {"utterance_id": "6467-97061-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0016.flac", "task_type": "lexical_detection", "query": "chain", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"chain\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.72, 5.12, "chain"]], "response_mode": "offline", "answer": "[{\"word\":\"chain\",\"start\":4.72,\"end\":5.12}]"} {"utterance_id": "4153-61735-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0002.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "When the utterance is complete, report how a family relation accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.62, "one"], [2.76, 3.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.16,\"end\":1.62},{\"word\":\"two\",\"start\":2.76,\"end\":3.42}]"} {"utterance_id": "1701-141759-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.7, "petya"]], "response_mode": "offline", "answer": "[{\"word\":\"petya\",\"start\":1.2,\"end\":1.7}]"} {"utterance_id": "4323-13259-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0010.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.37, 6.69, "whiskey"]], "response_mode": "offline", "answer": "[{\"word\":\"whiskey\",\"start\":6.37,\"end\":6.69}]"} {"utterance_id": "6123-59150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0000.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Do not answer until the audio ends. Then list each instance of a disease name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.37, 4.43, "epileptic fits"]], "response_mode": "offline", "answer": "[{\"word\":\"epileptic fits\",\"start\":3.37,\"end\":4.43}]"} {"utterance_id": "6123-59186-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0002.flac", "task_type": "lexical_detection", "query": "sufficiently", "prompt": "After listening to the whole recording, output every instance of the word \"sufficiently\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 3.09, "sufficiently"], [4.45, 5.02, "sufficiently"]], "response_mode": "offline", "answer": "[{\"word\":\"sufficiently\",\"start\":2.39,\"end\":3.09},{\"word\":\"sufficiently\",\"start\":4.45,\"end\":5.02}]"} {"utterance_id": "1701-141759-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0025.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Process the full audio offline. For every occurrence of a body part, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.11, 4.52, "one"], [10.7, 11.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.11,\"end\":4.52},{\"word\":\"two\",\"start\":10.7,\"end\":11.18}]"} {"utterance_id": "8288-274162-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0052.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.7, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":2.31,\"end\":2.7}]"} {"utterance_id": "4831-18525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.99, "jasper"], [6.1, 6.38, "joel's"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":1.56,\"end\":1.99},{\"word\":\"joel's\",\"start\":6.1,\"end\":6.38}]"} {"utterance_id": "6455-67803-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0031.flac", "task_type": "lexical_detection", "query": "start", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"start\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.36, 6.75, "start"]], "response_mode": "offline", "answer": "[{\"word\":\"start\",\"start\":6.36,\"end\":6.75}]"} {"utterance_id": "5543-27761-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0022.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[2.89, 3.26, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"lips\",\"start\":2.89,\"end\":3.26}]"} {"utterance_id": "4323-55228-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0020.flac", "task_type": "lexical_detection", "query": "pulled", "prompt": "After listening to the whole recording, output every instance of the word \"pulled\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.52, "pulled"]], "response_mode": "offline", "answer": "[{\"word\":\"pulled\",\"start\":0.33,\"end\":0.52}]"} {"utterance_id": "4515-11057-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0026.flac", "task_type": "lexical_detection", "query": "achieve", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"achieve\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.63, 7.21, "achieve"]], "response_mode": "offline", "answer": "[{\"word\":\"achieve\",\"start\":6.63,\"end\":7.21}]"} {"utterance_id": "3660-172182-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0004.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.62, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":3.22,\"end\":3.62}]"} {"utterance_id": "8254-84205-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0065.flac", "task_type": "lexical_detection", "query": "hit", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hit\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 2.82, "hit"]], "response_mode": "offline", "answer": "[{\"word\":\"hit\",\"start\":2.69,\"end\":2.82}]"} {"utterance_id": "7697-105815-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0037.flac", "task_type": "lexical_detection", "query": "chains", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"chains\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.81, 4.33, "chains"]], "response_mode": "offline", "answer": "[{\"word\":\"chains\",\"start\":3.81,\"end\":4.33}]"} {"utterance_id": "1585-131718-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0052.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "When the utterance is complete, report all instances of a state or province name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.42, 3.9, "florida"]], "response_mode": "offline", "answer": "[{\"word\":\"florida\",\"start\":3.42,\"end\":3.9}]"} {"utterance_id": "7641-96684-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0025.flac", "task_type": "lexical_detection", "query": "speech", "prompt": "When the utterance is complete, report all instances of the word \"speech\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.18, "speech"]], "response_mode": "offline", "answer": "[{\"word\":\"speech\",\"start\":0.8,\"end\":1.18}]"} {"utterance_id": "3663-172528-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0052.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Process the full audio offline. For every occurrence of a number, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.46, 8.78, "one"], [9.9, 10.35, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.46,\"end\":8.78},{\"word\":\"two\",\"start\":9.9,\"end\":10.35}]"} {"utterance_id": "8288-274162-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0058.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.62, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":2.25,\"end\":2.62}]"} {"utterance_id": "6599-38590-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0008.flac", "task_type": "lexical_detection", "query": "cambrai", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"cambrai\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.44, "cambrai"]], "response_mode": "offline", "answer": "[{\"word\":\"cambrai\",\"start\":0.99,\"end\":1.44}]"} {"utterance_id": "700-122867-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0022.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.76, 4.2, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":3.76,\"end\":4.2}]"} {"utterance_id": "4323-18416-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0002.flac", "task_type": "semantic_counting", "query": "color", "prompt": "Listen to the entire clip, then report the count sequence for a color name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.67, "one"], [2.07, 2.55, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.45,\"end\":0.67},{\"word\":\"two\",\"start\":2.07,\"end\":2.55}]"} {"utterance_id": "1686-142278-0093", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0093.flac", "task_type": "lexical_detection", "query": "doubting", "prompt": "After the audio ends, extract all occurrences of the word \"doubting\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.48, "doubting"]], "response_mode": "offline", "answer": "[{\"word\":\"doubting\",\"start\":2.06,\"end\":2.48}]"} {"utterance_id": "116-288048-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0010.flac", "task_type": "lexical_detection", "query": "immediate", "prompt": "Process the full audio offline. Find every instance of the word \"immediate\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.91, "immediate"]], "response_mode": "offline", "answer": "[{\"word\":\"immediate\",\"start\":0.51,\"end\":0.91}]"} {"utterance_id": "8288-274162-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0017.flac", "task_type": "lexical_detection", "query": "astonished", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"astonished\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 2.23, "astonished"]], "response_mode": "offline", "answer": "[{\"word\":\"astonished\",\"start\":1.58,\"end\":2.23}]"} {"utterance_id": "6841-88291-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0045.flac", "task_type": "lexical_detection", "query": "alone", "prompt": "After listening to the whole recording, output every instance of the word \"alone\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.84, "alone"]], "response_mode": "offline", "answer": "[{\"word\":\"alone\",\"start\":2.38,\"end\":2.84}]"} {"utterance_id": "1650-173551-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0005.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen carefully to the entire clip, then report every instance of a plant name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.99, 6.33, "bush"]], "response_mode": "offline", "answer": "[{\"word\":\"bush\",\"start\":5.99,\"end\":6.33}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "lexical_counting", "query": "polly", "prompt": "Listen to the full audio. Count every occurrence of the word \"polly\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.63, 4.01, "one"], [10.7, 11.0, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.63,\"end\":4.01},{\"word\":\"two\",\"start\":10.7,\"end\":11.0}]"} {"utterance_id": "5543-27761-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0037.flac", "task_type": "lexical_detection", "query": "mistress", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"mistress\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.39, "mistress"]], "response_mode": "offline", "answer": "[{\"word\":\"mistress\",\"start\":0.79,\"end\":1.39}]"} {"utterance_id": "1686-142278-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.08, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.5,\"end\":1.08}]"} {"utterance_id": "4323-55228-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0014.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a color name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.84, 3.06, "gray"]], "response_mode": "offline", "answer": "[{\"word\":\"gray\",\"start\":2.84,\"end\":3.06}]"} {"utterance_id": "700-122867-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0024.flac", "task_type": "lexical_detection", "query": "scissors", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"scissors\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.98, 2.53, "scissors"]], "response_mode": "offline", "answer": "[{\"word\":\"scissors\",\"start\":1.98,\"end\":2.53}]"} {"utterance_id": "3663-172528-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0005.flac", "task_type": "lexical_detection", "query": "grieved", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"grieved\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.53, 5.99, "grieved"]], "response_mode": "offline", "answer": "[{\"word\":\"grieved\",\"start\":5.53,\"end\":5.99}]"} {"utterance_id": "4323-55228-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0022.flac", "task_type": "lexical_detection", "query": "kate", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"kate\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.44, "kate"]], "response_mode": "offline", "answer": "[{\"word\":\"kate\",\"start\":1.19,\"end\":1.44}]"} {"utterance_id": "6841-88291-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0029.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[2.9, 3.11, "ear"]], "response_mode": "offline", "answer": "[{\"word\":\"ear\",\"start\":2.9,\"end\":3.11}]"} {"utterance_id": "6841-88291-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0028.flac", "task_type": "lexical_detection", "query": "doctrine", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"doctrine\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.55, 2.93, "doctrine"]], "response_mode": "offline", "answer": "[{\"word\":\"doctrine\",\"start\":2.55,\"end\":2.93}]"} {"utterance_id": "1686-142278-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0064.flac", "task_type": "lexical_detection", "query": "income", "prompt": "After the audio ends, extract all occurrences of the word \"income\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.64, 3.08, "income"]], "response_mode": "offline", "answer": "[{\"word\":\"income\",\"start\":2.64,\"end\":3.08}]"} {"utterance_id": "4323-55228-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0043.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.32, 1.64, "one"], [3.12, 3.52, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.32,\"end\":1.64},{\"word\":\"two\",\"start\":3.12,\"end\":3.52}]"} {"utterance_id": "8288-274162-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0057.flac", "task_type": "lexical_detection", "query": "herself", "prompt": "After the audio ends, extract all occurrences of the word \"herself\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.7, 6.17, "herself"]], "response_mode": "offline", "answer": "[{\"word\":\"herself\",\"start\":5.7,\"end\":6.17}]"} {"utterance_id": "7601-175351-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0012.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 3.07, "plover"]], "response_mode": "offline", "answer": "[{\"word\":\"plover\",\"start\":2.39,\"end\":3.07}]"} {"utterance_id": "116-288048-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.37, 2.69, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":2.37,\"end\":2.69}]"} {"utterance_id": "5543-27761-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0009.flac", "task_type": "lexical_detection", "query": "wife", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wife\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 2.01, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.71,\"end\":2.01}]"} {"utterance_id": "8254-84205-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0059.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.79, "water"]], "response_mode": "offline", "answer": "[{\"word\":\"water\",\"start\":1.44,\"end\":1.79}]"} {"utterance_id": "4570-102353-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0007.flac", "task_type": "lexical_detection", "query": "inform", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"inform\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.05, 4.45, "inform"]], "response_mode": "offline", "answer": "[{\"word\":\"inform\",\"start\":4.05,\"end\":4.45}]"} {"utterance_id": "6455-66379-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0014.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.97, 3.51, "salesman"]], "response_mode": "offline", "answer": "[{\"word\":\"salesman\",\"start\":2.97,\"end\":3.51}]"} {"utterance_id": "8173-294714-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0029.flac", "task_type": "lexical_detection", "query": "forgiven", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"forgiven\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 2.1, "forgiven"]], "response_mode": "offline", "answer": "[{\"word\":\"forgiven\",\"start\":1.4,\"end\":2.1}]"} {"utterance_id": "7641-96684-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0006.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.24, 2.65, "one"], [4.34, 4.9, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.24,\"end\":2.65},{\"word\":\"two\",\"start\":4.34,\"end\":4.9}]"} {"utterance_id": "116-288045-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0000.flac", "task_type": "lexical_detection", "query": "ringing", "prompt": "Process the full audio offline. Find every instance of the word \"ringing\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.68, 3.07, "ringing"]], "response_mode": "offline", "answer": "[{\"word\":\"ringing\",\"start\":2.68,\"end\":3.07}]"} {"utterance_id": "4570-56594-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0000.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen to the full audio. After it finishes, find every instance of an organization name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.69, "corps"]], "response_mode": "offline", "answer": "[{\"word\":\"corps\",\"start\":1.21,\"end\":1.69}]"} {"utterance_id": "6467-56885-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0003.flac", "task_type": "lexical_detection", "query": "birds", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"birds\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.57, "birds"]], "response_mode": "offline", "answer": "[{\"word\":\"birds\",\"start\":1.1,\"end\":1.57}]"} {"utterance_id": "4570-56594-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.2, 5.91, "indiany"]], "response_mode": "offline", "answer": "[{\"word\":\"indiany\",\"start\":5.2,\"end\":5.91}]"} {"utterance_id": "4323-55228-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0008.flac", "task_type": "lexical_detection", "query": "it'll", "prompt": "When the utterance is complete, report all instances of the word \"it'll\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.09, 0.39, "it'll"]], "response_mode": "offline", "answer": "[{\"word\":\"it'll\",\"start\":0.09,\"end\":0.39}]"} {"utterance_id": "1686-142278-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0026.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Process the full audio offline. Find every instance of an organization name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.15, "church"]], "response_mode": "offline", "answer": "[{\"word\":\"church\",\"start\":1.83,\"end\":2.15}]"} {"utterance_id": "8254-115543-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0016.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the full audio. After it finishes, find every instance of a plant name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.94, 6.54, "sugar cane"]], "response_mode": "offline", "answer": "[{\"word\":\"sugar cane\",\"start\":5.94,\"end\":6.54}]"} {"utterance_id": "1255-90413-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 2.97, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":2.46,\"end\":2.97}]"} {"utterance_id": "7601-291468-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0006.flac", "task_type": "lexical_detection", "query": "dutch", "prompt": "After listening to the whole recording, output every instance of the word \"dutch\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.91, 10.42, "dutch"], [19.84, 20.15, "dutch"]], "response_mode": "offline", "answer": "[{\"word\":\"dutch\",\"start\":9.91,\"end\":10.42},{\"word\":\"dutch\",\"start\":19.84,\"end\":20.15}]"} {"utterance_id": "3915-98647-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0036.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.41, 8.79, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":8.41,\"end\":8.79}]"} {"utterance_id": "1651-136854-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0021.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.47, "wife"], [4.28, 4.79, "children"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.04,\"end\":1.47},{\"word\":\"children\",\"start\":4.28,\"end\":4.79}]"} {"utterance_id": "3663-172528-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0000.flac", "task_type": "lexical_detection", "query": "thirst", "prompt": "Once the audio has finished, identify every instance of the word \"thirst\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.54, 2.07, "thirst"]], "response_mode": "offline", "answer": "[{\"word\":\"thirst\",\"start\":1.54,\"end\":2.07}]"} {"utterance_id": "4831-25894-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0031.flac", "task_type": "lexical_detection", "query": "screamed", "prompt": "After listening to the whole recording, output every instance of the word \"screamed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.18, "screamed"]], "response_mode": "offline", "answer": "[{\"word\":\"screamed\",\"start\":0.72,\"end\":1.18}]"} {"utterance_id": "5849-50873-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0037.flac", "task_type": "lexical_detection", "query": "afterward", "prompt": "Once the audio has finished, identify every instance of the word \"afterward\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.06, 4.59, "afterward"]], "response_mode": "offline", "answer": "[{\"word\":\"afterward\",\"start\":4.06,\"end\":4.59}]"} {"utterance_id": "116-288048-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0013.flac", "task_type": "lexical_detection", "query": "absolutely", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"absolutely\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.28, "absolutely"]], "response_mode": "offline", "answer": "[{\"word\":\"absolutely\",\"start\":0.83,\"end\":1.28}]"} {"utterance_id": "4323-55228-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.5, "billy"], [4.44, 4.97, "aunt hannah's"], [6.69, 7.01, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.23,\"end\":0.5},{\"word\":\"aunt hannah's\",\"start\":4.44,\"end\":4.97},{\"word\":\"billy\",\"start\":6.69,\"end\":7.01}]"} {"utterance_id": "6841-88294-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0068.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 0.57, "third"], [2.32, 2.58, "third"]], "response_mode": "offline", "answer": "[{\"word\":\"third\",\"start\":0.36,\"end\":0.57},{\"word\":\"third\",\"start\":2.32,\"end\":2.58}]"} {"utterance_id": "6455-67804-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0019.flac", "task_type": "lexical_detection", "query": "perhaps", "prompt": "Process the full audio offline. Find every instance of the word \"perhaps\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.15, 0.67, "perhaps"]], "response_mode": "offline", "answer": "[{\"word\":\"perhaps\",\"start\":0.15,\"end\":0.67}]"} {"utterance_id": "1585-131718-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0013.flac", "task_type": "lexical_detection", "query": "opposition", "prompt": "When the utterance is complete, report all instances of the word \"opposition\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.29, 1.99, "opposition"]], "response_mode": "offline", "answer": "[{\"word\":\"opposition\",\"start\":1.29,\"end\":1.99}]"} {"utterance_id": "1650-167613-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0030.flac", "task_type": "lexical_detection", "query": "it's", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"it's\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.11, 0.36, "it's"]], "response_mode": "offline", "answer": "[{\"word\":\"it's\",\"start\":0.11,\"end\":0.36}]"} {"utterance_id": "1255-90407-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.96, "jocelyn"], [5.08, 5.4, "montagues"], [6.83, 7.49, "capulets"]], "response_mode": "offline", "answer": "[{\"word\":\"jocelyn\",\"start\":0.54,\"end\":0.96},{\"word\":\"montagues\",\"start\":5.08,\"end\":5.4},{\"word\":\"capulets\",\"start\":6.83,\"end\":7.49}]"} {"utterance_id": "5849-50962-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.37, "george bingham"]], "response_mode": "offline", "answer": "[{\"word\":\"george bingham\",\"start\":0.77,\"end\":1.37}]"} {"utterance_id": "1585-131718-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0031.flac", "task_type": "lexical_detection", "query": "eighty", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"eighty\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.02, "eighty"]], "response_mode": "offline", "answer": "[{\"word\":\"eighty\",\"start\":1.84,\"end\":2.02}]"} {"utterance_id": "4831-18525-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0009.flac", "task_type": "lexical_detection", "query": "everybody", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"everybody\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.71, "everybody"]], "response_mode": "offline", "answer": "[{\"word\":\"everybody\",\"start\":2.27,\"end\":2.71}]"} {"utterance_id": "7601-175351-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0017.flac", "task_type": "lexical_detection", "query": "ought", "prompt": "Process the full audio offline. Find every instance of the word \"ought\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.77, "ought"]], "response_mode": "offline", "answer": "[{\"word\":\"ought\",\"start\":0.33,\"end\":0.77}]"} {"utterance_id": "6267-65525-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0009.flac", "task_type": "semantic_counting", "query": "food", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a food name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[5.31, 5.82, "one"], [8.8, 9.2, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.31,\"end\":5.82},{\"word\":\"two\",\"start\":8.8,\"end\":9.2}]"} {"utterance_id": "6467-94831-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0044.flac", "task_type": "lexical_detection", "query": "peleg", "prompt": "When the utterance is complete, report all instances of the word \"peleg\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.68, "peleg"]], "response_mode": "offline", "answer": "[{\"word\":\"peleg\",\"start\":0.31,\"end\":0.68}]"} {"utterance_id": "2506-169427-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0000.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Once the audio has finished, identify every instance of a money or currency term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.2, 0.76, "interest"], [5.88, 6.33, "payments"], [17.25, 17.84, "principal"], [20.07, 20.5, "interest"], [24.24, 25.28, "obligations"]], "response_mode": "offline", "answer": "[{\"word\":\"interest\",\"start\":0.2,\"end\":0.76},{\"word\":\"payments\",\"start\":5.88,\"end\":6.33},{\"word\":\"principal\",\"start\":17.25,\"end\":17.84},{\"word\":\"interest\",\"start\":20.07,\"end\":20.5},{\"word\":\"obligations\",\"start\":24.24,\"end\":25.28}]"} {"utterance_id": "116-288048-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0025.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.9, 10.59, "floods"]], "response_mode": "offline", "answer": "[{\"word\":\"floods\",\"start\":9.9,\"end\":10.59}]"} {"utterance_id": "116-288045-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0027.flac", "task_type": "lexical_detection", "query": "morning", "prompt": "After the audio ends, extract all occurrences of the word \"morning\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.77, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.37,\"end\":1.77}]"} {"utterance_id": "3660-6517-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0034.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.55, 2.88, "evans"]], "response_mode": "offline", "answer": "[{\"word\":\"evans\",\"start\":2.55,\"end\":2.88}]"} {"utterance_id": "4153-186222-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0001.flac", "task_type": "lexical_detection", "query": "why", "prompt": "Once the audio has finished, identify every instance of the word \"why\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.34, 1.52, "why"]], "response_mode": "offline", "answer": "[{\"word\":\"why\",\"start\":1.34,\"end\":1.52}]"} {"utterance_id": "700-122867-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.36, 5.87, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":5.36,\"end\":5.87}]"} {"utterance_id": "6455-66379-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0006.flac", "task_type": "lexical_detection", "query": "playing", "prompt": "When the utterance is complete, report all instances of the word \"playing\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.55, "playing"]], "response_mode": "offline", "answer": "[{\"word\":\"playing\",\"start\":2.18,\"end\":2.55}]"} {"utterance_id": "1585-157660-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0000.flac", "task_type": "lexical_detection", "query": "london", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"london\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.32, 0.85, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":0.32,\"end\":0.85}]"} {"utterance_id": "1585-131718-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0010.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.13, 2.47, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":2.13,\"end\":2.47}]"} {"utterance_id": "6467-56885-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0014.flac", "task_type": "lexical_detection", "query": "mad", "prompt": "Process the full audio offline. Find every instance of the word \"mad\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 0.87, "mad"]], "response_mode": "offline", "answer": "[{\"word\":\"mad\",\"start\":0.56,\"end\":0.87}]"} {"utterance_id": "5543-27761-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0070.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Once the audio has finished, identify every instance of a color name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.53, 1.77, "blue"]], "response_mode": "offline", "answer": "[{\"word\":\"blue\",\"start\":1.53,\"end\":1.77}]"} {"utterance_id": "1686-142278-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0003.flac", "task_type": "lexical_detection", "query": "unhappy", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"unhappy\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.69, 2.23, "unhappy"]], "response_mode": "offline", "answer": "[{\"word\":\"unhappy\",\"start\":1.69,\"end\":2.23}]"} {"utterance_id": "6841-88294-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0063.flac", "task_type": "semantic_counting", "query": "money", "prompt": "Once the audio has finished, list each occurrence of a money or currency term as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.45, 0.75, "one"], [2.07, 2.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.45,\"end\":0.75},{\"word\":\"two\",\"start\":2.07,\"end\":2.36}]"} {"utterance_id": "4831-25894-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0021.flac", "task_type": "lexical_detection", "query": "we'll", "prompt": "Once the audio has finished, identify every instance of the word \"we'll\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.31, 0.46, "we'll"]], "response_mode": "offline", "answer": "[{\"word\":\"we'll\",\"start\":0.31,\"end\":0.46}]"} {"utterance_id": "7641-96670-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.47, "face"], [3.64, 4.03, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":1.16,\"end\":1.47},{\"word\":\"hair\",\"start\":3.64,\"end\":4.03}]"} {"utterance_id": "6123-59186-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0039.flac", "task_type": "lexical_detection", "query": "experience", "prompt": "When the utterance is complete, report all instances of the word \"experience\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.77, "experience"]], "response_mode": "offline", "answer": "[{\"word\":\"experience\",\"start\":2.11,\"end\":2.77}]"} {"utterance_id": "1686-142278-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0017.flac", "task_type": "lexical_detection", "query": "speech", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"speech\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.54, "speech"]], "response_mode": "offline", "answer": "[{\"word\":\"speech\",\"start\":3.1,\"end\":3.54}]"} {"utterance_id": "2506-11278-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0012.flac", "task_type": "lexical_detection", "query": "victuals", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"victuals\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.41, 5.06, "victuals"]], "response_mode": "offline", "answer": "[{\"word\":\"victuals\",\"start\":4.41,\"end\":5.06}]"} {"utterance_id": "6267-65525-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0053.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.2, 1.01, "john lincoln"]], "response_mode": "offline", "answer": "[{\"word\":\"john lincoln\",\"start\":0.2,\"end\":1.01}]"} {"utterance_id": "7641-96670-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0005.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[2.68, 3.07, "one"], [4.68, 5.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.68,\"end\":3.07},{\"word\":\"two\",\"start\":4.68,\"end\":5.42}]"} {"utterance_id": "6123-59186-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0029.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 2.08, "hearts"]], "response_mode": "offline", "answer": "[{\"word\":\"hearts\",\"start\":1.53,\"end\":2.08}]"} {"utterance_id": "1686-142278-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0036.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[1.01, 1.24, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.01,\"end\":1.24}]"} {"utterance_id": "3660-6517-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0007.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 2.9, "ponies"]], "response_mode": "offline", "answer": "[{\"word\":\"ponies\",\"start\":2.46,\"end\":2.9}]"} {"utterance_id": "6841-88294-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0034.flac", "task_type": "lexical_detection", "query": "off", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"off\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 1.93, "off"]], "response_mode": "offline", "answer": "[{\"word\":\"off\",\"start\":1.68,\"end\":1.93}]"} {"utterance_id": "1255-90407-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0007.flac", "task_type": "lexical_detection", "query": "castle", "prompt": "Process the full audio offline. Find every instance of the word \"castle\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.77, "castle"]], "response_mode": "offline", "answer": "[{\"word\":\"castle\",\"start\":2.14,\"end\":2.77}]"} {"utterance_id": "4153-185072-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0001.flac", "task_type": "lexical_detection", "query": "enraged", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"enraged\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.34, "enraged"]], "response_mode": "offline", "answer": "[{\"word\":\"enraged\",\"start\":0.79,\"end\":1.34}]"} {"utterance_id": "8254-84205-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0052.flac", "task_type": "lexical_detection", "query": "evidently", "prompt": "After the audio ends, extract all occurrences of the word \"evidently\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.25, "evidently"]], "response_mode": "offline", "answer": "[{\"word\":\"evidently\",\"start\":2.76,\"end\":3.25}]"} {"utterance_id": "1651-136854-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0026.flac", "task_type": "lexical_counting", "query": "death", "prompt": "After the audio ends, tally every instance of the word \"death\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[5.83, 6.2, "one"], [12.0, 12.34, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.83,\"end\":6.2},{\"word\":\"two\",\"start\":12.0,\"end\":12.34}]"} {"utterance_id": "6841-88294-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0036.flac", "task_type": "lexical_detection", "query": "mind", "prompt": "Process the full audio offline. Find every instance of the word \"mind\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.69, "mind"]], "response_mode": "offline", "answer": "[{\"word\":\"mind\",\"start\":2.24,\"end\":2.69}]"} {"utterance_id": "2506-11267-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.53, 1.44, "lady marchioness"], [8.67, 9.24, "ladyship"]], "response_mode": "offline", "answer": "[{\"word\":\"lady marchioness\",\"start\":0.53,\"end\":1.44},{\"word\":\"ladyship\",\"start\":8.67,\"end\":9.24}]"} {"utterance_id": "6841-88291-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0004.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "When the utterance is complete, report all instances of a plant name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.68, "grass"]], "response_mode": "offline", "answer": "[{\"word\":\"grass\",\"start\":2.18,\"end\":2.68}]"} {"utterance_id": "3915-98647-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0014.flac", "task_type": "lexical_detection", "query": "style", "prompt": "After the audio ends, extract all occurrences of the word \"style\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.54, 7.88, "style"]], "response_mode": "offline", "answer": "[{\"word\":\"style\",\"start\":7.54,\"end\":7.88}]"} {"utterance_id": "3915-57461-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.56, 6.8, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.56,\"end\":6.8}]"} {"utterance_id": "700-122866-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0001.flac", "task_type": "lexical_detection", "query": "while", "prompt": "When the utterance is complete, report all instances of the word \"while\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.88, "while"]], "response_mode": "offline", "answer": "[{\"word\":\"while\",\"start\":1.46,\"end\":1.88}]"} {"utterance_id": "1650-167613-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0031.flac", "task_type": "lexical_detection", "query": "silent", "prompt": "Process the full audio offline. Find every instance of the word \"silent\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.86, "silent"]], "response_mode": "offline", "answer": "[{\"word\":\"silent\",\"start\":1.17,\"end\":1.86}]"} {"utterance_id": "7641-96684-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.31, "admetus"], [4.26, 4.91, "alcestis"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":0.76,\"end\":1.31},{\"word\":\"alcestis\",\"start\":4.26,\"end\":4.91}]"} {"utterance_id": "6123-59150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0008.flac", "task_type": "lexical_detection", "query": "cries", "prompt": "When the utterance is complete, report all instances of the word \"cries\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.87, "cries"]], "response_mode": "offline", "answer": "[{\"word\":\"cries\",\"start\":1.59,\"end\":1.87}]"} {"utterance_id": "3660-172183-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0014.flac", "task_type": "lexical_detection", "query": "say", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"say\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.52, 0.8, "say"]], "response_mode": "offline", "answer": "[{\"word\":\"say\",\"start\":0.52,\"end\":0.8}]"} {"utterance_id": "6841-88294-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0010.flac", "task_type": "lexical_counting", "query": "hair", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"hair\" (word, start, end). If none, reply [].", "labels": [[1.01, 1.3, "one"], [12.4, 12.69, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.01,\"end\":1.3},{\"word\":\"two\",\"start\":12.4,\"end\":12.69}]"} {"utterance_id": "6123-59150-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0004.flac", "task_type": "lexical_detection", "query": "cruel", "prompt": "After the audio ends, extract all occurrences of the word \"cruel\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 1.9, "cruel"]], "response_mode": "offline", "answer": "[{\"word\":\"cruel\",\"start\":1.42,\"end\":1.9}]"} {"utterance_id": "700-122868-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0039.flac", "task_type": "lexical_detection", "query": "marilla", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"marilla\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.36, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":1.97,\"end\":2.36}]"} {"utterance_id": "5543-27761-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0024.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[4.51, 4.81, "one"], [9.44, 9.75, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.51,\"end\":4.81},{\"word\":\"two\",\"start\":9.44,\"end\":9.75}]"} {"utterance_id": "4570-14911-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0010.flac", "task_type": "lexical_detection", "query": "brown", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"brown\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.94, 2.46, "brown"]], "response_mode": "offline", "answer": "[{\"word\":\"brown\",\"start\":1.94,\"end\":2.46}]"} {"utterance_id": "8254-84205-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.08, 4.39, "heels"]], "response_mode": "offline", "answer": "[{\"word\":\"heels\",\"start\":4.08,\"end\":4.39}]"} {"utterance_id": "5849-50873-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0020.flac", "task_type": "lexical_detection", "query": "means", "prompt": "Process the full audio offline. Find every instance of the word \"means\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.29, "means"]], "response_mode": "offline", "answer": "[{\"word\":\"means\",\"start\":0.87,\"end\":1.29}]"} {"utterance_id": "8254-84205-0072", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0072.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.47, "horses"], [3.54, 4.0, "horses"]], "response_mode": "offline", "answer": "[{\"word\":\"horses\",\"start\":2.01,\"end\":2.47},{\"word\":\"horses\",\"start\":3.54,\"end\":4.0}]"} {"utterance_id": "6841-88294-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0026.flac", "task_type": "lexical_detection", "query": "satisfy", "prompt": "After listening to the whole recording, output every instance of the word \"satisfy\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.44, "satisfy"]], "response_mode": "offline", "answer": "[{\"word\":\"satisfy\",\"start\":0.86,\"end\":1.44}]"} {"utterance_id": "7641-96684-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0001.flac", "task_type": "lexical_detection", "query": "admetus", "prompt": "When the utterance is complete, report all instances of the word \"admetus\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.5, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":1.95,\"end\":2.5}]"} {"utterance_id": "3660-172182-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0004.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.37, 8.69, "knight"]], "response_mode": "offline", "answer": "[{\"word\":\"knight\",\"start\":8.37,\"end\":8.69}]"} {"utterance_id": "6123-59150-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0037.flac", "task_type": "lexical_detection", "query": "attitude", "prompt": "When the utterance is complete, report all instances of the word \"attitude\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.5, "attitude"]], "response_mode": "offline", "answer": "[{\"word\":\"attitude\",\"start\":1.02,\"end\":1.5}]"} {"utterance_id": "4153-186223-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0002.flac", "task_type": "lexical_detection", "query": "riding", "prompt": "After listening to the whole recording, output every instance of the word \"riding\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.66, "riding"]], "response_mode": "offline", "answer": "[{\"word\":\"riding\",\"start\":1.3,\"end\":1.66}]"} {"utterance_id": "3660-6517-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0012.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Once the audio has finished, identify every instance of a plant name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.9, 4.51, "oaten hay"]], "response_mode": "offline", "answer": "[{\"word\":\"oaten hay\",\"start\":3.9,\"end\":4.51}]"} {"utterance_id": "116-288048-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.76, 11.01, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":10.76,\"end\":11.01}]"} {"utterance_id": "116-288045-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.85, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":1.3,\"end\":1.85}]"} {"utterance_id": "700-122867-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.87, 2.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.87,\"end\":2.13}]"} {"utterance_id": "1650-173552-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0005.flac", "task_type": "lexical_counting", "query": "kremlin", "prompt": "Once the audio has finished, list each occurrence of the word \"kremlin\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.48, 2.15, "one"], [3.98, 4.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.48,\"end\":2.15},{\"word\":\"two\",\"start\":3.98,\"end\":4.63}]"} {"utterance_id": "4153-186222-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0010.flac", "task_type": "lexical_detection", "query": "jane's", "prompt": "When the utterance is complete, report all instances of the word \"jane's\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.66, "jane's"]], "response_mode": "offline", "answer": "[{\"word\":\"jane's\",\"start\":3.33,\"end\":3.66}]"} {"utterance_id": "6123-59150-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 1.02, "mister murry"]], "response_mode": "offline", "answer": "[{\"word\":\"mister murry\",\"start\":0.46,\"end\":1.02}]"} {"utterance_id": "1630-141772-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.67, "leg"]], "response_mode": "offline", "answer": "[{\"word\":\"leg\",\"start\":1.44,\"end\":1.67}]"} {"utterance_id": "3663-172528-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.83, 5.29, "pietro pagolo"]], "response_mode": "offline", "answer": "[{\"word\":\"pietro pagolo\",\"start\":3.83,\"end\":5.29}]"} {"utterance_id": "4323-13259-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0026.flac", "task_type": "lexical_detection", "query": "recently", "prompt": "Process the full audio offline. Find every instance of the word \"recently\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.04, 5.56, "recently"]], "response_mode": "offline", "answer": "[{\"word\":\"recently\",\"start\":5.04,\"end\":5.56}]"} {"utterance_id": "4323-18416-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.73, 1.96, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":1.73,\"end\":1.96}]"} {"utterance_id": "700-122868-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0011.flac", "task_type": "lexical_detection", "query": "arrange", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"arrange\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.37, "arrange"]], "response_mode": "offline", "answer": "[{\"word\":\"arrange\",\"start\":1.09,\"end\":1.37}]"} {"utterance_id": "4323-55228-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.35, 0.68, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":0.35,\"end\":0.68}]"} {"utterance_id": "3663-172005-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.07, 1.53, "legs"], [3.37, 3.98, "body"]], "response_mode": "offline", "answer": "[{\"word\":\"legs\",\"start\":1.07,\"end\":1.53},{\"word\":\"body\",\"start\":3.37,\"end\":3.98}]"} {"utterance_id": "4515-11057-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0000.flac", "task_type": "lexical_detection", "query": "eager", "prompt": "After listening to the whole recording, output every instance of the word \"eager\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.74, 10.19, "eager"]], "response_mode": "offline", "answer": "[{\"word\":\"eager\",\"start\":9.74,\"end\":10.19}]"} {"utterance_id": "3660-6517-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0014.flac", "task_type": "lexical_detection", "query": "mass", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"mass\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.09, "mass"]], "response_mode": "offline", "answer": "[{\"word\":\"mass\",\"start\":0.82,\"end\":1.09}]"} {"utterance_id": "6267-53049-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0007.flac", "task_type": "lexical_detection", "query": "penelope", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"penelope\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.37, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.69,\"end\":1.37}]"} {"utterance_id": "1255-90413-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0011.flac", "task_type": "lexical_detection", "query": "painted", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"painted\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.54, 5.0, "painted"]], "response_mode": "offline", "answer": "[{\"word\":\"painted\",\"start\":4.54,\"end\":5.0}]"} {"utterance_id": "1701-141760-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0017.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.72, 4.24, "wine"]], "response_mode": "offline", "answer": "[{\"word\":\"wine\",\"start\":3.72,\"end\":4.24}]"} {"utterance_id": "8254-115543-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.65, 2.15, "mahala"]], "response_mode": "offline", "answer": "[{\"word\":\"mahala\",\"start\":1.65,\"end\":2.15}]"} {"utterance_id": "1255-90413-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0022.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.01, "one"], [8.89, 9.93, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.4,\"end\":1.01},{\"word\":\"two\",\"start\":8.89,\"end\":9.93}]"} {"utterance_id": "3660-172183-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0014.flac", "task_type": "lexical_counting", "query": "unto", "prompt": "Listen to the entire clip, then report the count sequence for the word \"unto\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 1.97, "one"], [3.83, 4.14, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.74,\"end\":1.97},{\"word\":\"two\",\"start\":3.83,\"end\":4.14}]"} {"utterance_id": "4323-18416-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0015.flac", "task_type": "lexical_detection", "query": "suppose", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"suppose\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.78, 2.07, "suppose"]], "response_mode": "offline", "answer": "[{\"word\":\"suppose\",\"start\":1.78,\"end\":2.07}]"} {"utterance_id": "4323-18416-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0020.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Do not answer until the audio ends. Then list each instance of a money or currency term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.34, 6.73, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":6.34,\"end\":6.73}]"} {"utterance_id": "8254-115543-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.17, 2.74, "servants"]], "response_mode": "offline", "answer": "[{\"word\":\"servants\",\"start\":2.17,\"end\":2.74}]"} {"utterance_id": "6455-67804-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0016.flac", "task_type": "lexical_detection", "query": "advantage", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"advantage\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.99, 3.61, "advantage"]], "response_mode": "offline", "answer": "[{\"word\":\"advantage\",\"start\":2.99,\"end\":3.61}]"} {"utterance_id": "5849-50962-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0017.flac", "task_type": "lexical_detection", "query": "pressed", "prompt": "Process the full audio offline. Find every instance of the word \"pressed\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.6, 5.91, "pressed"]], "response_mode": "offline", "answer": "[{\"word\":\"pressed\",\"start\":5.6,\"end\":5.91}]"} {"utterance_id": "7601-101622-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0007.flac", "task_type": "lexical_detection", "query": "set", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"set\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.55, "set"]], "response_mode": "offline", "answer": "[{\"word\":\"set\",\"start\":2.26,\"end\":2.55}]"} {"utterance_id": "8254-84205-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0008.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Do not answer until the audio ends. Then list each instance of a time of day as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.35, 1.72, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.35,\"end\":1.72}]"} {"utterance_id": "6123-59186-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0005.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.0, 5.38, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":5.0,\"end\":5.38}]"} {"utterance_id": "6467-56885-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0011.flac", "task_type": "lexical_detection", "query": "darkness", "prompt": "After the audio ends, extract all occurrences of the word \"darkness\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 3.21, "darkness"]], "response_mode": "offline", "answer": "[{\"word\":\"darkness\",\"start\":2.57,\"end\":3.21}]"} {"utterance_id": "4323-55228-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.46, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":2.1,\"end\":2.46}]"} {"utterance_id": "4572-112375-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0006.flac", "task_type": "lexical_detection", "query": "fraught", "prompt": "When the utterance is complete, report all instances of the word \"fraught\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.85, 11.39, "fraught"]], "response_mode": "offline", "answer": "[{\"word\":\"fraught\",\"start\":10.85,\"end\":11.39}]"} {"utterance_id": "3915-98647-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0012.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.32, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":0.94,\"end\":1.32}]"} {"utterance_id": "116-288046-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.55, 5.22, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":4.55,\"end\":5.22}]"} {"utterance_id": "7601-101619-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0000.flac", "task_type": "lexical_detection", "query": "pleasure", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"pleasure\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.99, "pleasure"]], "response_mode": "offline", "answer": "[{\"word\":\"pleasure\",\"start\":1.48,\"end\":1.99}]"} {"utterance_id": "1630-73710-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0001.flac", "task_type": "lexical_detection", "query": "hope", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hope\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.25, "hope"], [3.76, 4.29, "hope"]], "response_mode": "offline", "answer": "[{\"word\":\"hope\",\"start\":1.79,\"end\":2.25},{\"word\":\"hope\",\"start\":3.76,\"end\":4.29}]"} {"utterance_id": "1650-167613-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0006.flac", "task_type": "lexical_detection", "query": "barclay", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"barclay\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.68, "barclay"]], "response_mode": "offline", "answer": "[{\"word\":\"barclay\",\"start\":2.1,\"end\":2.68}]"} {"utterance_id": "6599-38591-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.47, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":2.08,\"end\":2.47}]"} {"utterance_id": "1255-138279-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0012.flac", "task_type": "lexical_detection", "query": "morsel", "prompt": "Process the full audio offline. Find every instance of the word \"morsel\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.41, "morsel"]], "response_mode": "offline", "answer": "[{\"word\":\"morsel\",\"start\":1.08,\"end\":1.41}]"} {"utterance_id": "6123-59186-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0030.flac", "task_type": "lexical_detection", "query": "wives", "prompt": "Process the full audio offline. Find every instance of the word \"wives\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.39, "wives"]], "response_mode": "offline", "answer": "[{\"word\":\"wives\",\"start\":0.81,\"end\":1.39}]"} {"utterance_id": "1255-74899-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0004.flac", "task_type": "lexical_detection", "query": "delay", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"delay\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.67, 2.16, "delay"]], "response_mode": "offline", "answer": "[{\"word\":\"delay\",\"start\":1.67,\"end\":2.16}]"} {"utterance_id": "6599-38590-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0004.flac", "task_type": "lexical_detection", "query": "sir", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"sir\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.03, 3.58, "sir"]], "response_mode": "offline", "answer": "[{\"word\":\"sir\",\"start\":3.03,\"end\":3.58}]"} {"utterance_id": "5543-27761-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0063.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.65, 2.01, "mother's"]], "response_mode": "offline", "answer": "[{\"word\":\"mother's\",\"start\":1.65,\"end\":2.01}]"} {"utterance_id": "8254-115543-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0012.flac", "task_type": "lexical_counting", "query": "holy", "prompt": "After listening to the whole recording, output timed count words for the word \"holy\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[3.21, 3.58, "one"], [9.08, 9.43, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.21,\"end\":3.58},{\"word\":\"two\",\"start\":9.08,\"end\":9.43}]"} {"utterance_id": "4831-18525-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.74, "polly"], [2.87, 3.25, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":1.38,\"end\":1.74},{\"word\":\"phronsie\",\"start\":2.87,\"end\":3.25}]"} {"utterance_id": "6267-65525-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0008.flac", "task_type": "lexical_detection", "query": "slept", "prompt": "After the audio ends, extract all occurrences of the word \"slept\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.23, 3.69, "slept"]], "response_mode": "offline", "answer": "[{\"word\":\"slept\",\"start\":3.23,\"end\":3.69}]"} {"utterance_id": "6599-38590-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 3.23, "bourbon rome"]], "response_mode": "offline", "answer": "[{\"word\":\"bourbon rome\",\"start\":1.97,\"end\":3.23}]"} {"utterance_id": "5849-50962-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0025.flac", "task_type": "lexical_detection", "query": "beach", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"beach\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.2, 2.69, "beach"]], "response_mode": "offline", "answer": "[{\"word\":\"beach\",\"start\":2.2,\"end\":2.69}]"} {"utterance_id": "6123-59186-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0007.flac", "task_type": "lexical_detection", "query": "example", "prompt": "After the audio ends, extract all occurrences of the word \"example\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.36, "example"]], "response_mode": "offline", "answer": "[{\"word\":\"example\",\"start\":0.78,\"end\":1.36}]"} {"utterance_id": "3660-6517-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0011.flac", "task_type": "lexical_detection", "query": "sacks", "prompt": "When the utterance is complete, report all instances of the word \"sacks\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.94, "sacks"]], "response_mode": "offline", "answer": "[{\"word\":\"sacks\",\"start\":0.6,\"end\":0.94}]"} {"utterance_id": "3660-172183-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0024.flac", "task_type": "lexical_detection", "query": "conduct", "prompt": "After listening to the whole recording, output every instance of the word \"conduct\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.18, "conduct"]], "response_mode": "offline", "answer": "[{\"word\":\"conduct\",\"start\":2.76,\"end\":3.18}]"} {"utterance_id": "6267-65525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0011.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[3.05, 3.66, "one"], [6.8, 7.62, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.05,\"end\":3.66},{\"word\":\"two\",\"start\":6.8,\"end\":7.62}]"} {"utterance_id": "4515-11057-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0068.flac", "task_type": "lexical_detection", "query": "trust", "prompt": "Process the full audio offline. Find every instance of the word \"trust\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.68, "trust"]], "response_mode": "offline", "answer": "[{\"word\":\"trust\",\"start\":1.17,\"end\":1.68}]"} {"utterance_id": "8173-294714-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0059.flac", "task_type": "lexical_detection", "query": "kiss", "prompt": "Once the audio has finished, identify every instance of the word \"kiss\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.4, 0.72, "kiss"]], "response_mode": "offline", "answer": "[{\"word\":\"kiss\",\"start\":0.4,\"end\":0.72}]"} {"utterance_id": "7697-105815-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0028.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.94, 6.43, "flesh"]], "response_mode": "offline", "answer": "[{\"word\":\"flesh\",\"start\":5.94,\"end\":6.43}]"} {"utterance_id": "3915-98647-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.89, 3.43, "servants"]], "response_mode": "offline", "answer": "[{\"word\":\"servants\",\"start\":2.89,\"end\":3.43}]"} {"utterance_id": "4831-29134-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.49, 6.79, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":6.49,\"end\":6.79}]"} {"utterance_id": "1255-138279-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0003.flac", "task_type": "lexical_detection", "query": "difficulty", "prompt": "Once the audio has finished, identify every instance of the word \"difficulty\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.47, 8.0, "difficulty"]], "response_mode": "offline", "answer": "[{\"word\":\"difficulty\",\"start\":7.47,\"end\":8.0}]"} {"utterance_id": "4323-18416-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0023.flac", "task_type": "lexical_detection", "query": "hastily", "prompt": "When the utterance is complete, report all instances of the word \"hastily\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.45, "hastily"]], "response_mode": "offline", "answer": "[{\"word\":\"hastily\",\"start\":1.92,\"end\":2.45}]"} {"utterance_id": "1255-138279-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0023.flac", "task_type": "lexical_detection", "query": "forgot", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"forgot\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.79, 2.31, "forgot"]], "response_mode": "offline", "answer": "[{\"word\":\"forgot\",\"start\":1.79,\"end\":2.31}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After listening to the whole recording, output every instance of an organization name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.82, 11.3, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"army\",\"start\":10.82,\"end\":11.3}]"} {"utterance_id": "6467-56885-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0000.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After the audio ends, tally every instance of a body part. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.56, 0.87, "one"], [4.57, 4.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.56,\"end\":0.87},{\"word\":\"two\",\"start\":4.57,\"end\":4.92}]"} {"utterance_id": "6467-56885-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0012.flac", "task_type": "lexical_detection", "query": "scoundrel", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"scoundrel\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.25, 3.92, "scoundrel"]], "response_mode": "offline", "answer": "[{\"word\":\"scoundrel\",\"start\":3.25,\"end\":3.92}]"} {"utterance_id": "6123-59186-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.29, 2.78, "browning's"]], "response_mode": "offline", "answer": "[{\"word\":\"browning's\",\"start\":2.29,\"end\":2.78}]"} {"utterance_id": "3660-172182-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0017.flac", "task_type": "lexical_detection", "query": "known", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"known\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.76, 5.16, "known"]], "response_mode": "offline", "answer": "[{\"word\":\"known\",\"start\":4.76,\"end\":5.16}]"} {"utterance_id": "7697-245712-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0009.flac", "task_type": "lexical_detection", "query": "obeyed", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"obeyed\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.76, "obeyed"]], "response_mode": "offline", "answer": "[{\"word\":\"obeyed\",\"start\":1.46,\"end\":1.76}]"} {"utterance_id": "4572-112381-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0019.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.77, "kinsman"]], "response_mode": "offline", "answer": "[{\"word\":\"kinsman\",\"start\":1.08,\"end\":1.77}]"} {"utterance_id": "8254-115543-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 4.83, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.52,\"end\":4.83}]"} {"utterance_id": "7641-96684-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0027.flac", "task_type": "lexical_detection", "query": "sacrifice", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"sacrifice\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.23, 1.83, "sacrifice"]], "response_mode": "offline", "answer": "[{\"word\":\"sacrifice\",\"start\":1.23,\"end\":1.83}]"} {"utterance_id": "4323-18416-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0020.flac", "task_type": "lexical_counting", "query": "charlotte", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"charlotte\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.06, 2.55, "one"], [5.09, 5.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.06,\"end\":2.55},{\"word\":\"two\",\"start\":5.09,\"end\":5.56}]"} {"utterance_id": "6599-38591-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0004.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.72, 2.23, "one"], [4.95, 5.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.72,\"end\":2.23},{\"word\":\"two\",\"start\":4.95,\"end\":5.6}]"} {"utterance_id": "1701-141760-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0044.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.98, 7.48, "captains"]], "response_mode": "offline", "answer": "[{\"word\":\"captains\",\"start\":6.98,\"end\":7.48}]"} {"utterance_id": "6455-67803-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0017.flac", "task_type": "lexical_detection", "query": "length", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"length\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.57, "length"]], "response_mode": "offline", "answer": "[{\"word\":\"length\",\"start\":2.14,\"end\":2.57}]"} {"utterance_id": "1585-157660-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0008.flac", "task_type": "lexical_counting", "query": "oft", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"oft\" (word, start, end). If none, reply [].", "labels": [[0.45, 0.86, "one"], [2.74, 3.06, "two"], [5.94, 6.27, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.45,\"end\":0.86},{\"word\":\"two\",\"start\":2.74,\"end\":3.06},{\"word\":\"three\",\"start\":5.94,\"end\":6.27}]"} {"utterance_id": "7601-101622-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0004.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.86, 6.61, "fathers"]], "response_mode": "offline", "answer": "[{\"word\":\"fathers\",\"start\":5.86,\"end\":6.61}]"} {"utterance_id": "1650-167613-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0004.flac", "task_type": "lexical_detection", "query": "fully", "prompt": "When the utterance is complete, report all instances of the word \"fully\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.89, 4.39, "fully"]], "response_mode": "offline", "answer": "[{\"word\":\"fully\",\"start\":3.89,\"end\":4.39}]"} {"utterance_id": "6841-88291-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0011.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Do not answer until the audio ends. Then list each instance of a state or province name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.43, 2.01, "california"]], "response_mode": "offline", "answer": "[{\"word\":\"california\",\"start\":1.43,\"end\":2.01}]"} {"utterance_id": "1686-142278-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0020.flac", "task_type": "lexical_detection", "query": "bishop", "prompt": "After the audio ends, extract all occurrences of the word \"bishop\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.42, "bishop"]], "response_mode": "offline", "answer": "[{\"word\":\"bishop\",\"start\":1.09,\"end\":1.42}]"} {"utterance_id": "4515-11057-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.55, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":1.12,\"end\":1.55}]"} {"utterance_id": "116-288045-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0001.flac", "task_type": "lexical_detection", "query": "extended", "prompt": "Process the full audio offline. Find every instance of the word \"extended\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.24, 6.66, "extended"]], "response_mode": "offline", "answer": "[{\"word\":\"extended\",\"start\":6.24,\"end\":6.66}]"} {"utterance_id": "1686-142278-0069", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0069.flac", "task_type": "lexical_detection", "query": "something", "prompt": "After listening to the whole recording, output every instance of the word \"something\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.07, 4.58, "something"]], "response_mode": "offline", "answer": "[{\"word\":\"something\",\"start\":4.07,\"end\":4.58}]"} {"utterance_id": "3915-98647-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0029.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.86, 2.19, "body"]], "response_mode": "offline", "answer": "[{\"word\":\"body\",\"start\":1.86,\"end\":2.19}]"} {"utterance_id": "116-288045-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.52, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":2.09,\"end\":2.52}]"} {"utterance_id": "6467-97061-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0012.flac", "task_type": "lexical_detection", "query": "bowing", "prompt": "After the audio ends, extract all occurrences of the word \"bowing\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.13, "bowing"]], "response_mode": "offline", "answer": "[{\"word\":\"bowing\",\"start\":0.79,\"end\":1.13}]"} {"utterance_id": "700-122866-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.8, 2.32, "nose"]], "response_mode": "offline", "answer": "[{\"word\":\"nose\",\"start\":1.8,\"end\":2.32}]"} {"utterance_id": "7697-105815-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0002.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.8, 7.45, "zion"]], "response_mode": "offline", "answer": "[{\"word\":\"zion\",\"start\":6.8,\"end\":7.45}]"} {"utterance_id": "8173-294714-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0002.flac", "task_type": "lexical_detection", "query": "arrested", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"arrested\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.92, 3.52, "arrested"]], "response_mode": "offline", "answer": "[{\"word\":\"arrested\",\"start\":2.92,\"end\":3.52}]"} {"utterance_id": "6841-88291-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0006.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After listening to the whole recording, output every instance of a plant name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 2.16, "greasewood"]], "response_mode": "offline", "answer": "[{\"word\":\"greasewood\",\"start\":1.54,\"end\":2.16}]"} {"utterance_id": "6841-88294-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0056.flac", "task_type": "lexical_detection", "query": "plain", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"plain\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.57, "plain"]], "response_mode": "offline", "answer": "[{\"word\":\"plain\",\"start\":2.31,\"end\":2.57}]"} {"utterance_id": "4323-18416-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0022.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen carefully to the entire clip, then report every instance of a money or currency term as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.94, 3.22, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":2.94,\"end\":3.22}]"} {"utterance_id": "4515-11057-0113", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0113.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.12, 0.53, "several"]], "response_mode": "offline", "answer": "[{\"word\":\"several\",\"start\":0.12,\"end\":0.53}]"} {"utterance_id": "1701-141759-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0031.flac", "task_type": "lexical_detection", "query": "cry", "prompt": "After listening to the whole recording, output every instance of the word \"cry\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.0, "cry"]], "response_mode": "offline", "answer": "[{\"word\":\"cry\",\"start\":2.61,\"end\":3.0}]"} {"utterance_id": "3660-172183-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0021.flac", "task_type": "lexical_detection", "query": "counsel", "prompt": "When the utterance is complete, report all instances of the word \"counsel\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.99, "counsel"]], "response_mode": "offline", "answer": "[{\"word\":\"counsel\",\"start\":1.47,\"end\":1.99}]"} {"utterance_id": "3663-172528-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.76, "messer francesco"]], "response_mode": "offline", "answer": "[{\"word\":\"messer francesco\",\"start\":0.4,\"end\":1.76}]"} {"utterance_id": "116-288048-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0008.flac", "task_type": "lexical_detection", "query": "origin", "prompt": "When the utterance is complete, report all instances of the word \"origin\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.27, "origin"]], "response_mode": "offline", "answer": "[{\"word\":\"origin\",\"start\":1.88,\"end\":2.27}]"} {"utterance_id": "4153-186223-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.01, 1.6, "tinsmith"]], "response_mode": "offline", "answer": "[{\"word\":\"tinsmith\",\"start\":1.01,\"end\":1.6}]"} {"utterance_id": "700-122866-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0008.flac", "task_type": "lexical_detection", "query": "chatter", "prompt": "When the utterance is complete, report all instances of the word \"chatter\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.73, 6.19, "chatter"]], "response_mode": "offline", "answer": "[{\"word\":\"chatter\",\"start\":5.73,\"end\":6.19}]"} {"utterance_id": "116-288045-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.01, 1.69, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":1.01,\"end\":1.69}]"} {"utterance_id": "6841-88291-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0049.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.78, 2.27, "cattleman"]], "response_mode": "offline", "answer": "[{\"word\":\"cattleman\",\"start\":1.78,\"end\":2.27}]"} {"utterance_id": "1686-142278-0081", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0081.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.74, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":1.25,\"end\":1.74}]"} {"utterance_id": "4323-55228-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0051.flac", "task_type": "lexical_detection", "query": "cut", "prompt": "Once the audio has finished, identify every instance of the word \"cut\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.25, 2.43, "cut"]], "response_mode": "offline", "answer": "[{\"word\":\"cut\",\"start\":2.25,\"end\":2.43}]"} {"utterance_id": "4831-29134-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0002.flac", "task_type": "lexical_detection", "query": "spell", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"spell\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.57, 3.87, "spell"]], "response_mode": "offline", "answer": "[{\"word\":\"spell\",\"start\":3.57,\"end\":3.87}]"} {"utterance_id": "1585-131718-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0043.flac", "task_type": "lexical_detection", "query": "nights", "prompt": "After the audio ends, extract all occurrences of the word \"nights\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.18, "nights"]], "response_mode": "offline", "answer": "[{\"word\":\"nights\",\"start\":1.77,\"end\":2.18}]"} {"utterance_id": "6123-59150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.7, 1.35, "dostoevsky"], [2.96, 3.55, "strindberg"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":0.7,\"end\":1.35},{\"word\":\"strindberg\",\"start\":2.96,\"end\":3.55}]"} {"utterance_id": "1255-90413-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0008.flac", "task_type": "lexical_detection", "query": "pull", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"pull\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.99, 2.32, "pull"]], "response_mode": "offline", "answer": "[{\"word\":\"pull\",\"start\":1.99,\"end\":2.32}]"} {"utterance_id": "1686-142278-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0055.flac", "task_type": "lexical_detection", "query": "mirror", "prompt": "Once the audio has finished, identify every instance of the word \"mirror\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.56, 5.97, "mirror"]], "response_mode": "offline", "answer": "[{\"word\":\"mirror\",\"start\":5.56,\"end\":5.97}]"} {"utterance_id": "7697-105815-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.01, 1.52, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":1.01,\"end\":1.52}]"} {"utterance_id": "7697-105815-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0008.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.73, 1.12, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":0.73,\"end\":1.12}]"} {"utterance_id": "6267-65525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0007.flac", "task_type": "lexical_counting", "query": "small", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"small\" with word/start/end. If none, reply [].", "labels": [[0.79, 1.23, "one"], [4.63, 5.09, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.79,\"end\":1.23},{\"word\":\"two\",\"start\":4.63,\"end\":5.09}]"} {"utterance_id": "6455-67803-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0019.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Once the audio has finished, list each occurrence of a body part as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.87, 1.31, "one"], [2.48, 2.69, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.87,\"end\":1.31},{\"word\":\"two\",\"start\":2.48,\"end\":2.69}]"} {"utterance_id": "700-122867-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0010.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Process the full audio offline. Find every instance of a color name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.8, "green"]], "response_mode": "offline", "answer": "[{\"word\":\"green\",\"start\":1.16,\"end\":1.8}]"} {"utterance_id": "1686-142278-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0064.flac", "task_type": "semantic_detection", "query": "money", "prompt": "You will hear a complete utterance. When it ends, list every instance of a money or currency term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.62, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":1.26,\"end\":1.62}]"} {"utterance_id": "6599-38590-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0002.flac", "task_type": "lexical_detection", "query": "office", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"office\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.91, 3.55, "office"]], "response_mode": "offline", "answer": "[{\"word\":\"office\",\"start\":2.91,\"end\":3.55}]"} {"utterance_id": "1701-141760-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.19, 2.45, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":2.19,\"end\":2.45}]"} {"utterance_id": "4153-185072-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0012.flac", "task_type": "lexical_detection", "query": "expected", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"expected\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.92, 4.57, "expected"]], "response_mode": "offline", "answer": "[{\"word\":\"expected\",\"start\":3.92,\"end\":4.57}]"} {"utterance_id": "4153-185072-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.43, 0.77, "allen"]], "response_mode": "offline", "answer": "[{\"word\":\"allen\",\"start\":0.43,\"end\":0.77}]"} {"utterance_id": "5849-50962-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.05, 3.39, "nellie"]], "response_mode": "offline", "answer": "[{\"word\":\"nellie\",\"start\":3.05,\"end\":3.39}]"} {"utterance_id": "6841-88291-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0056.flac", "task_type": "lexical_detection", "query": "periodical", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"periodical\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.57, 3.28, "periodical"]], "response_mode": "offline", "answer": "[{\"word\":\"periodical\",\"start\":2.57,\"end\":3.28}]"} {"utterance_id": "4570-14911-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.88, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":0.5,\"end\":0.88}]"} {"utterance_id": "6455-67804-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0025.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.34, 0.89, "one"], [2.45, 3.11, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.34,\"end\":0.89},{\"word\":\"two\",\"start\":2.45,\"end\":3.11}]"} {"utterance_id": "8288-274162-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0064.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[3.12, 3.46, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":3.12,\"end\":3.46}]"} {"utterance_id": "4570-14911-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0013.flac", "task_type": "lexical_detection", "query": "shoulders", "prompt": "When the utterance is complete, report all instances of the word \"shoulders\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 1.11, "shoulders"]], "response_mode": "offline", "answer": "[{\"word\":\"shoulders\",\"start\":0.56,\"end\":1.11}]"} {"utterance_id": "5849-50963-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0005.flac", "task_type": "lexical_detection", "query": "started", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"started\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.48, 3.88, "started"]], "response_mode": "offline", "answer": "[{\"word\":\"started\",\"start\":3.48,\"end\":3.88}]"} {"utterance_id": "1255-74899-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.85, 5.38, "beast"]], "response_mode": "offline", "answer": "[{\"word\":\"beast\",\"start\":4.85,\"end\":5.38}]"} {"utterance_id": "6123-59186-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0031.flac", "task_type": "lexical_detection", "query": "reality", "prompt": "When the utterance is complete, report all instances of the word \"reality\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.42, 2.99, "reality"]], "response_mode": "offline", "answer": "[{\"word\":\"reality\",\"start\":2.42,\"end\":2.99}]"} {"utterance_id": "4570-14911-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0004.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.91, 11.33, "second"]], "response_mode": "offline", "answer": "[{\"word\":\"second\",\"start\":10.91,\"end\":11.33}]"} {"utterance_id": "6267-65525-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0032.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.53, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":2.14,\"end\":2.53}]"} {"utterance_id": "8288-274162-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0054.flac", "task_type": "lexical_counting", "query": "going", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"going\" (word, start, end). If none, reply [].", "labels": [[3.24, 3.53, "one"], [4.94, 5.77, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.24,\"end\":3.53},{\"word\":\"two\",\"start\":4.94,\"end\":5.77}]"} {"utterance_id": "2506-11278-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0006.flac", "task_type": "lexical_detection", "query": "exercised", "prompt": "After listening to the whole recording, output every instance of the word \"exercised\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.83, "exercised"]], "response_mode": "offline", "answer": "[{\"word\":\"exercised\",\"start\":3.22,\"end\":3.83}]"} {"utterance_id": "6267-53049-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.77, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.97,\"end\":1.77}]"} {"utterance_id": "4323-18416-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0006.flac", "task_type": "lexical_detection", "query": "polly's", "prompt": "Process the full audio offline. Find every instance of the word \"polly's\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.9, 3.23, "polly's"]], "response_mode": "offline", "answer": "[{\"word\":\"polly's\",\"start\":2.9,\"end\":3.23}]"} {"utterance_id": "1701-141760-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0040.flac", "task_type": "lexical_detection", "query": "assistance", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"assistance\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.51, 5.23, "assistance"]], "response_mode": "offline", "answer": "[{\"word\":\"assistance\",\"start\":4.51,\"end\":5.23}]"} {"utterance_id": "700-122867-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0021.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After listening to the whole recording, output every instance of a color name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.61, "raven black"]], "response_mode": "offline", "answer": "[{\"word\":\"raven black\",\"start\":2.81,\"end\":3.61}]"} {"utterance_id": "7641-96670-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0013.flac", "task_type": "lexical_counting", "query": "eyes", "prompt": "When the utterance is complete, report how the word \"eyes\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.77, 4.07, "one"], [5.12, 5.59, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.77,\"end\":4.07},{\"word\":\"two\",\"start\":5.12,\"end\":5.59}]"} {"utterance_id": "8288-274162-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0053.flac", "task_type": "lexical_detection", "query": "concern", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"concern\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.61, 2.17, "concern"]], "response_mode": "offline", "answer": "[{\"word\":\"concern\",\"start\":1.61,\"end\":2.17}]"} {"utterance_id": "8254-115543-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0033.flac", "task_type": "lexical_detection", "query": "orders", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"orders\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.56, "orders"]], "response_mode": "offline", "answer": "[{\"word\":\"orders\",\"start\":1.19,\"end\":1.56}]"} {"utterance_id": "3915-98647-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0024.flac", "task_type": "lexical_detection", "query": "watched", "prompt": "After the audio ends, extract all occurrences of the word \"watched\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.84, 3.29, "watched"]], "response_mode": "offline", "answer": "[{\"word\":\"watched\",\"start\":2.84,\"end\":3.29}]"} {"utterance_id": "2506-11278-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0031.flac", "task_type": "lexical_detection", "query": "harm", "prompt": "Process the full audio offline. Find every instance of the word \"harm\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.31, "harm"]], "response_mode": "offline", "answer": "[{\"word\":\"harm\",\"start\":0.97,\"end\":1.31}]"} {"utterance_id": "4323-18416-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.97, 2.27, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.97,\"end\":2.27}]"} {"utterance_id": "6467-94831-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0018.flac", "task_type": "lexical_detection", "query": "erect", "prompt": "When the utterance is complete, report all instances of the word \"erect\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.76, "erect"]], "response_mode": "offline", "answer": "[{\"word\":\"erect\",\"start\":3.39,\"end\":3.76}]"} {"utterance_id": "4831-25894-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0028.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[4.19, 4.57, "one"], [9.33, 9.73, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.19,\"end\":4.57},{\"word\":\"two\",\"start\":9.33,\"end\":9.73}]"} {"utterance_id": "6467-97061-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0010.flac", "task_type": "lexical_detection", "query": "elephant", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"elephant\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 2.86, "elephant"], [13.65, 14.08, "elephant"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant\",\"start\":2.4,\"end\":2.86},{\"word\":\"elephant\",\"start\":13.65,\"end\":14.08}]"} {"utterance_id": "7697-105815-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0005.flac", "task_type": "lexical_detection", "query": "does", "prompt": "After listening to the whole recording, output every instance of the word \"does\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 1.93, "does"]], "response_mode": "offline", "answer": "[{\"word\":\"does\",\"start\":1.74,\"end\":1.93}]"} {"utterance_id": "1650-167613-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0052.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.4, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":2.09,\"end\":2.4}]"} {"utterance_id": "7641-96684-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0029.flac", "task_type": "lexical_detection", "query": "dragged", "prompt": "After listening to the whole recording, output every instance of the word \"dragged\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.86, "dragged"]], "response_mode": "offline", "answer": "[{\"word\":\"dragged\",\"start\":0.46,\"end\":0.86}]"} {"utterance_id": "3660-6517-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0021.flac", "task_type": "lexical_counting", "query": "pathetic", "prompt": "Process the full audio offline. For every occurrence of the word \"pathetic\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.2, "one"], [6.96, 7.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.67,\"end\":1.2},{\"word\":\"two\",\"start\":6.96,\"end\":7.44}]"} {"utterance_id": "1630-102884-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0016.flac", "task_type": "lexical_detection", "query": "slavery", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"slavery\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.14, 9.87, "slavery"]], "response_mode": "offline", "answer": "[{\"word\":\"slavery\",\"start\":9.14,\"end\":9.87}]"} {"utterance_id": "4515-11057-0104", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0104.flac", "task_type": "lexical_detection", "query": "lose", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"lose\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.41, "lose"]], "response_mode": "offline", "answer": "[{\"word\":\"lose\",\"start\":2.91,\"end\":3.41}]"} {"utterance_id": "4515-11057-0093", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0093.flac", "task_type": "lexical_detection", "query": "say", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"say\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.49, "say"]], "response_mode": "offline", "answer": "[{\"word\":\"say\",\"start\":1.11,\"end\":1.49}]"} {"utterance_id": "4515-11057-0100", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0100.flac", "task_type": "lexical_detection", "query": "mingle", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mingle\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.02, 4.55, "mingle"]], "response_mode": "offline", "answer": "[{\"word\":\"mingle\",\"start\":4.02,\"end\":4.55}]"} {"utterance_id": "4572-64670-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0006.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Process the full audio offline. For every occurrence of a body part, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.65, 5.15, "one"], [7.72, 8.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.65,\"end\":5.15},{\"word\":\"two\",\"start\":7.72,\"end\":8.13}]"} {"utterance_id": "8288-274162-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.82, 1.43, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":0.82,\"end\":1.43}]"} {"utterance_id": "2506-11267-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0015.flac", "task_type": "lexical_detection", "query": "passes", "prompt": "Process the full audio offline. Find every instance of the word \"passes\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.17, "passes"]], "response_mode": "offline", "answer": "[{\"word\":\"passes\",\"start\":0.65,\"end\":1.17}]"} {"utterance_id": "3915-98647-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0018.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.09, 4.54, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":4.09,\"end\":4.54}]"} {"utterance_id": "8288-274162-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0043.flac", "task_type": "lexical_detection", "query": "taught", "prompt": "After the audio ends, extract all occurrences of the word \"taught\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 1.83, "taught"]], "response_mode": "offline", "answer": "[{\"word\":\"taught\",\"start\":1.49,\"end\":1.83}]"} {"utterance_id": "1701-141760-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0048.flac", "task_type": "lexical_detection", "query": "weight", "prompt": "After listening to the whole recording, output every instance of the word \"weight\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.61, 2.09, "weight"]], "response_mode": "offline", "answer": "[{\"word\":\"weight\",\"start\":1.61,\"end\":2.09}]"} {"utterance_id": "6123-59150-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0042.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.28, 5.65, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":5.28,\"end\":5.65}]"} {"utterance_id": "3663-172528-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0047.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 2.05, "florence"]], "response_mode": "offline", "answer": "[{\"word\":\"florence\",\"start\":1.28,\"end\":2.05}]"} {"utterance_id": "6841-88294-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0003.flac", "task_type": "lexical_detection", "query": "fellow", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fellow\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.02, 3.34, "fellow"]], "response_mode": "offline", "answer": "[{\"word\":\"fellow\",\"start\":3.02,\"end\":3.34}]"} {"utterance_id": "8254-84205-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0047.flac", "task_type": "lexical_detection", "query": "turn", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"turn\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 0.86, "turn"]], "response_mode": "offline", "answer": "[{\"word\":\"turn\",\"start\":0.67,\"end\":0.86}]"} {"utterance_id": "1630-73710-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0014.flac", "task_type": "lexical_detection", "query": "pangs", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pangs\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.45, 0.84, "pangs"]], "response_mode": "offline", "answer": "[{\"word\":\"pangs\",\"start\":0.45,\"end\":0.84}]"} {"utterance_id": "116-288048-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.54, 8.11, "mothers"]], "response_mode": "offline", "answer": "[{\"word\":\"mothers\",\"start\":7.54,\"end\":8.11}]"} {"utterance_id": "8173-294714-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0005.flac", "task_type": "lexical_counting", "query": "man", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"man\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[4.32, 4.68, "one"], [6.4, 6.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.32,\"end\":4.68},{\"word\":\"two\",\"start\":6.4,\"end\":6.78}]"} {"utterance_id": "6267-65525-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0009.flac", "task_type": "lexical_detection", "query": "size", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"size\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.94, 8.46, "size"]], "response_mode": "offline", "answer": "[{\"word\":\"size\",\"start\":7.94,\"end\":8.46}]"} {"utterance_id": "1630-102884-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0009.flac", "task_type": "lexical_detection", "query": "construction", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"construction\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.3, 3.1, "construction"]], "response_mode": "offline", "answer": "[{\"word\":\"construction\",\"start\":2.3,\"end\":3.1}]"} {"utterance_id": "3915-98647-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0036.flac", "task_type": "lexical_detection", "query": "daily", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"daily\" with fields word, start, end (seconds). If none, reply [].", "labels": [[10.22, 10.68, "daily"]], "response_mode": "offline", "answer": "[{\"word\":\"daily\",\"start\":10.22,\"end\":10.68}]"} {"utterance_id": "1651-136854-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0018.flac", "task_type": "lexical_detection", "query": "private", "prompt": "After listening to the whole recording, output every instance of the word \"private\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.63, "private"]], "response_mode": "offline", "answer": "[{\"word\":\"private\",\"start\":2.13,\"end\":2.63}]"} {"utterance_id": "700-122868-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0027.flac", "task_type": "lexical_detection", "query": "gilbert", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"gilbert\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.86, "gilbert"]], "response_mode": "offline", "answer": "[{\"word\":\"gilbert\",\"start\":0.48,\"end\":0.86}]"} {"utterance_id": "6467-56885-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.57, "birds"]], "response_mode": "offline", "answer": "[{\"word\":\"birds\",\"start\":1.1,\"end\":1.57}]"} {"utterance_id": "4570-24733-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0002.flac", "task_type": "lexical_detection", "query": "nature", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"nature\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.01, 1.59, "nature"]], "response_mode": "offline", "answer": "[{\"word\":\"nature\",\"start\":1.01,\"end\":1.59}]"} {"utterance_id": "3663-172005-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0005.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[1.07, 1.53, "one"], [3.37, 3.98, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.07,\"end\":1.53},{\"word\":\"two\",\"start\":3.37,\"end\":3.98}]"} {"utterance_id": "7697-245712-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0006.flac", "task_type": "lexical_detection", "query": "natural", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"natural\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.2, 3.85, "natural"]], "response_mode": "offline", "answer": "[{\"word\":\"natural\",\"start\":3.2,\"end\":3.85}]"} {"utterance_id": "3915-57461-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0020.flac", "task_type": "lexical_detection", "query": "journey", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"journey\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.05, "journey"]], "response_mode": "offline", "answer": "[{\"word\":\"journey\",\"start\":0.73,\"end\":1.05}]"} {"utterance_id": "4572-112383-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 0.8, "seven"]], "response_mode": "offline", "answer": "[{\"word\":\"seven\",\"start\":0.35,\"end\":0.8}]"} {"utterance_id": "7697-245712-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.27, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":1.92,\"end\":2.27}]"} {"utterance_id": "4323-18416-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0014.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "When the utterance is complete, report how a family relation accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 1.19, "one"], [4.58, 5.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.64,\"end\":1.19},{\"word\":\"two\",\"start\":4.58,\"end\":5.13}]"} {"utterance_id": "700-122866-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.59, "missus allan"]], "response_mode": "offline", "answer": "[{\"word\":\"missus allan\",\"start\":1.92,\"end\":2.59}]"} {"utterance_id": "700-122867-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0022.flac", "task_type": "lexical_detection", "query": "practice", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"practice\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.81, "practice"]], "response_mode": "offline", "answer": "[{\"word\":\"practice\",\"start\":4.32,\"end\":4.81}]"} {"utterance_id": "4570-102353-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0007.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.17, 9.65, "father"], [13.13, 13.65, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":9.17,\"end\":9.65},{\"word\":\"son\",\"start\":13.13,\"end\":13.65}]"} {"utterance_id": "1585-131718-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.47, 2.95, "edison"]], "response_mode": "offline", "answer": "[{\"word\":\"edison\",\"start\":2.47,\"end\":2.95}]"} {"utterance_id": "3660-172182-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0029.flac", "task_type": "lexical_detection", "query": "neighboring", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"neighboring\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.13, "neighboring"]], "response_mode": "offline", "answer": "[{\"word\":\"neighboring\",\"start\":0.72,\"end\":1.13}]"} {"utterance_id": "1686-142278-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0063.flac", "task_type": "lexical_detection", "query": "plans", "prompt": "After the audio ends, extract all occurrences of the word \"plans\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.77, 5.15, "plans"]], "response_mode": "offline", "answer": "[{\"word\":\"plans\",\"start\":4.77,\"end\":5.15}]"} {"utterance_id": "3915-98647-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0029.flac", "task_type": "lexical_detection", "query": "clothing", "prompt": "Once the audio has finished, identify every instance of the word \"clothing\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.59, 5.14, "clothing"]], "response_mode": "offline", "answer": "[{\"word\":\"clothing\",\"start\":4.59,\"end\":5.14}]"} {"utterance_id": "7641-96684-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.3, "heracles"]], "response_mode": "offline", "answer": "[{\"word\":\"heracles\",\"start\":0.71,\"end\":1.3}]"} {"utterance_id": "6841-88294-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0061.flac", "task_type": "lexical_detection", "query": "got", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"got\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.09, "got"]], "response_mode": "offline", "answer": "[{\"word\":\"got\",\"start\":1.9,\"end\":2.09}]"} {"utterance_id": "700-122868-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.02, "anne's"], [2.11, 2.52, "elaine"]], "response_mode": "offline", "answer": "[{\"word\":\"anne's\",\"start\":0.77,\"end\":1.02},{\"word\":\"elaine\",\"start\":2.11,\"end\":2.52}]"} {"utterance_id": "6455-67803-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0031.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.47, 0.84, "one"], [4.07, 4.96, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.47,\"end\":0.84},{\"word\":\"two\",\"start\":4.07,\"end\":4.96}]"} {"utterance_id": "1701-141759-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0028.flac", "task_type": "lexical_detection", "query": "anna", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"anna\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.63, "anna"]], "response_mode": "offline", "answer": "[{\"word\":\"anna\",\"start\":0.37,\"end\":0.63}]"} {"utterance_id": "8173-294714-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0013.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 2.01, "wife"], [5.12, 5.58, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.64,\"end\":2.01},{\"word\":\"father\",\"start\":5.12,\"end\":5.58}]"} {"utterance_id": "1701-141760-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0005.flac", "task_type": "lexical_detection", "query": "dressed", "prompt": "When the utterance is complete, report all instances of the word \"dressed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.1, 6.69, "dressed"]], "response_mode": "offline", "answer": "[{\"word\":\"dressed\",\"start\":6.1,\"end\":6.69}]"} {"utterance_id": "6123-59150-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.23, 4.02, "dostoevsky s"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky s\",\"start\":3.23,\"end\":4.02}]"} {"utterance_id": "5849-50873-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.66, "charles"], [3.56, 4.49, "captain wadsworth"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":1.22,\"end\":1.66},{\"word\":\"captain wadsworth\",\"start\":3.56,\"end\":4.49}]"} {"utterance_id": "7697-105815-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.89, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.26,\"end\":0.89}]"} {"utterance_id": "3660-172182-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0011.flac", "task_type": "lexical_detection", "query": "greets", "prompt": "After the audio ends, extract all occurrences of the word \"greets\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.26, "greets"]], "response_mode": "offline", "answer": "[{\"word\":\"greets\",\"start\":0.96,\"end\":1.26}]"} {"utterance_id": "1701-141760-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0021.flac", "task_type": "lexical_detection", "query": "boris", "prompt": "After the audio ends, extract all occurrences of the word \"boris\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.59, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":3.05,\"end\":3.59}]"} {"utterance_id": "2506-11278-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0029.flac", "task_type": "lexical_detection", "query": "ladies", "prompt": "After the audio ends, extract all occurrences of the word \"ladies\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.5, "ladies"]], "response_mode": "offline", "answer": "[{\"word\":\"ladies\",\"start\":1.99,\"end\":2.5}]"} {"utterance_id": "5543-27761-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0006.flac", "task_type": "lexical_detection", "query": "wherever", "prompt": "Once the audio has finished, identify every instance of the word \"wherever\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.09, 3.35, "wherever"]], "response_mode": "offline", "answer": "[{\"word\":\"wherever\",\"start\":3.09,\"end\":3.35}]"} {"utterance_id": "116-288045-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0023.flac", "task_type": "lexical_detection", "query": "suggestion", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"suggestion\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 3.01, "suggestion"]], "response_mode": "offline", "answer": "[{\"word\":\"suggestion\",\"start\":2.43,\"end\":3.01}]"} {"utterance_id": "4515-11057-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0060.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.57, 5.62, "missus milvain"]], "response_mode": "offline", "answer": "[{\"word\":\"missus milvain\",\"start\":4.57,\"end\":5.62}]"} {"utterance_id": "4831-29134-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0008.flac", "task_type": "lexical_detection", "query": "consciousness", "prompt": "After the audio ends, extract all occurrences of the word \"consciousness\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.42, "consciousness"]], "response_mode": "offline", "answer": "[{\"word\":\"consciousness\",\"start\":1.74,\"end\":2.42}]"} {"utterance_id": "7641-96252-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0005.flac", "task_type": "lexical_detection", "query": "virtue", "prompt": "After the audio ends, extract all occurrences of the word \"virtue\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.15, 2.64, "virtue"]], "response_mode": "offline", "answer": "[{\"word\":\"virtue\",\"start\":2.15,\"end\":2.64}]"} {"utterance_id": "1585-157660-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0010.flac", "task_type": "semantic_detection", "query": "money", "prompt": "You will hear a complete utterance. When it ends, list every instance of a money or currency term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.3, 3.72, "guinea"]], "response_mode": "offline", "answer": "[{\"word\":\"guinea\",\"start\":3.3,\"end\":3.72}]"} {"utterance_id": "1686-142278-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0037.flac", "task_type": "lexical_counting", "query": "tried", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"tried\" (word, start, end). If none, reply [].", "labels": [[1.8, 2.1, "one"], [3.24, 3.51, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.8,\"end\":2.1},{\"word\":\"two\",\"start\":3.24,\"end\":3.51}]"} {"utterance_id": "4153-186223-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.29, 9.53, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.29,\"end\":9.53}]"} {"utterance_id": "4831-18525-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0014.flac", "task_type": "lexical_detection", "query": "bag", "prompt": "After listening to the whole recording, output every instance of the word \"bag\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.7, 5.99, "bag"]], "response_mode": "offline", "answer": "[{\"word\":\"bag\",\"start\":5.7,\"end\":5.99}]"} {"utterance_id": "4153-186222-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0028.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.45, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":2.08,\"end\":2.45}]"} {"utterance_id": "8288-274150-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.52, 1.09, "la valliere"]], "response_mode": "offline", "answer": "[{\"word\":\"la valliere\",\"start\":0.52,\"end\":1.09}]"} {"utterance_id": "4153-186222-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.7, 3.9, "beth"]], "response_mode": "offline", "answer": "[{\"word\":\"beth\",\"start\":3.7,\"end\":3.9}]"} {"utterance_id": "7601-175351-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0019.flac", "task_type": "lexical_counting", "query": "accustomed", "prompt": "Listen to the entire clip, then report the count sequence for the word \"accustomed\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.88, "one"], [11.74, 12.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.14,\"end\":2.88},{\"word\":\"two\",\"start\":11.74,\"end\":12.36}]"} {"utterance_id": "700-122867-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 1.18, "missus allan"]], "response_mode": "offline", "answer": "[{\"word\":\"missus allan\",\"start\":0.51,\"end\":1.18}]"} {"utterance_id": "6123-59150-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0044.flac", "task_type": "lexical_detection", "query": "images", "prompt": "When the utterance is complete, report all instances of the word \"images\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.13, 5.62, "images"]], "response_mode": "offline", "answer": "[{\"word\":\"images\",\"start\":5.13,\"end\":5.62}]"} {"utterance_id": "6841-88294-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0050.flac", "task_type": "lexical_detection", "query": "table", "prompt": "After the audio ends, extract all occurrences of the word \"table\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 3.79, "table"]], "response_mode": "offline", "answer": "[{\"word\":\"table\",\"start\":3.36,\"end\":3.79}]"} {"utterance_id": "1650-167613-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0046.flac", "task_type": "lexical_detection", "query": "above", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"above\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.02, 1.45, "above"]], "response_mode": "offline", "answer": "[{\"word\":\"above\",\"start\":1.02,\"end\":1.45}]"} {"utterance_id": "8254-84205-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0001.flac", "task_type": "lexical_detection", "query": "glided", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"glided\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.39, "glided"]], "response_mode": "offline", "answer": "[{\"word\":\"glided\",\"start\":0.99,\"end\":1.39}]"} {"utterance_id": "4572-112375-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0000.flac", "task_type": "lexical_detection", "query": "tread", "prompt": "After listening to the whole recording, output every instance of the word \"tread\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.23, 2.85, "tread"]], "response_mode": "offline", "answer": "[{\"word\":\"tread\",\"start\":2.23,\"end\":2.85}]"} {"utterance_id": "6267-65525-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0006.flac", "task_type": "lexical_detection", "query": "baby", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"baby\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.43, "baby"]], "response_mode": "offline", "answer": "[{\"word\":\"baby\",\"start\":2.93,\"end\":3.43}]"} {"utterance_id": "1651-136854-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0032.flac", "task_type": "lexical_detection", "query": "cloth", "prompt": "After listening to the whole recording, output every instance of the word \"cloth\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.62, "cloth"]], "response_mode": "offline", "answer": "[{\"word\":\"cloth\",\"start\":2.07,\"end\":2.62}]"} {"utterance_id": "4323-55228-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0038.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.64, "five"]], "response_mode": "offline", "answer": "[{\"word\":\"five\",\"start\":0.21,\"end\":0.64}]"} {"utterance_id": "7697-245712-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0001.flac", "task_type": "lexical_detection", "query": "mastership", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mastership\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.63, 3.23, "mastership"]], "response_mode": "offline", "answer": "[{\"word\":\"mastership\",\"start\":2.63,\"end\":3.23}]"} {"utterance_id": "4323-55228-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0005.flac", "task_type": "lexical_detection", "query": "wait", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"wait\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.52, 2.82, "wait"]], "response_mode": "offline", "answer": "[{\"word\":\"wait\",\"start\":2.52,\"end\":2.82}]"} {"utterance_id": "3915-57461-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0022.flac", "task_type": "lexical_detection", "query": "become", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"become\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.61, 2.02, "become"]], "response_mode": "offline", "answer": "[{\"word\":\"become\",\"start\":1.61,\"end\":2.02}]"} {"utterance_id": "1650-157641-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0013.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen carefully to the entire clip, then report every instance of a plant name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.04, "tobacco"]], "response_mode": "offline", "answer": "[{\"word\":\"tobacco\",\"start\":1.55,\"end\":2.04}]"} {"utterance_id": "6467-94831-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0012.flac", "task_type": "lexical_counting", "query": "put", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"put\" with word/start/end. If none, reply [].", "labels": [[1.94, 2.18, "one"], [11.25, 11.43, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.94,\"end\":2.18},{\"word\":\"two\",\"start\":11.25,\"end\":11.43}]"} {"utterance_id": "1701-141759-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0005.flac", "task_type": "lexical_detection", "query": "crossing", "prompt": "Once the audio has finished, identify every instance of the word \"crossing\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.23, 3.63, "crossing"]], "response_mode": "offline", "answer": "[{\"word\":\"crossing\",\"start\":3.23,\"end\":3.63}]"} {"utterance_id": "5849-50873-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0035.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Do not answer until the audio ends. Then list each instance of a state or province name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.1, 2.69, "connecticut"]], "response_mode": "offline", "answer": "[{\"word\":\"connecticut\",\"start\":2.1,\"end\":2.69}]"} {"utterance_id": "1585-157660-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0002.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.61, 5.61, "primrose hill"]], "response_mode": "offline", "answer": "[{\"word\":\"primrose hill\",\"start\":4.61,\"end\":5.61}]"} {"utterance_id": "1585-131718-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0019.flac", "task_type": "lexical_detection", "query": "plenty", "prompt": "When the utterance is complete, report all instances of the word \"plenty\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.57, 4.08, "plenty"]], "response_mode": "offline", "answer": "[{\"word\":\"plenty\",\"start\":3.57,\"end\":4.08}]"} {"utterance_id": "4831-18525-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.28, 2.76, "polly"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":2.28,\"end\":2.76}]"} {"utterance_id": "4323-18416-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0011.flac", "task_type": "lexical_detection", "query": "folded", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"folded\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.96, 7.26, "folded"]], "response_mode": "offline", "answer": "[{\"word\":\"folded\",\"start\":6.96,\"end\":7.26}]"} {"utterance_id": "1686-142278-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0016.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Do not answer until the audio ends. Then list each instance of an organization name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.47, 2.82, "church"]], "response_mode": "offline", "answer": "[{\"word\":\"church\",\"start\":2.47,\"end\":2.82}]"} {"utterance_id": "8254-115543-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0007.flac", "task_type": "lexical_counting", "query": "bye", "prompt": "Listen to the entire clip, then report the count sequence for the word \"bye\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.31, "one"], [4.64, 5.17, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.71,\"end\":1.31},{\"word\":\"two\",\"start\":4.64,\"end\":5.17}]"} {"utterance_id": "2506-11278-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.9, "moustaches"]], "response_mode": "offline", "answer": "[{\"word\":\"moustaches\",\"start\":2.27,\"end\":2.9}]"} {"utterance_id": "1255-90407-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 2.05, "avice caro"]], "response_mode": "offline", "answer": "[{\"word\":\"avice caro\",\"start\":1.27,\"end\":2.05}]"} {"utterance_id": "3915-57461-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0025.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.69, 7.2, "goblin"]], "response_mode": "offline", "answer": "[{\"word\":\"goblin\",\"start\":6.69,\"end\":7.2}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_counting", "query": "color", "prompt": "After the audio ends, tally every instance of a color name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[12.08, 12.49, "one"], [26.75, 27.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":12.08,\"end\":12.49},{\"word\":\"two\",\"start\":26.75,\"end\":27.18}]"} {"utterance_id": "1630-96099-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0024.flac", "task_type": "lexical_detection", "query": "cab", "prompt": "When the utterance is complete, report all instances of the word \"cab\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.38, 5.78, "cab"]], "response_mode": "offline", "answer": "[{\"word\":\"cab\",\"start\":5.38,\"end\":5.78}]"} {"utterance_id": "8288-274162-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0026.flac", "task_type": "lexical_detection", "query": "big", "prompt": "When the utterance is complete, report all instances of the word \"big\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.57, "big"]], "response_mode": "offline", "answer": "[{\"word\":\"big\",\"start\":1.33,\"end\":1.57}]"} {"utterance_id": "1650-167613-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0014.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.01, 10.53, "kin"]], "response_mode": "offline", "answer": "[{\"word\":\"kin\",\"start\":10.01,\"end\":10.53}]"} {"utterance_id": "6455-67803-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.62, 1.37, "twenty three"]], "response_mode": "offline", "answer": "[{\"word\":\"twenty three\",\"start\":0.62,\"end\":1.37}]"} {"utterance_id": "8254-115543-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0006.flac", "task_type": "lexical_detection", "query": "cook", "prompt": "After the audio ends, extract all occurrences of the word \"cook\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.84, "cook"]], "response_mode": "offline", "answer": "[{\"word\":\"cook\",\"start\":0.47,\"end\":0.84}]"} {"utterance_id": "6841-88291-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0036.flac", "task_type": "lexical_detection", "query": "scramble", "prompt": "When the utterance is complete, report all instances of the word \"scramble\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.8, "scramble"]], "response_mode": "offline", "answer": "[{\"word\":\"scramble\",\"start\":2.25,\"end\":2.8}]"} {"utterance_id": "7641-96684-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0024.flac", "task_type": "lexical_detection", "query": "lifted", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"lifted\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.91, "lifted"]], "response_mode": "offline", "answer": "[{\"word\":\"lifted\",\"start\":0.53,\"end\":0.91}]"} {"utterance_id": "8288-274162-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0049.flac", "task_type": "lexical_detection", "query": "whose", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"whose\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.09, "whose"]], "response_mode": "offline", "answer": "[{\"word\":\"whose\",\"start\":0.79,\"end\":1.09}]"} {"utterance_id": "6267-65525-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0009.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.31, 5.82, "orange"], [8.8, 9.2, "pea"]], "response_mode": "offline", "answer": "[{\"word\":\"orange\",\"start\":5.31,\"end\":5.82},{\"word\":\"pea\",\"start\":8.8,\"end\":9.2}]"} {"utterance_id": "7697-105815-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.43, 12.04, "limbs"]], "response_mode": "offline", "answer": "[{\"word\":\"limbs\",\"start\":11.43,\"end\":12.04}]"} {"utterance_id": "6467-94831-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0042.flac", "task_type": "lexical_detection", "query": "picture", "prompt": "After listening to the whole recording, output every instance of the word \"picture\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.99, "picture"]], "response_mode": "offline", "answer": "[{\"word\":\"picture\",\"start\":1.44,\"end\":1.99}]"} {"utterance_id": "2506-11267-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0016.flac", "task_type": "lexical_detection", "query": "already", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"already\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.11, 2.69, "already"]], "response_mode": "offline", "answer": "[{\"word\":\"already\",\"start\":2.11,\"end\":2.69}]"} {"utterance_id": "4831-18525-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.88, 3.3, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":2.88,\"end\":3.3}]"} {"utterance_id": "1650-167613-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0001.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.43, "tea"]], "response_mode": "offline", "answer": "[{\"word\":\"tea\",\"start\":2.1,\"end\":2.43}]"} {"utterance_id": "6841-88294-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0019.flac", "task_type": "lexical_detection", "query": "deadly", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"deadly\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.44, 2.78, "deadly"]], "response_mode": "offline", "answer": "[{\"word\":\"deadly\",\"start\":2.44,\"end\":2.78}]"} {"utterance_id": "7601-101619-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.53, 8.04, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":7.53,\"end\":8.04}]"} {"utterance_id": "4831-18525-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.97, 1.19, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":0.97,\"end\":1.19}]"} {"utterance_id": "3915-57461-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0012.flac", "task_type": "lexical_detection", "query": "favouring", "prompt": "After the audio ends, extract all occurrences of the word \"favouring\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.45, "favouring"]], "response_mode": "offline", "answer": "[{\"word\":\"favouring\",\"start\":0.98,\"end\":1.45}]"} {"utterance_id": "4572-112383-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0004.flac", "task_type": "lexical_detection", "query": "element", "prompt": "After listening to the whole recording, output every instance of the word \"element\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.15, 5.95, "element"]], "response_mode": "offline", "answer": "[{\"word\":\"element\",\"start\":5.15,\"end\":5.95}]"} {"utterance_id": "4570-24733-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.35, 7.95, "god"], [10.99, 11.49, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":7.35,\"end\":7.95},{\"word\":\"god\",\"start\":10.99,\"end\":11.49}]"} {"utterance_id": "700-122867-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.3, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":0.97,\"end\":1.3}]"} {"utterance_id": "4831-29134-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0010.flac", "task_type": "semantic_counting", "query": "date", "prompt": "Listen to the full audio. Count every occurrence of a date. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.18, 5.47, "one"], [7.1, 7.45, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.18,\"end\":5.47},{\"word\":\"two\",\"start\":7.1,\"end\":7.45}]"} {"utterance_id": "3660-172183-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.37, 3.68, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":3.37,\"end\":3.68}]"} {"utterance_id": "6123-59150-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0018.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.71, "one"], [3.11, 3.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.52,\"end\":0.71},{\"word\":\"two\",\"start\":3.11,\"end\":3.46}]"} {"utterance_id": "4515-11057-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0049.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.7, 3.19, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":2.7,\"end\":3.19}]"} {"utterance_id": "700-122868-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0037.flac", "task_type": "lexical_detection", "query": "trustworthy", "prompt": "After the audio ends, extract all occurrences of the word \"trustworthy\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.65, "trustworthy"]], "response_mode": "offline", "answer": "[{\"word\":\"trustworthy\",\"start\":1.95,\"end\":2.65}]"} {"utterance_id": "3663-172528-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0034.flac", "task_type": "lexical_detection", "query": "certain", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"certain\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.49, 3.23, "certain"]], "response_mode": "offline", "answer": "[{\"word\":\"certain\",\"start\":2.49,\"end\":3.23}]"} {"utterance_id": "6841-88291-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0002.flac", "task_type": "lexical_detection", "query": "seemingly", "prompt": "When the utterance is complete, report all instances of the word \"seemingly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.29, 0.88, "seemingly"]], "response_mode": "offline", "answer": "[{\"word\":\"seemingly\",\"start\":0.29,\"end\":0.88}]"} {"utterance_id": "5849-50962-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0007.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Do not answer until the audio ends. Then list each instance of a weather term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.84, 3.24, "storm"]], "response_mode": "offline", "answer": "[{\"word\":\"storm\",\"start\":2.84,\"end\":3.24}]"} {"utterance_id": "6841-88294-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0051.flac", "task_type": "lexical_detection", "query": "fire", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fire\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.16, "fire"]], "response_mode": "offline", "answer": "[{\"word\":\"fire\",\"start\":1.84,\"end\":2.16}]"} {"utterance_id": "2506-11267-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.78, 2.14, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":1.78,\"end\":2.14}]"} {"utterance_id": "4323-55228-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0013.flac", "task_type": "lexical_detection", "query": "spread", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"spread\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.59, 4.91, "spread"]], "response_mode": "offline", "answer": "[{\"word\":\"spread\",\"start\":4.59,\"end\":4.91}]"} {"utterance_id": "6467-97061-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0013.flac", "task_type": "lexical_detection", "query": "rays", "prompt": "After the audio ends, extract all occurrences of the word \"rays\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.67, "rays"]], "response_mode": "offline", "answer": "[{\"word\":\"rays\",\"start\":2.39,\"end\":2.67}]"} {"utterance_id": "6841-88291-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0009.flac", "task_type": "lexical_detection", "query": "fence", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"fence\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.66, "fence"]], "response_mode": "offline", "answer": "[{\"word\":\"fence\",\"start\":2.11,\"end\":2.66}]"} {"utterance_id": "4831-29134-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0016.flac", "task_type": "lexical_detection", "query": "stands", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"stands\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.19, 4.53, "stands"]], "response_mode": "offline", "answer": "[{\"word\":\"stands\",\"start\":4.19,\"end\":4.53}]"} {"utterance_id": "1701-141759-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.34, "sonya"], [4.29, 4.65, "sonya"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":0.9,\"end\":1.34},{\"word\":\"sonya\",\"start\":4.29,\"end\":4.65}]"} {"utterance_id": "4323-55228-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.1, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":2.75,\"end\":3.1}]"} {"utterance_id": "1686-142278-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.99, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.51,\"end\":0.99}]"} {"utterance_id": "1650-173551-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0004.flac", "task_type": "lexical_detection", "query": "staircase", "prompt": "When the utterance is complete, report all instances of the word \"staircase\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.29, 2.02, "staircase"]], "response_mode": "offline", "answer": "[{\"word\":\"staircase\",\"start\":1.29,\"end\":2.02}]"} {"utterance_id": "7601-175351-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0027.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.12, 6.68, "mist"]], "response_mode": "offline", "answer": "[{\"word\":\"mist\",\"start\":6.12,\"end\":6.68}]"} {"utterance_id": "4570-14911-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0003.flac", "task_type": "lexical_detection", "query": "procured", "prompt": "After the audio ends, extract all occurrences of the word \"procured\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.38, 5.04, "procured"]], "response_mode": "offline", "answer": "[{\"word\":\"procured\",\"start\":4.38,\"end\":5.04}]"} {"utterance_id": "116-288048-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.91, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":1.56,\"end\":1.91}]"} {"utterance_id": "6841-88294-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0052.flac", "task_type": "lexical_detection", "query": "handy", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"handy\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.68, 1.94, "handy"]], "response_mode": "offline", "answer": "[{\"word\":\"handy\",\"start\":1.68,\"end\":1.94}]"} {"utterance_id": "6455-67804-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0006.flac", "task_type": "lexical_detection", "query": "doesn't", "prompt": "When the utterance is complete, report all instances of the word \"doesn't\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.61, "doesn't"]], "response_mode": "offline", "answer": "[{\"word\":\"doesn't\",\"start\":0.39,\"end\":0.61}]"} {"utterance_id": "1255-138279-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 2.0, "two three"]], "response_mode": "offline", "answer": "[{\"word\":\"two three\",\"start\":0.4,\"end\":2.0}]"} {"utterance_id": "7697-105815-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0031.flac", "task_type": "lexical_detection", "query": "bench", "prompt": "When the utterance is complete, report all instances of the word \"bench\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.2, 9.7, "bench"]], "response_mode": "offline", "answer": "[{\"word\":\"bench\",\"start\":9.2,\"end\":9.7}]"} {"utterance_id": "8254-115543-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0031.flac", "task_type": "lexical_detection", "query": "litter", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"litter\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.62, 0.88, "litter"]], "response_mode": "offline", "answer": "[{\"word\":\"litter\",\"start\":0.62,\"end\":0.88}]"} {"utterance_id": "6267-65525-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.77, 6.35, "twenty"], [7.74, 8.31, "eighteen"]], "response_mode": "offline", "answer": "[{\"word\":\"twenty\",\"start\":5.77,\"end\":6.35},{\"word\":\"eighteen\",\"start\":7.74,\"end\":8.31}]"} {"utterance_id": "700-122867-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0032.flac", "task_type": "lexical_detection", "query": "sensation", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"sensation\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 2.07, "sensation"]], "response_mode": "offline", "answer": "[{\"word\":\"sensation\",\"start\":1.52,\"end\":2.07}]"} {"utterance_id": "1630-96099-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0017.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.49, "hand"], [4.4, 4.77, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":0.21,\"end\":0.49},{\"word\":\"hands\",\"start\":4.4,\"end\":4.77}]"} {"utterance_id": "7697-245712-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.75, 2.07, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":1.75,\"end\":2.07}]"} {"utterance_id": "5849-50962-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0026.flac", "task_type": "lexical_detection", "query": "flossie", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"flossie\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.51, 2.98, "flossie"]], "response_mode": "offline", "answer": "[{\"word\":\"flossie\",\"start\":2.51,\"end\":2.98}]"} {"utterance_id": "1255-90413-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0014.flac", "task_type": "lexical_detection", "query": "uncivil", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"uncivil\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.39, "uncivil"]], "response_mode": "offline", "answer": "[{\"word\":\"uncivil\",\"start\":0.85,\"end\":1.39}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "lexical_counting", "query": "smoking", "prompt": "Listen to the full audio. Count every occurrence of the word \"smoking\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.0, 5.67, "one"], [8.89, 9.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.0,\"end\":5.67},{\"word\":\"two\",\"start\":8.89,\"end\":9.44}]"} {"utterance_id": "1686-142278-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0005.flac", "task_type": "lexical_detection", "query": "hale", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"hale\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.11, "hale"]], "response_mode": "offline", "answer": "[{\"word\":\"hale\",\"start\":0.83,\"end\":1.11}]"} {"utterance_id": "4323-55228-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.1, 1.37, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":1.1,\"end\":1.37}]"} {"utterance_id": "3660-6517-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.02, 3.56, "scientists"], [4.94, 5.21, "cook"]], "response_mode": "offline", "answer": "[{\"word\":\"scientists\",\"start\":3.02,\"end\":3.56},{\"word\":\"cook\",\"start\":4.94,\"end\":5.21}]"} {"utterance_id": "1255-74899-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0014.flac", "task_type": "lexical_counting", "query": "boil", "prompt": "Once the audio has finished, list each occurrence of the word \"boil\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.61, 2.05, "one"], [3.72, 4.08, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.61,\"end\":2.05},{\"word\":\"two\",\"start\":3.72,\"end\":4.08}]"} {"utterance_id": "6455-67804-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.63, 5.33, "octavia's"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia's\",\"start\":4.63,\"end\":5.33}]"} {"utterance_id": "1630-141772-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.67, 3.26, "napoleon"]], "response_mode": "offline", "answer": "[{\"word\":\"napoleon\",\"start\":2.67,\"end\":3.26}]"} {"utterance_id": "4515-11057-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0012.flac", "task_type": "lexical_detection", "query": "hadn't", "prompt": "Once the audio has finished, identify every instance of the word \"hadn't\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.64, 2.97, "hadn't"]], "response_mode": "offline", "answer": "[{\"word\":\"hadn't\",\"start\":2.64,\"end\":2.97}]"} {"utterance_id": "3915-98647-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0010.flac", "task_type": "lexical_detection", "query": "roof", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"roof\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.42, 2.62, "roof"]], "response_mode": "offline", "answer": "[{\"word\":\"roof\",\"start\":2.42,\"end\":2.62}]"} {"utterance_id": "7601-175351-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0006.flac", "task_type": "lexical_detection", "query": "hatteras", "prompt": "After the audio ends, extract all occurrences of the word \"hatteras\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.95, "hatteras"]], "response_mode": "offline", "answer": "[{\"word\":\"hatteras\",\"start\":2.16,\"end\":2.95}]"} {"utterance_id": "6467-94831-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.83, 2.56, "master rover"]], "response_mode": "offline", "answer": "[{\"word\":\"master rover\",\"start\":1.83,\"end\":2.56}]"} {"utterance_id": "3915-57461-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0004.flac", "task_type": "lexical_counting", "query": "boat", "prompt": "Process the full audio offline. For every occurrence of the word \"boat\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.74, 9.06, "one"], [15.6, 15.85, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.74,\"end\":9.06},{\"word\":\"two\",\"start\":15.6,\"end\":15.85}]"} {"utterance_id": "3663-172528-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0004.flac", "task_type": "semantic_detection", "query": "food", "prompt": "You will hear a complete utterance. When it ends, list every instance of a food name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.15, 4.65, "water"]], "response_mode": "offline", "answer": "[{\"word\":\"water\",\"start\":4.15,\"end\":4.65}]"} {"utterance_id": "4323-55228-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0049.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.38, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.19,\"end\":1.38}]"} {"utterance_id": "3663-172528-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0032.flac", "task_type": "lexical_detection", "query": "denied", "prompt": "After the audio ends, extract all occurrences of the word \"denied\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.88, "denied"]], "response_mode": "offline", "answer": "[{\"word\":\"denied\",\"start\":0.33,\"end\":0.88}]"} {"utterance_id": "4153-186223-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0000.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a plant name with fields word, start, end (seconds). If none, reply [].", "labels": [[5.46, 5.81, "pine"]], "response_mode": "offline", "answer": "[{\"word\":\"pine\",\"start\":5.46,\"end\":5.81}]"} {"utterance_id": "4570-14911-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.55, 2.15, "harriers"]], "response_mode": "offline", "answer": "[{\"word\":\"harriers\",\"start\":1.55,\"end\":2.15}]"} {"utterance_id": "4572-112375-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0013.flac", "task_type": "lexical_detection", "query": "leave", "prompt": "After the audio ends, extract all occurrences of the word \"leave\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 3.03, "leave"]], "response_mode": "offline", "answer": "[{\"word\":\"leave\",\"start\":2.39,\"end\":3.03}]"} {"utterance_id": "8254-84205-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0062.flac", "task_type": "lexical_detection", "query": "exactly", "prompt": "When the utterance is complete, report all instances of the word \"exactly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 1.02, "exactly"]], "response_mode": "offline", "answer": "[{\"word\":\"exactly\",\"start\":0.42,\"end\":1.02}]"} {"utterance_id": "4153-186223-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0003.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.94, 8.21, "one"], [9.53, 9.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.94,\"end\":8.21},{\"word\":\"two\",\"start\":9.53,\"end\":9.78}]"} {"utterance_id": "5849-50873-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0014.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.35, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":2.91,\"end\":3.35}]"} {"utterance_id": "7697-245712-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0009.flac", "task_type": "lexical_counting", "query": "animals", "prompt": "When the utterance is complete, report how the word \"animals\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.13, "one"], [4.82, 5.23, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.73,\"end\":1.13},{\"word\":\"two\",\"start\":4.82,\"end\":5.23}]"} {"utterance_id": "6123-59186-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0024.flac", "task_type": "lexical_detection", "query": "enter", "prompt": "When the utterance is complete, report all instances of the word \"enter\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.03, "enter"]], "response_mode": "offline", "answer": "[{\"word\":\"enter\",\"start\":0.71,\"end\":1.03}]"} {"utterance_id": "700-122867-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.79, "diana"]], "response_mode": "offline", "answer": "[{\"word\":\"diana\",\"start\":0.39,\"end\":0.79}]"} {"utterance_id": "1701-141760-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0025.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[3.76, 4.42, "one"], [6.22, 7.12, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.76,\"end\":4.42},{\"word\":\"two\",\"start\":6.22,\"end\":7.12}]"} {"utterance_id": "4153-185072-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0003.flac", "task_type": "lexical_detection", "query": "continuance", "prompt": "Once the audio has finished, identify every instance of the word \"continuance\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.35, 8.09, "continuance"]], "response_mode": "offline", "answer": "[{\"word\":\"continuance\",\"start\":7.35,\"end\":8.09}]"} {"utterance_id": "1686-142278-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0002.flac", "task_type": "lexical_counting", "query": "walk", "prompt": "After the audio ends, tally every instance of the word \"walk\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[7.6, 7.99, "one"], [12.56, 12.83, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.6,\"end\":7.99},{\"word\":\"two\",\"start\":12.56,\"end\":12.83}]"} {"utterance_id": "4831-18525-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0005.flac", "task_type": "lexical_detection", "query": "grow", "prompt": "When the utterance is complete, report all instances of the word \"grow\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 2.85, "grow"]], "response_mode": "offline", "answer": "[{\"word\":\"grow\",\"start\":2.57,\"end\":2.85}]"} {"utterance_id": "1255-138279-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0016.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After the audio ends, extract all occurrences of a weather term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 2.08, "snow"]], "response_mode": "offline", "answer": "[{\"word\":\"snow\",\"start\":1.58,\"end\":2.08}]"} {"utterance_id": "1686-142278-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0098.flac", "task_type": "lexical_detection", "query": "startled", "prompt": "When the utterance is complete, report all instances of the word \"startled\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.28, "startled"]], "response_mode": "offline", "answer": "[{\"word\":\"startled\",\"start\":0.75,\"end\":1.28}]"} {"utterance_id": "7697-105815-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0051.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.9, "jaw bone"]], "response_mode": "offline", "answer": "[{\"word\":\"jaw bone\",\"start\":1.23,\"end\":1.9}]"} {"utterance_id": "4153-186222-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0005.flac", "task_type": "lexical_detection", "query": "surely", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"surely\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.2, 0.71, "surely"]], "response_mode": "offline", "answer": "[{\"word\":\"surely\",\"start\":0.2,\"end\":0.71}]"} {"utterance_id": "1255-90407-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0026.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.99, 7.6, "maid servant"]], "response_mode": "offline", "answer": "[{\"word\":\"maid servant\",\"start\":6.99,\"end\":7.6}]"} {"utterance_id": "1585-157660-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.9, 4.4, "landlady"]], "response_mode": "offline", "answer": "[{\"word\":\"landlady\",\"start\":3.9,\"end\":4.4}]"} {"utterance_id": "1650-173551-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0005.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.72, 4.49, "ostrich"]], "response_mode": "offline", "answer": "[{\"word\":\"ostrich\",\"start\":3.72,\"end\":4.49}]"} {"utterance_id": "6267-53049-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0028.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 2.24, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.58,\"end\":2.24}]"} {"utterance_id": "4515-11057-0114", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0114.flac", "task_type": "lexical_detection", "query": "troubles", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"troubles\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.86, 5.44, "troubles"]], "response_mode": "offline", "answer": "[{\"word\":\"troubles\",\"start\":4.86,\"end\":5.44}]"} {"utterance_id": "4572-64670-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0015.flac", "task_type": "lexical_detection", "query": "conquered", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"conquered\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.82, "conquered"]], "response_mode": "offline", "answer": "[{\"word\":\"conquered\",\"start\":2.35,\"end\":2.82}]"} {"utterance_id": "4515-11057-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0018.flac", "task_type": "lexical_detection", "query": "twice", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"twice\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.61, 4.99, "twice"]], "response_mode": "offline", "answer": "[{\"word\":\"twice\",\"start\":4.61,\"end\":4.99}]"} {"utterance_id": "4153-186223-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0002.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[12.61, 13.14, "elmhurst"]], "response_mode": "offline", "answer": "[{\"word\":\"elmhurst\",\"start\":12.61,\"end\":13.14}]"} {"utterance_id": "3660-172182-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0009.flac", "task_type": "lexical_detection", "query": "replied", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"replied\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.39, "replied"]], "response_mode": "offline", "answer": "[{\"word\":\"replied\",\"start\":0.84,\"end\":1.39}]"} {"utterance_id": "1701-141759-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0021.flac", "task_type": "lexical_detection", "query": "smiled", "prompt": "Process the full audio offline. Find every instance of the word \"smiled\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.57, "smiled"]], "response_mode": "offline", "answer": "[{\"word\":\"smiled\",\"start\":0.95,\"end\":1.57}]"} {"utterance_id": "1255-90413-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0005.flac", "task_type": "lexical_detection", "query": "hitherto", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"hitherto\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.14, 4.73, "hitherto"]], "response_mode": "offline", "answer": "[{\"word\":\"hitherto\",\"start\":4.14,\"end\":4.73}]"} {"utterance_id": "1651-136854-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0014.flac", "task_type": "lexical_detection", "query": "passions", "prompt": "When the utterance is complete, report all instances of the word \"passions\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.96, 3.63, "passions"]], "response_mode": "offline", "answer": "[{\"word\":\"passions\",\"start\":2.96,\"end\":3.63}]"} {"utterance_id": "7697-105817-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0004.flac", "task_type": "lexical_detection", "query": "hopeful", "prompt": "Process the full audio offline. Find every instance of the word \"hopeful\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.7, "hopeful"]], "response_mode": "offline", "answer": "[{\"word\":\"hopeful\",\"start\":1.07,\"end\":1.7}]"} {"utterance_id": "6455-66379-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0013.flac", "task_type": "lexical_detection", "query": "deck", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"deck\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.39, "deck"]], "response_mode": "offline", "answer": "[{\"word\":\"deck\",\"start\":2.1,\"end\":2.39}]"} {"utterance_id": "4515-11057-0102", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0102.flac", "task_type": "lexical_detection", "query": "going", "prompt": "After listening to the whole recording, output every instance of the word \"going\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.17, "going"]], "response_mode": "offline", "answer": "[{\"word\":\"going\",\"start\":0.84,\"end\":1.17}]"} {"utterance_id": "4515-11057-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0055.flac", "task_type": "lexical_detection", "query": "feel", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"feel\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.8, "feel"]], "response_mode": "offline", "answer": "[{\"word\":\"feel\",\"start\":0.48,\"end\":0.8}]"} {"utterance_id": "8254-84205-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0058.flac", "task_type": "lexical_detection", "query": "making", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"making\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.68, "making"]], "response_mode": "offline", "answer": "[{\"word\":\"making\",\"start\":1.33,\"end\":1.68}]"} {"utterance_id": "1585-157660-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 1.24, "james burbage"], [3.36, 3.74, "william"]], "response_mode": "offline", "answer": "[{\"word\":\"james burbage\",\"start\":0.42,\"end\":1.24},{\"word\":\"william\",\"start\":3.36,\"end\":3.74}]"} {"utterance_id": "7641-96252-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0016.flac", "task_type": "lexical_detection", "query": "particular", "prompt": "After the audio ends, extract all occurrences of the word \"particular\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.37, 6.97, "particular"]], "response_mode": "offline", "answer": "[{\"word\":\"particular\",\"start\":6.37,\"end\":6.97}]"} {"utterance_id": "8254-115543-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0023.flac", "task_type": "lexical_detection", "query": "howdah", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"howdah\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.78, "howdah"]], "response_mode": "offline", "answer": "[{\"word\":\"howdah\",\"start\":2.22,\"end\":2.78}]"} {"utterance_id": "7641-96684-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0031.flac", "task_type": "lexical_detection", "query": "husband", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"husband\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.73, 5.25, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":4.73,\"end\":5.25}]"} {"utterance_id": "7697-105815-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0030.flac", "task_type": "lexical_detection", "query": "far", "prompt": "Once the audio has finished, identify every instance of the word \"far\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.59, 0.85, "far"]], "response_mode": "offline", "answer": "[{\"word\":\"far\",\"start\":0.59,\"end\":0.85}]"} {"utterance_id": "3660-6517-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.68, 5.11, "legs"]], "response_mode": "offline", "answer": "[{\"word\":\"legs\",\"start\":4.68,\"end\":5.11}]"} {"utterance_id": "1630-73710-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0008.flac", "task_type": "lexical_detection", "query": "instant", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"instant\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.89, 2.53, "instant"]], "response_mode": "offline", "answer": "[{\"word\":\"instant\",\"start\":1.89,\"end\":2.53}]"} {"utterance_id": "4153-186223-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0017.flac", "task_type": "lexical_detection", "query": "exercise", "prompt": "After listening to the whole recording, output every instance of the word \"exercise\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.47, 4.96, "exercise"]], "response_mode": "offline", "answer": "[{\"word\":\"exercise\",\"start\":4.47,\"end\":4.96}]"} {"utterance_id": "8173-294714-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0050.flac", "task_type": "lexical_detection", "query": "policy", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"policy\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.6, 6.14, "policy"]], "response_mode": "offline", "answer": "[{\"word\":\"policy\",\"start\":5.6,\"end\":6.14}]"} {"utterance_id": "700-122868-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0026.flac", "task_type": "lexical_detection", "query": "happened", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"happened\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.2, "happened"]], "response_mode": "offline", "answer": "[{\"word\":\"happened\",\"start\":0.78,\"end\":1.2}]"} {"utterance_id": "1651-136854-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0028.flac", "task_type": "lexical_detection", "query": "soothing", "prompt": "After the audio ends, extract all occurrences of the word \"soothing\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.65, 9.15, "soothing"]], "response_mode": "offline", "answer": "[{\"word\":\"soothing\",\"start\":8.65,\"end\":9.15}]"} {"utterance_id": "4515-11057-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0024.flac", "task_type": "lexical_detection", "query": "difficulties", "prompt": "When the utterance is complete, report all instances of the word \"difficulties\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.58, 1.27, "difficulties"]], "response_mode": "offline", "answer": "[{\"word\":\"difficulties\",\"start\":0.58,\"end\":1.27}]"} {"utterance_id": "5543-27761-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0019.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.68, 0.94, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.68,\"end\":0.94}]"} {"utterance_id": "7697-245712-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0009.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.13, "animals"], [4.82, 5.23, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":0.73,\"end\":1.13},{\"word\":\"animals\",\"start\":4.82,\"end\":5.23}]"} {"utterance_id": "7697-245715-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0002.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.97, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":1.59,\"end\":1.97}]"} {"utterance_id": "6455-67803-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0028.flac", "task_type": "lexical_detection", "query": "closer", "prompt": "When the utterance is complete, report all instances of the word \"closer\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 2.01, "closer"]], "response_mode": "offline", "answer": "[{\"word\":\"closer\",\"start\":1.49,\"end\":2.01}]"} {"utterance_id": "4323-55228-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0007.flac", "task_type": "lexical_detection", "query": "angel", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"angel\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.44, "angel"]], "response_mode": "offline", "answer": "[{\"word\":\"angel\",\"start\":2.1,\"end\":2.44}]"} {"utterance_id": "6455-67803-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0034.flac", "task_type": "lexical_detection", "query": "fierce", "prompt": "After listening to the whole recording, output every instance of the word \"fierce\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.02, "fierce"]], "response_mode": "offline", "answer": "[{\"word\":\"fierce\",\"start\":0.67,\"end\":1.02}]"} {"utterance_id": "116-288048-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0005.flac", "task_type": "lexical_detection", "query": "friday", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"friday\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.45, 2.04, "friday"]], "response_mode": "offline", "answer": "[{\"word\":\"friday\",\"start\":1.45,\"end\":2.04}]"} {"utterance_id": "4570-24733-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0005.flac", "task_type": "lexical_detection", "query": "limits", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"limits\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.63, "limits"]], "response_mode": "offline", "answer": "[{\"word\":\"limits\",\"start\":1.14,\"end\":1.63}]"} {"utterance_id": "3915-98647-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0001.flac", "task_type": "lexical_counting", "query": "each", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"each\" (word, start, end). If none, reply [].", "labels": [[5.63, 5.85, "one"], [7.46, 7.8, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.63,\"end\":5.85},{\"word\":\"two\",\"start\":7.46,\"end\":7.8}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.63, 1.0, "fourth"]], "response_mode": "offline", "answer": "[{\"word\":\"fourth\",\"start\":0.63,\"end\":1.0}]"} {"utterance_id": "4153-61735-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0006.flac", "task_type": "lexical_detection", "query": "sailing", "prompt": "Process the full audio offline. Find every instance of the word \"sailing\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.46, 6.87, "sailing"]], "response_mode": "offline", "answer": "[{\"word\":\"sailing\",\"start\":6.46,\"end\":6.87}]"} {"utterance_id": "700-122867-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0040.flac", "task_type": "lexical_detection", "query": "head", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"head\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.87, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":0.66,\"end\":0.87}]"} {"utterance_id": "7697-105815-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0015.flac", "task_type": "lexical_detection", "query": "door", "prompt": "Once the audio has finished, identify every instance of the word \"door\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.56, 2.93, "door"]], "response_mode": "offline", "answer": "[{\"word\":\"door\",\"start\":2.56,\"end\":2.93}]"} {"utterance_id": "6599-38590-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.58, "francis"]], "response_mode": "offline", "answer": "[{\"word\":\"francis\",\"start\":0.94,\"end\":1.58}]"} {"utterance_id": "4323-18416-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0004.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.21, 1.52, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":1.21,\"end\":1.52}]"} {"utterance_id": "3663-172528-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 3.15, "maestro bernardino"]], "response_mode": "offline", "answer": "[{\"word\":\"maestro bernardino\",\"start\":1.63,\"end\":3.15}]"} {"utterance_id": "4831-25894-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0001.flac", "task_type": "lexical_detection", "query": "christmas", "prompt": "After the audio ends, extract all occurrences of the word \"christmas\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.37, "christmas"]], "response_mode": "offline", "answer": "[{\"word\":\"christmas\",\"start\":2.75,\"end\":3.37}]"} {"utterance_id": "7641-96684-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0026.flac", "task_type": "lexical_detection", "query": "beside", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"beside\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.17, 3.54, "beside"]], "response_mode": "offline", "answer": "[{\"word\":\"beside\",\"start\":3.17,\"end\":3.54}]"} {"utterance_id": "5543-27761-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0052.flac", "task_type": "lexical_detection", "query": "understand", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"understand\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.96, "understand"]], "response_mode": "offline", "answer": "[{\"word\":\"understand\",\"start\":0.39,\"end\":0.96}]"} {"utterance_id": "4515-11057-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0030.flac", "task_type": "lexical_detection", "query": "encouragement", "prompt": "After listening to the whole recording, output every instance of the word \"encouragement\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 3.17, "encouragement"]], "response_mode": "offline", "answer": "[{\"word\":\"encouragement\",\"start\":2.34,\"end\":3.17}]"} {"utterance_id": "116-288047-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0012.flac", "task_type": "lexical_detection", "query": "discretion", "prompt": "Process the full audio offline. Find every instance of the word \"discretion\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.97, "discretion"]], "response_mode": "offline", "answer": "[{\"word\":\"discretion\",\"start\":4.23,\"end\":4.97}]"} {"utterance_id": "7697-105815-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0001.flac", "task_type": "lexical_detection", "query": "christian", "prompt": "Once the audio has finished, identify every instance of the word \"christian\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.26, 0.89, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.26,\"end\":0.89}]"} {"utterance_id": "8254-115543-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0028.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.56, 7.0, "suppers"], [8.05, 8.3, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"suppers\",\"start\":6.56,\"end\":7.0},{\"word\":\"food\",\"start\":8.05,\"end\":8.3}]"} {"utterance_id": "5849-50873-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 1.43, "captain wadsworth"]], "response_mode": "offline", "answer": "[{\"word\":\"captain wadsworth\",\"start\":0.49,\"end\":1.43}]"} {"utterance_id": "8254-84205-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0046.flac", "task_type": "lexical_detection", "query": "doctor", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"doctor\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.31, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":1.88,\"end\":2.31}]"} {"utterance_id": "4515-11057-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0079.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.9, "parents"]], "response_mode": "offline", "answer": "[{\"word\":\"parents\",\"start\":4.23,\"end\":4.9}]"} {"utterance_id": "8173-294714-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0013.flac", "task_type": "lexical_detection", "query": "second", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"second\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.74, 5.12, "second"]], "response_mode": "offline", "answer": "[{\"word\":\"second\",\"start\":4.74,\"end\":5.12}]"} {"utterance_id": "8254-84205-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0067.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.1, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.74,\"end\":2.1}]"} {"utterance_id": "4831-18525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0013.flac", "task_type": "lexical_detection", "query": "yours", "prompt": "After listening to the whole recording, output every instance of the word \"yours\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.85, "yours"]], "response_mode": "offline", "answer": "[{\"word\":\"yours\",\"start\":0.62,\"end\":0.85}]"} {"utterance_id": "6267-65525-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0033.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, list every instance of a plant name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.78, 5.3, "thorns"]], "response_mode": "offline", "answer": "[{\"word\":\"thorns\",\"start\":4.78,\"end\":5.3}]"} {"utterance_id": "6267-53049-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0026.flac", "task_type": "semantic_detection", "query": "color", "prompt": "When the utterance is complete, report all instances of a color name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.93, 6.74, "rosebud"]], "response_mode": "offline", "answer": "[{\"word\":\"rosebud\",\"start\":5.93,\"end\":6.74}]"} {"utterance_id": "3915-57461-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0016.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.3, 4.84, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":4.3,\"end\":4.84}]"} {"utterance_id": "7697-105815-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0033.flac", "task_type": "lexical_detection", "query": "thanks", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"thanks\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.01, "thanks"]], "response_mode": "offline", "answer": "[{\"word\":\"thanks\",\"start\":2.69,\"end\":3.01}]"} {"utterance_id": "6841-88291-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0027.flac", "task_type": "lexical_detection", "query": "extent", "prompt": "Once the audio has finished, identify every instance of the word \"extent\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.8, 2.23, "extent"]], "response_mode": "offline", "answer": "[{\"word\":\"extent\",\"start\":1.8,\"end\":2.23}]"} {"utterance_id": "6841-88294-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0009.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of a country name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[11.79, 12.44, "barbaree"]], "response_mode": "offline", "answer": "[{\"word\":\"barbaree\",\"start\":11.79,\"end\":12.44}]"} {"utterance_id": "6267-65525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0022.flac", "task_type": "lexical_detection", "query": "plainfield", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"plainfield\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.55, "plainfield"]], "response_mode": "offline", "answer": "[{\"word\":\"plainfield\",\"start\":1.97,\"end\":2.55}]"} {"utterance_id": "4515-11057-0081", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0081.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.68, 3.16, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.68,\"end\":3.16}]"} {"utterance_id": "8173-294714-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.29, 2.2, "mister van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"mister van brandt\",\"start\":1.29,\"end\":2.2}]"} {"utterance_id": "4570-56594-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0006.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Process the full audio offline. Find every instance of a country name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.07, 9.68, "mexicans"]], "response_mode": "offline", "answer": "[{\"word\":\"mexicans\",\"start\":9.07,\"end\":9.68}]"} {"utterance_id": "8254-84205-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0061.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After the audio ends, extract all occurrences of a time of day. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.57, "dusk"]], "response_mode": "offline", "answer": "[{\"word\":\"dusk\",\"start\":1.28,\"end\":1.57}]"} {"utterance_id": "1686-142278-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0066.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.32, 1.59, "pay"]], "response_mode": "offline", "answer": "[{\"word\":\"pay\",\"start\":1.32,\"end\":1.59}]"} {"utterance_id": "8254-115543-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0036.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.72, 5.02, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":4.72,\"end\":5.02}]"} {"utterance_id": "116-288045-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0010.flac", "task_type": "lexical_detection", "query": "god", "prompt": "After listening to the whole recording, output every instance of the word \"god\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.47, 2.92, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":2.47,\"end\":2.92}]"} {"utterance_id": "5849-50962-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0008.flac", "task_type": "lexical_detection", "query": "daylight", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"daylight\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.2, "daylight"]], "response_mode": "offline", "answer": "[{\"word\":\"daylight\",\"start\":0.77,\"end\":1.2}]"} {"utterance_id": "7601-101622-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.66, 11.11, "blood"]], "response_mode": "offline", "answer": "[{\"word\":\"blood\",\"start\":10.66,\"end\":11.11}]"} {"utterance_id": "3915-98647-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0012.flac", "task_type": "lexical_detection", "query": "pole", "prompt": "Process the full audio offline. Find every instance of the word \"pole\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[19.79, 20.04, "pole"]], "response_mode": "offline", "answer": "[{\"word\":\"pole\",\"start\":19.79,\"end\":20.04}]"} {"utterance_id": "3915-57461-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0016.flac", "task_type": "lexical_detection", "query": "reached", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"reached\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.43, "reached"]], "response_mode": "offline", "answer": "[{\"word\":\"reached\",\"start\":2.12,\"end\":2.43}]"} {"utterance_id": "1255-138279-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0022.flac", "task_type": "lexical_detection", "query": "course", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"course\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 1.69, "course"]], "response_mode": "offline", "answer": "[{\"word\":\"course\",\"start\":1.42,\"end\":1.69}]"} {"utterance_id": "7601-101622-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.26, 5.57, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":5.26,\"end\":5.57}]"} {"utterance_id": "4515-11057-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0041.flac", "task_type": "lexical_detection", "query": "soft", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"soft\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.72, 3.21, "soft"]], "response_mode": "offline", "answer": "[{\"word\":\"soft\",\"start\":2.72,\"end\":3.21}]"} {"utterance_id": "1585-131718-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0040.flac", "task_type": "lexical_detection", "query": "matter", "prompt": "Process the full audio offline. Find every instance of the word \"matter\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.76, "matter"]], "response_mode": "offline", "answer": "[{\"word\":\"matter\",\"start\":1.39,\"end\":1.76}]"} {"utterance_id": "4153-61735-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0005.flac", "task_type": "lexical_detection", "query": "copse", "prompt": "Process the full audio offline. Find every instance of the word \"copse\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.13, 1.55, "copse"]], "response_mode": "offline", "answer": "[{\"word\":\"copse\",\"start\":1.13,\"end\":1.55}]"} {"utterance_id": "6467-97061-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0014.flac", "task_type": "semantic_detection", "query": "color", "prompt": "You will hear a complete utterance. When it ends, list every instance of a color name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.04, 4.37, "white"]], "response_mode": "offline", "answer": "[{\"word\":\"white\",\"start\":4.04,\"end\":4.37}]"} {"utterance_id": "4515-11057-0085", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0085.flac", "task_type": "lexical_detection", "query": "sobbing", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"sobbing\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.33, "sobbing"]], "response_mode": "offline", "answer": "[{\"word\":\"sobbing\",\"start\":4.82,\"end\":5.33}]"} {"utterance_id": "6123-59150-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0040.flac", "task_type": "lexical_detection", "query": "bears", "prompt": "After the audio ends, extract all occurrences of the word \"bears\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.57, "bears"]], "response_mode": "offline", "answer": "[{\"word\":\"bears\",\"start\":2.25,\"end\":2.57}]"} {"utterance_id": "6467-94831-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0034.flac", "task_type": "lexical_detection", "query": "square", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"square\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.36, 4.94, "square"]], "response_mode": "offline", "answer": "[{\"word\":\"square\",\"start\":4.36,\"end\":4.94}]"} {"utterance_id": "4323-55228-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0036.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[5.71, 5.95, "kate"]], "response_mode": "offline", "answer": "[{\"word\":\"kate\",\"start\":5.71,\"end\":5.95}]"} {"utterance_id": "4515-11057-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0067.flac", "task_type": "lexical_detection", "query": "decided", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"decided\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.57, 2.18, "decided"]], "response_mode": "offline", "answer": "[{\"word\":\"decided\",\"start\":1.57,\"end\":2.18}]"} {"utterance_id": "3663-172528-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0012.flac", "task_type": "lexical_detection", "query": "bottle", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"bottle\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.74, 7.28, "bottle"]], "response_mode": "offline", "answer": "[{\"word\":\"bottle\",\"start\":6.74,\"end\":7.28}]"} {"utterance_id": "8254-115543-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0030.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.75, "supper"]], "response_mode": "offline", "answer": "[{\"word\":\"supper\",\"start\":0.5,\"end\":0.75}]"} {"utterance_id": "6841-88294-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0044.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.17, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.02,\"end\":1.17}]"} {"utterance_id": "6455-67804-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0015.flac", "task_type": "lexical_detection", "query": "friend", "prompt": "Process the full audio offline. Find every instance of the word \"friend\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.19, 0.4, "friend"]], "response_mode": "offline", "answer": "[{\"word\":\"friend\",\"start\":0.19,\"end\":0.4}]"} {"utterance_id": "7601-175351-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0018.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.99, 7.71, "twenty four"]], "response_mode": "offline", "answer": "[{\"word\":\"twenty four\",\"start\":6.99,\"end\":7.71}]"} {"utterance_id": "1686-142278-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0010.flac", "task_type": "lexical_detection", "query": "serious", "prompt": "Once the audio has finished, identify every instance of the word \"serious\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.8, 1.25, "serious"]], "response_mode": "offline", "answer": "[{\"word\":\"serious\",\"start\":0.8,\"end\":1.25}]"} {"utterance_id": "1686-142278-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0034.flac", "task_type": "lexical_counting", "query": "read", "prompt": "Listen to the full audio. Count every occurrence of the word \"read\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.14, "one"], [3.42, 3.77, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.88,\"end\":1.14},{\"word\":\"two\",\"start\":3.42,\"end\":3.77}]"} {"utterance_id": "700-122868-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0004.flac", "task_type": "lexical_detection", "query": "earnestly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"earnestly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.9, "earnestly"]], "response_mode": "offline", "answer": "[{\"word\":\"earnestly\",\"start\":3.33,\"end\":3.9}]"} {"utterance_id": "5849-50873-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0025.flac", "task_type": "lexical_detection", "query": "windows", "prompt": "Process the full audio offline. Find every instance of the word \"windows\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.32, "windows"]], "response_mode": "offline", "answer": "[{\"word\":\"windows\",\"start\":2.75,\"end\":3.32}]"} {"utterance_id": "5849-50873-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0039.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.61, 3.88, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":3.61,\"end\":3.88}]"} {"utterance_id": "116-288047-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.53, 2.87, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":2.53,\"end\":2.87}]"} {"utterance_id": "4153-185072-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.9, 9.32, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":8.9,\"end\":9.32}]"} {"utterance_id": "6599-38590-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 2.96, "francis"]], "response_mode": "offline", "answer": "[{\"word\":\"francis\",\"start\":2.4,\"end\":2.96}]"} {"utterance_id": "700-122866-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.39, "junior avonlea"]], "response_mode": "offline", "answer": "[{\"word\":\"junior avonlea\",\"start\":0.4,\"end\":1.39}]"} {"utterance_id": "6267-65525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.6, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":3.14,\"end\":3.6}]"} {"utterance_id": "6599-38591-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0002.flac", "task_type": "lexical_detection", "query": "nine", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"nine\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.26, 3.67, "nine"]], "response_mode": "offline", "answer": "[{\"word\":\"nine\",\"start\":3.26,\"end\":3.67}]"} {"utterance_id": "6467-94831-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0039.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.42, 2.02, "donkey"]], "response_mode": "offline", "answer": "[{\"word\":\"donkey\",\"start\":1.42,\"end\":2.02}]"} {"utterance_id": "4153-186222-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0005.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a number (word, start, end). If none, reply [].", "labels": [[0.71, 0.91, "one"], [3.77, 4.0, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.71,\"end\":0.91},{\"word\":\"two\",\"start\":3.77,\"end\":4.0}]"} {"utterance_id": "7641-96684-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0012.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.79, 5.2, "corn"]], "response_mode": "offline", "answer": "[{\"word\":\"corn\",\"start\":4.79,\"end\":5.2}]"} {"utterance_id": "7641-96670-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0020.flac", "task_type": "lexical_detection", "query": "feet", "prompt": "After listening to the whole recording, output every instance of the word \"feet\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.28, 4.43, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":4.28,\"end\":4.43}]"} {"utterance_id": "7641-96684-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.46, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":2.07,\"end\":2.46}]"} {"utterance_id": "6267-65525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0021.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.23, 1.02, "one"], [11.84, 12.32, "two"], [19.09, 19.67, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.23,\"end\":1.02},{\"word\":\"two\",\"start\":11.84,\"end\":12.32},{\"word\":\"three\",\"start\":19.09,\"end\":19.67}]"} {"utterance_id": "6841-88291-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.8, 3.22, "cattle"]], "response_mode": "offline", "answer": "[{\"word\":\"cattle\",\"start\":2.8,\"end\":3.22}]"} {"utterance_id": "8254-115543-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0021.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.89, 7.05, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.89,\"end\":7.05}]"} {"utterance_id": "4515-11057-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0064.flac", "task_type": "lexical_detection", "query": "isn't", "prompt": "Process the full audio offline. Find every instance of the word \"isn't\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 0.99, "isn't"]], "response_mode": "offline", "answer": "[{\"word\":\"isn't\",\"start\":0.72,\"end\":0.99}]"} {"utterance_id": "3660-6517-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[3.46, 3.89, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":3.46,\"end\":3.89}]"} {"utterance_id": "3663-172005-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0005.flac", "task_type": "lexical_counting", "query": "tired", "prompt": "Listen to the full audio. Count every occurrence of the word \"tired\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.07, "one"], [2.89, 3.37, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":1.07},{\"word\":\"two\",\"start\":2.89,\"end\":3.37}]"} {"utterance_id": "1650-167613-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0050.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.51, "gorki"]], "response_mode": "offline", "answer": "[{\"word\":\"gorki\",\"start\":0.9,\"end\":1.51}]"} {"utterance_id": "8254-115543-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0027.flac", "task_type": "lexical_detection", "query": "talked", "prompt": "Once the audio has finished, identify every instance of the word \"talked\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.13, 2.44, "talked"]], "response_mode": "offline", "answer": "[{\"word\":\"talked\",\"start\":2.13,\"end\":2.44}]"} {"utterance_id": "700-122867-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0023.flac", "task_type": "lexical_detection", "query": "looking", "prompt": "Process the full audio offline. Find every instance of the word \"looking\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.68, "looking"]], "response_mode": "offline", "answer": "[{\"word\":\"looking\",\"start\":3.39,\"end\":3.68}]"} {"utterance_id": "8288-274162-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0052.flac", "task_type": "lexical_detection", "query": "mouth", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"mouth\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.7, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":2.31,\"end\":2.7}]"} {"utterance_id": "3660-172182-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0033.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.25, 3.73, "severn"]], "response_mode": "offline", "answer": "[{\"word\":\"severn\",\"start\":3.25,\"end\":3.73}]"} {"utterance_id": "4153-186222-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0029.flac", "task_type": "lexical_detection", "query": "things", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"things\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.18, 2.66, "things"]], "response_mode": "offline", "answer": "[{\"word\":\"things\",\"start\":2.18,\"end\":2.66}]"} {"utterance_id": "2506-11278-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0033.flac", "task_type": "lexical_detection", "query": "confidence", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"confidence\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.5, "confidence"]], "response_mode": "offline", "answer": "[{\"word\":\"confidence\",\"start\":0.87,\"end\":1.5}]"} {"utterance_id": "4323-18416-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0011.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.69, 1.33, "grandpapa"]], "response_mode": "offline", "answer": "[{\"word\":\"grandpapa\",\"start\":0.69,\"end\":1.33}]"} {"utterance_id": "4831-25894-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0004.flac", "task_type": "lexical_detection", "query": "bites", "prompt": "Process the full audio offline. Find every instance of the word \"bites\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.72, 3.09, "bites"]], "response_mode": "offline", "answer": "[{\"word\":\"bites\",\"start\":2.72,\"end\":3.09}]"} {"utterance_id": "1650-167613-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0002.flac", "task_type": "lexical_detection", "query": "stout", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"stout\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.24, 3.65, "stout"]], "response_mode": "offline", "answer": "[{\"word\":\"stout\",\"start\":3.24,\"end\":3.65}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "semantic_counting", "query": "money", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a money or currency term (word, start, end). If none, reply [].", "labels": [[2.95, 3.35, "one"], [4.44, 4.75, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.95,\"end\":3.35},{\"word\":\"two\",\"start\":4.44,\"end\":4.75}]"} {"utterance_id": "116-288047-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0013.flac", "task_type": "lexical_detection", "query": "matured", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"matured\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.94, 2.46, "matured"]], "response_mode": "offline", "answer": "[{\"word\":\"matured\",\"start\":1.94,\"end\":2.46}]"} {"utterance_id": "4515-11057-0084", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0084.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.61, 11.11, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":10.61,\"end\":11.11}]"} {"utterance_id": "3663-172528-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0021.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.22, 13.72, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":13.22,\"end\":13.72}]"} {"utterance_id": "6267-65525-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0055.flac", "task_type": "lexical_detection", "query": "think", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"think\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.4, 0.72, "think"]], "response_mode": "offline", "answer": "[{\"word\":\"think\",\"start\":0.4,\"end\":0.72}]"} {"utterance_id": "3915-57461-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0003.flac", "task_type": "lexical_detection", "query": "embarked", "prompt": "Once the audio has finished, identify every instance of the word \"embarked\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.52, 10.01, "embarked"]], "response_mode": "offline", "answer": "[{\"word\":\"embarked\",\"start\":9.52,\"end\":10.01}]"} {"utterance_id": "6467-62797-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0004.flac", "task_type": "lexical_detection", "query": "blood", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"blood\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.32, 0.65, "blood"]], "response_mode": "offline", "answer": "[{\"word\":\"blood\",\"start\":0.32,\"end\":0.65}]"} {"utterance_id": "7641-96684-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0007.flac", "task_type": "lexical_detection", "query": "joy", "prompt": "After the audio ends, extract all occurrences of the word \"joy\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.47, 3.85, "joy"]], "response_mode": "offline", "answer": "[{\"word\":\"joy\",\"start\":3.47,\"end\":3.85}]"} {"utterance_id": "116-288045-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0029.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.88, 3.38, "hand"], [4.69, 5.22, "feet"], [10.48, 11.0, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":2.88,\"end\":3.38},{\"word\":\"feet\",\"start\":4.69,\"end\":5.22},{\"word\":\"eyes\",\"start\":10.48,\"end\":11.0}]"} {"utterance_id": "7601-175351-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0016.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.93, 4.45, "seventy"]], "response_mode": "offline", "answer": "[{\"word\":\"seventy\",\"start\":3.93,\"end\":4.45}]"} {"utterance_id": "3660-172183-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0018.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.45, "dawn"]], "response_mode": "offline", "answer": "[{\"word\":\"dawn\",\"start\":1.21,\"end\":1.45}]"} {"utterance_id": "1651-136854-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0003.flac", "task_type": "lexical_detection", "query": "fixed", "prompt": "After listening to the whole recording, output every instance of the word \"fixed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.29, 5.62, "fixed"]], "response_mode": "offline", "answer": "[{\"word\":\"fixed\",\"start\":5.29,\"end\":5.62}]"} {"utterance_id": "8173-294714-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0011.flac", "task_type": "lexical_detection", "query": "forget", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"forget\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.44, 1.85, "forget"]], "response_mode": "offline", "answer": "[{\"word\":\"forget\",\"start\":1.44,\"end\":1.85}]"} {"utterance_id": "700-122868-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0005.flac", "task_type": "lexical_detection", "query": "duck", "prompt": "After the audio ends, extract all occurrences of the word \"duck\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.55, 7.78, "duck"]], "response_mode": "offline", "answer": "[{\"word\":\"duck\",\"start\":7.55,\"end\":7.78}]"} {"utterance_id": "116-288045-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0021.flac", "task_type": "lexical_detection", "query": "incense", "prompt": "After the audio ends, extract all occurrences of the word \"incense\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 2.24, "incense"]], "response_mode": "offline", "answer": "[{\"word\":\"incense\",\"start\":1.4,\"end\":2.24}]"} {"utterance_id": "700-122867-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0038.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.57, 5.78, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.57,\"end\":5.78}]"} {"utterance_id": "6467-94831-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0022.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a color name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.99, 2.3, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":1.99,\"end\":2.3}]"} {"utterance_id": "3915-98647-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.27, 6.6, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":6.27,\"end\":6.6}]"} {"utterance_id": "6123-59186-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0011.flac", "task_type": "lexical_detection", "query": "stroked", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"stroked\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[18.52, 18.98, "stroked"]], "response_mode": "offline", "answer": "[{\"word\":\"stroked\",\"start\":18.52,\"end\":18.98}]"} {"utterance_id": "5849-50873-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0014.flac", "task_type": "lexical_detection", "query": "uncle", "prompt": "Process the full audio offline. Find every instance of the word \"uncle\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.35, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":2.91,\"end\":3.35}]"} {"utterance_id": "116-288048-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.61, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":2.91,\"end\":3.61}]"} {"utterance_id": "5543-27761-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0009.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 2.01, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.71,\"end\":2.01}]"} {"utterance_id": "6267-53049-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.02, 4.36, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":4.02,\"end\":4.36}]"} {"utterance_id": "116-288045-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0006.flac", "task_type": "lexical_detection", "query": "idol", "prompt": "After listening to the whole recording, output every instance of the word \"idol\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.86, "idol"]], "response_mode": "offline", "answer": "[{\"word\":\"idol\",\"start\":0.45,\"end\":0.86}]"} {"utterance_id": "3660-172183-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0004.flac", "task_type": "lexical_detection", "query": "awoke", "prompt": "Process the full audio offline. Find every instance of the word \"awoke\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.51, 2.82, "awoke"]], "response_mode": "offline", "answer": "[{\"word\":\"awoke\",\"start\":2.51,\"end\":2.82}]"} {"utterance_id": "8288-274162-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0064.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.16, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":0.75,\"end\":1.16}]"} {"utterance_id": "116-288046-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0009.flac", "task_type": "lexical_detection", "query": "wish", "prompt": "Process the full audio offline. Find every instance of the word \"wish\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 0.77, "wish"]], "response_mode": "offline", "answer": "[{\"word\":\"wish\",\"start\":0.61,\"end\":0.77}]"} {"utterance_id": "4323-18416-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0027.flac", "task_type": "lexical_detection", "query": "phronsie", "prompt": "Once the audio has finished, identify every instance of the word \"phronsie\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.2, 1.66, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.2,\"end\":1.66}]"} {"utterance_id": "6841-88294-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0016.flac", "task_type": "lexical_detection", "query": "bother", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"bother\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.19, "bother"]], "response_mode": "offline", "answer": "[{\"word\":\"bother\",\"start\":0.91,\"end\":1.19}]"} {"utterance_id": "1650-167613-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0036.flac", "task_type": "lexical_detection", "query": "foes", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"foes\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.23, 1.71, "foes"]], "response_mode": "offline", "answer": "[{\"word\":\"foes\",\"start\":1.23,\"end\":1.71}]"} {"utterance_id": "1630-141772-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.31, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":0.5,\"end\":1.31}]"} {"utterance_id": "5849-50962-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.8, "bert"]], "response_mode": "offline", "answer": "[{\"word\":\"bert\",\"start\":0.39,\"end\":0.8}]"} {"utterance_id": "2506-11278-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0009.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.27, 4.77, "york"]], "response_mode": "offline", "answer": "[{\"word\":\"york\",\"start\":4.27,\"end\":4.77}]"} {"utterance_id": "3915-98647-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0018.flac", "task_type": "lexical_detection", "query": "younger", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"younger\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.84, 1.15, "younger"]], "response_mode": "offline", "answer": "[{\"word\":\"younger\",\"start\":0.84,\"end\":1.15}]"} {"utterance_id": "6267-53049-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.95, 5.35, "arm"], [10.43, 10.83, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":4.95,\"end\":5.35},{\"word\":\"face\",\"start\":10.43,\"end\":10.83}]"} {"utterance_id": "3915-98647-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0033.flac", "task_type": "lexical_detection", "query": "feels", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"feels\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.94, 5.25, "feels"]], "response_mode": "offline", "answer": "[{\"word\":\"feels\",\"start\":4.94,\"end\":5.25}]"} {"utterance_id": "6841-88294-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0068.flac", "task_type": "lexical_detection", "query": "third", "prompt": "When the utterance is complete, report all instances of the word \"third\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 0.57, "third"], [2.32, 2.58, "third"]], "response_mode": "offline", "answer": "[{\"word\":\"third\",\"start\":0.36,\"end\":0.57},{\"word\":\"third\",\"start\":2.32,\"end\":2.58}]"} {"utterance_id": "6599-38591-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0013.flac", "task_type": "lexical_detection", "query": "everywhere", "prompt": "After the audio ends, extract all occurrences of the word \"everywhere\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.7, "everywhere"]], "response_mode": "offline", "answer": "[{\"word\":\"everywhere\",\"start\":1.23,\"end\":1.7}]"} {"utterance_id": "4515-11057-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0008.flac", "task_type": "lexical_detection", "query": "biffen", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"biffen\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.52, 1.86, "biffen"]], "response_mode": "offline", "answer": "[{\"word\":\"biffen\",\"start\":1.52,\"end\":1.86}]"} {"utterance_id": "4323-13259-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0017.flac", "task_type": "lexical_detection", "query": "page", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"page\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.65, 1.97, "page"]], "response_mode": "offline", "answer": "[{\"word\":\"page\",\"start\":1.65,\"end\":1.97}]"} {"utterance_id": "6467-94831-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.29, 0.61, "sam"]], "response_mode": "offline", "answer": "[{\"word\":\"sam\",\"start\":0.29,\"end\":0.61}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.95, "william"], [9.2, 9.82, "morpheus"]], "response_mode": "offline", "answer": "[{\"word\":\"william\",\"start\":0.4,\"end\":0.95},{\"word\":\"morpheus\",\"start\":9.2,\"end\":9.82}]"} {"utterance_id": "4572-64670-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0001.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.69, "brother"], [2.78, 3.23, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":1.21,\"end\":1.69},{\"word\":\"father\",\"start\":2.78,\"end\":3.23}]"} {"utterance_id": "6455-66379-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0018.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.01, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.57,\"end\":2.01}]"} {"utterance_id": "1650-167613-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.12, 5.98, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":5.12,\"end\":5.98}]"} {"utterance_id": "4323-55228-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 1.74, "miss billy miss billy"]], "response_mode": "offline", "answer": "[{\"word\":\"miss billy miss billy\",\"start\":0.34,\"end\":1.74}]"} {"utterance_id": "2506-11278-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0030.flac", "task_type": "lexical_detection", "query": "man's", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"man's\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.12, 1.46, "man's"]], "response_mode": "offline", "answer": "[{\"word\":\"man's\",\"start\":1.12,\"end\":1.46}]"} {"utterance_id": "5543-27761-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0037.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.4, 2.81, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.4,\"end\":2.81}]"} {"utterance_id": "4515-11057-0108", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0108.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[0.76, 1.07, "sister"]], "response_mode": "offline", "answer": "[{\"word\":\"sister\",\"start\":0.76,\"end\":1.07}]"} {"utterance_id": "6841-88294-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0048.flac", "task_type": "lexical_detection", "query": "treasure", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"treasure\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.56, 3.04, "treasure"]], "response_mode": "offline", "answer": "[{\"word\":\"treasure\",\"start\":2.56,\"end\":3.04}]"} {"utterance_id": "1701-141760-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.22, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":3.82,\"end\":4.22}]"} {"utterance_id": "1585-157660-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0009.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen to the full audio. After it finishes, find every instance of a money or currency term. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.89, 3.11, "coin"]], "response_mode": "offline", "answer": "[{\"word\":\"coin\",\"start\":2.89,\"end\":3.11}]"} {"utterance_id": "6455-67804-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0013.flac", "task_type": "lexical_detection", "query": "belinda", "prompt": "Once the audio has finished, identify every instance of the word \"belinda\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.2, 1.56, "belinda"]], "response_mode": "offline", "answer": "[{\"word\":\"belinda\",\"start\":1.2,\"end\":1.56}]"} {"utterance_id": "5849-50873-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0007.flac", "task_type": "lexical_detection", "query": "consented", "prompt": "When the utterance is complete, report all instances of the word \"consented\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.21, "consented"]], "response_mode": "offline", "answer": "[{\"word\":\"consented\",\"start\":0.65,\"end\":1.21}]"} {"utterance_id": "1650-157641-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0010.flac", "task_type": "lexical_detection", "query": "guinea", "prompt": "Once the audio has finished, identify every instance of the word \"guinea\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.36, 1.72, "guinea"]], "response_mode": "offline", "answer": "[{\"word\":\"guinea\",\"start\":1.36,\"end\":1.72}]"} {"utterance_id": "6467-94831-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0045.flac", "task_type": "lexical_detection", "query": "joke", "prompt": "When the utterance is complete, report all instances of the word \"joke\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.38, "joke"]], "response_mode": "offline", "answer": "[{\"word\":\"joke\",\"start\":1.08,\"end\":1.38}]"} {"utterance_id": "116-288048-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0006.flac", "task_type": "lexical_counting", "query": "spring", "prompt": "After listening to the whole recording, output timed count words for the word \"spring\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.17, 2.52, "one"], [4.54, 4.88, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.17,\"end\":2.52},{\"word\":\"two\",\"start\":4.54,\"end\":4.88}]"} {"utterance_id": "4153-186222-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0000.flac", "task_type": "lexical_detection", "query": "quarter", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"quarter\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.84, "quarter"]], "response_mode": "offline", "answer": "[{\"word\":\"quarter\",\"start\":1.4,\"end\":1.84}]"} {"utterance_id": "3915-57461-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.5, "tokubei"]], "response_mode": "offline", "answer": "[{\"word\":\"tokubei\",\"start\":3.94,\"end\":4.5}]"} {"utterance_id": "8254-84205-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0027.flac", "task_type": "lexical_detection", "query": "lads", "prompt": "Process the full audio offline. Find every instance of the word \"lads\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 0.94, "lads"]], "response_mode": "offline", "answer": "[{\"word\":\"lads\",\"start\":0.56,\"end\":0.94}]"} {"utterance_id": "4323-55228-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.65, 3.54, "mister arkwright"]], "response_mode": "offline", "answer": "[{\"word\":\"mister arkwright\",\"start\":2.65,\"end\":3.54}]"} {"utterance_id": "7641-96684-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0012.flac", "task_type": "lexical_detection", "query": "stones", "prompt": "Process the full audio offline. Find every instance of the word \"stones\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.18, "stones"], [5.83, 6.42, "stones"]], "response_mode": "offline", "answer": "[{\"word\":\"stones\",\"start\":2.69,\"end\":3.18},{\"word\":\"stones\",\"start\":5.83,\"end\":6.42}]"} {"utterance_id": "8173-294714-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0052.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.28, 2.84, "lawyers"]], "response_mode": "offline", "answer": "[{\"word\":\"lawyers\",\"start\":2.28,\"end\":2.84}]"} {"utterance_id": "6455-67804-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0038.flac", "task_type": "lexical_detection", "query": "intended", "prompt": "After listening to the whole recording, output every instance of the word \"intended\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.23, "intended"]], "response_mode": "offline", "answer": "[{\"word\":\"intended\",\"start\":1.68,\"end\":2.23}]"} {"utterance_id": "4515-11057-0110", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0110.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.47, 2.98, "final"]], "response_mode": "offline", "answer": "[{\"word\":\"final\",\"start\":2.47,\"end\":2.98}]"} {"utterance_id": "6267-65525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0021.flac", "task_type": "lexical_counting", "query": "life", "prompt": "Process the full audio offline. For every occurrence of the word \"life\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.53, "one"], [11.22, 11.58, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.15,\"end\":3.53},{\"word\":\"two\",\"start\":11.22,\"end\":11.58}]"} {"utterance_id": "8173-294714-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.67, 4.6, "mister van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"mister van brandt\",\"start\":3.67,\"end\":4.6}]"} {"utterance_id": "5543-27761-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0022.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.65, 2.89, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":2.65,\"end\":2.89}]"} {"utterance_id": "6467-94831-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0009.flac", "task_type": "lexical_detection", "query": "back", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"back\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.43, "back"]], "response_mode": "offline", "answer": "[{\"word\":\"back\",\"start\":1.09,\"end\":1.43}]"} {"utterance_id": "1585-157660-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.32, 6.68, "william"]], "response_mode": "offline", "answer": "[{\"word\":\"william\",\"start\":6.32,\"end\":6.68}]"} {"utterance_id": "4153-186222-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0008.flac", "task_type": "lexical_detection", "query": "arrangement", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"arrangement\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.49, 3.1, "arrangement"]], "response_mode": "offline", "answer": "[{\"word\":\"arrangement\",\"start\":2.49,\"end\":3.1}]"} {"utterance_id": "6467-94831-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0011.flac", "task_type": "lexical_detection", "query": "fell", "prompt": "Process the full audio offline. Find every instance of the word \"fell\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.07, "fell"]], "response_mode": "offline", "answer": "[{\"word\":\"fell\",\"start\":2.81,\"end\":3.07}]"} {"utterance_id": "7601-291468-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0002.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.95, 5.53, "cities"]], "response_mode": "offline", "answer": "[{\"word\":\"cities\",\"start\":4.95,\"end\":5.53}]"} {"utterance_id": "7601-101622-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0002.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.6, 1.87, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.6,\"end\":1.87}]"} {"utterance_id": "2506-11278-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0024.flac", "task_type": "lexical_detection", "query": "ages", "prompt": "After the audio ends, extract all occurrences of the word \"ages\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.38, "ages"]], "response_mode": "offline", "answer": "[{\"word\":\"ages\",\"start\":3.82,\"end\":4.38}]"} {"utterance_id": "6599-38591-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.4, 0.87, "francis"]], "response_mode": "offline", "answer": "[{\"word\":\"francis\",\"start\":0.4,\"end\":0.87}]"} {"utterance_id": "1701-141760-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0014.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Once the audio has finished, identify every instance of an organization name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.31, 3.81, "regiment"]], "response_mode": "offline", "answer": "[{\"word\":\"regiment\",\"start\":3.31,\"end\":3.81}]"} {"utterance_id": "116-288045-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0027.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After the audio ends, extract all occurrences of a time of day. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.77, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.37,\"end\":1.77}]"} {"utterance_id": "6267-65525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0000.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 2.88, "afternoon"]], "response_mode": "offline", "answer": "[{\"word\":\"afternoon\",\"start\":2.3,\"end\":2.88}]"} {"utterance_id": "1255-138279-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.98, 4.41, "frank"]], "response_mode": "offline", "answer": "[{\"word\":\"frank\",\"start\":3.98,\"end\":4.41}]"} {"utterance_id": "3663-172005-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.57, 1.0, "hounds"], [3.91, 4.29, "fawn"]], "response_mode": "offline", "answer": "[{\"word\":\"hounds\",\"start\":0.57,\"end\":1.0},{\"word\":\"fawn\",\"start\":3.91,\"end\":4.29}]"} {"utterance_id": "1701-141759-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.24, 6.6, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":6.24,\"end\":6.6}]"} {"utterance_id": "1686-142278-0089", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0089.flac", "task_type": "lexical_detection", "query": "decidedly", "prompt": "After the audio ends, extract all occurrences of the word \"decidedly\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.65, 3.38, "decidedly"]], "response_mode": "offline", "answer": "[{\"word\":\"decidedly\",\"start\":2.65,\"end\":3.38}]"} {"utterance_id": "6841-88294-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0068.flac", "task_type": "lexical_counting", "query": "third", "prompt": "Once the audio has finished, list each occurrence of the word \"third\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.36, 0.57, "one"], [2.32, 2.58, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.36,\"end\":0.57},{\"word\":\"two\",\"start\":2.32,\"end\":2.58}]"} {"utterance_id": "6455-67804-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0002.flac", "task_type": "lexical_detection", "query": "binnie", "prompt": "After listening to the whole recording, output every instance of the word \"binnie\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.19, 3.5, "binnie"]], "response_mode": "offline", "answer": "[{\"word\":\"binnie\",\"start\":3.19,\"end\":3.5}]"} {"utterance_id": "3663-172528-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.96, "giorgio"]], "response_mode": "offline", "answer": "[{\"word\":\"giorgio\",\"start\":0.42,\"end\":0.96}]"} {"utterance_id": "5849-50873-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0029.flac", "task_type": "lexical_detection", "query": "train", "prompt": "Process the full audio offline. Find every instance of the word \"train\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.58, 5.97, "train"]], "response_mode": "offline", "answer": "[{\"word\":\"train\",\"start\":5.58,\"end\":5.97}]"} {"utterance_id": "1585-157660-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0016.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 1.24, "one"], [3.36, 3.74, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.42,\"end\":1.24},{\"word\":\"two\",\"start\":3.36,\"end\":3.74}]"} {"utterance_id": "8254-84205-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0016.flac", "task_type": "lexical_detection", "query": "seeing", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"seeing\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.16, 1.49, "seeing"]], "response_mode": "offline", "answer": "[{\"word\":\"seeing\",\"start\":1.16,\"end\":1.49}]"} {"utterance_id": "4153-186222-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0023.flac", "task_type": "lexical_detection", "query": "lightly", "prompt": "Process the full audio offline. Find every instance of the word \"lightly\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.51, 2.97, "lightly"]], "response_mode": "offline", "answer": "[{\"word\":\"lightly\",\"start\":2.51,\"end\":2.97}]"} {"utterance_id": "4515-11057-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0063.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.78, 4.2, "father"], [6.43, 6.94, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":3.78,\"end\":4.2},{\"word\":\"wife\",\"start\":6.43,\"end\":6.94}]"} {"utterance_id": "4831-29134-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0000.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Listen carefully to the entire clip, then report every instance of a date as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.96, 5.31, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":4.96,\"end\":5.31}]"} {"utterance_id": "7697-245715-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0003.flac", "task_type": "lexical_detection", "query": "established", "prompt": "Once the audio has finished, identify every instance of the word \"established\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[10.72, 11.25, "established"]], "response_mode": "offline", "answer": "[{\"word\":\"established\",\"start\":10.72,\"end\":11.25}]"} {"utterance_id": "1701-141759-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0023.flac", "task_type": "lexical_detection", "query": "stupid", "prompt": "After the audio ends, extract all occurrences of the word \"stupid\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.94, "stupid"]], "response_mode": "offline", "answer": "[{\"word\":\"stupid\",\"start\":1.33,\"end\":1.94}]"} {"utterance_id": "4153-186222-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0020.flac", "task_type": "lexical_detection", "query": "persisted", "prompt": "When the utterance is complete, report all instances of the word \"persisted\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.67, 3.27, "persisted"]], "response_mode": "offline", "answer": "[{\"word\":\"persisted\",\"start\":2.67,\"end\":3.27}]"} {"utterance_id": "8288-274162-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.44, 3.38, "d'artagnan"]], "response_mode": "offline", "answer": "[{\"word\":\"d'artagnan\",\"start\":2.44,\"end\":3.38}]"} {"utterance_id": "1585-157660-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0012.flac", "task_type": "lexical_detection", "query": "cast", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"cast\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.7, "cast"]], "response_mode": "offline", "answer": "[{\"word\":\"cast\",\"start\":2.01,\"end\":2.7}]"} {"utterance_id": "7697-105815-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0029.flac", "task_type": "lexical_detection", "query": "grief", "prompt": "When the utterance is complete, report all instances of the word \"grief\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.63, 8.96, "grief"]], "response_mode": "offline", "answer": "[{\"word\":\"grief\",\"start\":8.63,\"end\":8.96}]"} {"utterance_id": "6267-65525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0030.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[0.38, 0.73, "aunt"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt\",\"start\":0.38,\"end\":0.73}]"} {"utterance_id": "1701-141760-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0052.flac", "task_type": "lexical_detection", "query": "coldly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"coldly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.76, 3.18, "coldly"]], "response_mode": "offline", "answer": "[{\"word\":\"coldly\",\"start\":2.76,\"end\":3.18}]"} {"utterance_id": "6599-38591-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0009.flac", "task_type": "lexical_detection", "query": "obliged", "prompt": "Once the audio has finished, identify every instance of the word \"obliged\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.27, 3.73, "obliged"]], "response_mode": "offline", "answer": "[{\"word\":\"obliged\",\"start\":3.27,\"end\":3.73}]"} {"utterance_id": "700-122868-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0019.flac", "task_type": "lexical_detection", "query": "lower", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"lower\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.84, 4.1, "lower"]], "response_mode": "offline", "answer": "[{\"word\":\"lower\",\"start\":3.84,\"end\":4.1}]"} {"utterance_id": "6267-53049-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.63, 1.15, "dorrie"], [11.32, 12.66, "grandmother hunter"]], "response_mode": "offline", "answer": "[{\"word\":\"dorrie\",\"start\":0.63,\"end\":1.15},{\"word\":\"grandmother hunter\",\"start\":11.32,\"end\":12.66}]"} {"utterance_id": "8288-274162-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0063.flac", "task_type": "lexical_detection", "query": "worse", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"worse\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.09, "worse"]], "response_mode": "offline", "answer": "[{\"word\":\"worse\",\"start\":2.7,\"end\":3.09}]"} {"utterance_id": "4831-29134-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0000.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a profession: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.54, 0.92, "one"], [2.38, 2.81, "two"], [4.05, 4.77, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.54,\"end\":0.92},{\"word\":\"two\",\"start\":2.38,\"end\":2.81},{\"word\":\"three\",\"start\":4.05,\"end\":4.77}]"} {"utterance_id": "6467-94831-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.11, 4.48, "tongue"]], "response_mode": "offline", "answer": "[{\"word\":\"tongue\",\"start\":4.11,\"end\":4.48}]"} {"utterance_id": "6467-94831-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0000.flac", "task_type": "lexical_detection", "query": "snuggers", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"snuggers\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.48, "snuggers"]], "response_mode": "offline", "answer": "[{\"word\":\"snuggers\",\"start\":0.82,\"end\":1.48}]"} {"utterance_id": "1255-90407-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0010.flac", "task_type": "lexical_detection", "query": "heard", "prompt": "After the audio ends, extract all occurrences of the word \"heard\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 0.98, "heard"]], "response_mode": "offline", "answer": "[{\"word\":\"heard\",\"start\":0.77,\"end\":0.98}]"} {"utterance_id": "6841-88291-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0045.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.86, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":1.46,\"end\":1.86}]"} {"utterance_id": "1686-142278-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0060.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.99, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.51,\"end\":0.99}]"} {"utterance_id": "1686-142278-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0014.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.67, 5.06, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":4.67,\"end\":5.06}]"} {"utterance_id": "4831-18525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.38, "polly"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":1.05,\"end\":1.38}]"} {"utterance_id": "700-122866-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0016.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.51, "rabbit"]], "response_mode": "offline", "answer": "[{\"word\":\"rabbit\",\"start\":2.13,\"end\":2.51}]"} {"utterance_id": "4570-14911-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0016.flac", "task_type": "lexical_detection", "query": "variety", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"variety\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.37, "variety"]], "response_mode": "offline", "answer": "[{\"word\":\"variety\",\"start\":1.74,\"end\":2.37}]"} {"utterance_id": "4323-13259-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0002.flac", "task_type": "lexical_detection", "query": "saloon", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"saloon\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 2.96, "saloon"]], "response_mode": "offline", "answer": "[{\"word\":\"saloon\",\"start\":2.52,\"end\":2.96}]"} {"utterance_id": "4153-186223-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.94, "nora"]], "response_mode": "offline", "answer": "[{\"word\":\"nora\",\"start\":0.53,\"end\":0.94}]"} {"utterance_id": "7697-245712-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0012.flac", "task_type": "lexical_counting", "query": "man", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"man\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.33, 2.53, "one"], [5.75, 5.98, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.33,\"end\":2.53},{\"word\":\"two\",\"start\":5.75,\"end\":5.98}]"} {"utterance_id": "4323-18416-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0019.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "After the audio ends, tally every instance of a family relation. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.5, 1.14, "one"], [3.43, 3.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":1.14},{\"word\":\"two\",\"start\":3.43,\"end\":3.76}]"} {"utterance_id": "6841-88291-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.77, "heels"]], "response_mode": "offline", "answer": "[{\"word\":\"heels\",\"start\":1.41,\"end\":1.77}]"} {"utterance_id": "1701-141759-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0016.flac", "task_type": "lexical_counting", "query": "sonya", "prompt": "When the utterance is complete, report how the word \"sonya\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.34, "one"], [4.29, 4.65, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.9,\"end\":1.34},{\"word\":\"two\",\"start\":4.29,\"end\":4.65}]"} {"utterance_id": "5543-27761-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0001.flac", "task_type": "lexical_detection", "query": "child", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"child\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.98, "child"]], "response_mode": "offline", "answer": "[{\"word\":\"child\",\"start\":1.3,\"end\":1.98}]"} {"utterance_id": "4515-11057-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0020.flac", "task_type": "lexical_detection", "query": "women", "prompt": "After the audio ends, extract all occurrences of the word \"women\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.26, 3.54, "women"]], "response_mode": "offline", "answer": "[{\"word\":\"women\",\"start\":3.26,\"end\":3.54}]"} {"utterance_id": "4831-25894-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0016.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen carefully to the entire clip, then report every instance of a disease name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 2.0, "chilblains"]], "response_mode": "offline", "answer": "[{\"word\":\"chilblains\",\"start\":1.45,\"end\":2.0}]"} {"utterance_id": "6123-59186-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0008.flac", "task_type": "lexical_detection", "query": "catering", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"catering\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.21, 2.79, "catering"]], "response_mode": "offline", "answer": "[{\"word\":\"catering\",\"start\":2.21,\"end\":2.79}]"} {"utterance_id": "6267-65525-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0017.flac", "task_type": "lexical_detection", "query": "east", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"east\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.37, "east"]], "response_mode": "offline", "answer": "[{\"word\":\"east\",\"start\":1.95,\"end\":2.37}]"} {"utterance_id": "7697-245712-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0013.flac", "task_type": "lexical_detection", "query": "master", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"master\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.64, "master"]], "response_mode": "offline", "answer": "[{\"word\":\"master\",\"start\":2.14,\"end\":2.64}]"} {"utterance_id": "3663-172528-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0029.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.1, 4.68, "tailor"]], "response_mode": "offline", "answer": "[{\"word\":\"tailor\",\"start\":4.1,\"end\":4.68}]"} {"utterance_id": "700-122867-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0041.flac", "task_type": "lexical_detection", "query": "afternoon", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"afternoon\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.89, "afternoon"]], "response_mode": "offline", "answer": "[{\"word\":\"afternoon\",\"start\":1.41,\"end\":1.89}]"} {"utterance_id": "6599-38591-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.58, "charles"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":1.02,\"end\":1.58}]"} {"utterance_id": "1585-131718-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0032.flac", "task_type": "lexical_detection", "query": "homes", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"homes\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.94, 2.39, "homes"]], "response_mode": "offline", "answer": "[{\"word\":\"homes\",\"start\":1.94,\"end\":2.39}]"} {"utterance_id": "7641-96684-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0030.flac", "task_type": "lexical_detection", "query": "strive", "prompt": "Once the audio has finished, identify every instance of the word \"strive\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.05, 1.47, "strive"]], "response_mode": "offline", "answer": "[{\"word\":\"strive\",\"start\":1.05,\"end\":1.47}]"} {"utterance_id": "4515-11057-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0052.flac", "task_type": "lexical_detection", "query": "affairs", "prompt": "When the utterance is complete, report all instances of the word \"affairs\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.37, 2.93, "affairs"]], "response_mode": "offline", "answer": "[{\"word\":\"affairs\",\"start\":2.37,\"end\":2.93}]"} {"utterance_id": "8254-84205-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.16, 6.61, "mustang"]], "response_mode": "offline", "answer": "[{\"word\":\"mustang\",\"start\":6.16,\"end\":6.61}]"} {"utterance_id": "4831-25894-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0032.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a food name with fields word, start, end (seconds). If none, reply [].", "labels": [[6.47, 7.03, "gingerbread"]], "response_mode": "offline", "answer": "[{\"word\":\"gingerbread\",\"start\":6.47,\"end\":7.03}]"} {"utterance_id": "1630-96099-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0006.flac", "task_type": "lexical_detection", "query": "great", "prompt": "Process the full audio offline. Find every instance of the word \"great\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.56, "great"]], "response_mode": "offline", "answer": "[{\"word\":\"great\",\"start\":1.18,\"end\":1.56}]"} {"utterance_id": "5849-50873-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0033.flac", "task_type": "lexical_detection", "query": "cried", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"cried\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.35, 1.67, "cried"]], "response_mode": "offline", "answer": "[{\"word\":\"cried\",\"start\":1.35,\"end\":1.67}]"} {"utterance_id": "2506-11278-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.78, "nose"]], "response_mode": "offline", "answer": "[{\"word\":\"nose\",\"start\":1.46,\"end\":1.78}]"} {"utterance_id": "3660-172183-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.1, 3.49, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.1,\"end\":3.49}]"} {"utterance_id": "1650-167613-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[2.31, 3.02, "commander"]], "response_mode": "offline", "answer": "[{\"word\":\"commander\",\"start\":2.31,\"end\":3.02}]"} {"utterance_id": "4570-56594-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0009.flac", "task_type": "semantic_detection", "query": "color", "prompt": "You will hear a complete utterance. When it ends, list every instance of a color name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.44, 9.2, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":8.44,\"end\":9.2}]"} {"utterance_id": "4153-186222-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0030.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.34, "nieces"]], "response_mode": "offline", "answer": "[{\"word\":\"nieces\",\"start\":3.94,\"end\":4.34}]"} {"utterance_id": "6123-59150-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0037.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.06, 4.46, "apostle"]], "response_mode": "offline", "answer": "[{\"word\":\"apostle\",\"start\":4.06,\"end\":4.46}]"} {"utterance_id": "6841-88291-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0001.flac", "task_type": "semantic_counting", "query": "color", "prompt": "When the utterance is complete, report how a color name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.67, 5.96, "one"], [7.39, 7.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.67,\"end\":5.96},{\"word\":\"two\",\"start\":7.39,\"end\":7.78}]"} {"utterance_id": "6455-67804-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0033.flac", "task_type": "lexical_detection", "query": "ejaculated", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"ejaculated\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.25, "ejaculated"]], "response_mode": "offline", "answer": "[{\"word\":\"ejaculated\",\"start\":0.6,\"end\":1.25}]"} {"utterance_id": "8254-115543-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0045.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.96, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":0.6,\"end\":0.96}]"} {"utterance_id": "6841-88294-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0013.flac", "task_type": "lexical_detection", "query": "gave", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"gave\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.22, 2.48, "gave"]], "response_mode": "offline", "answer": "[{\"word\":\"gave\",\"start\":2.22,\"end\":2.48}]"} {"utterance_id": "4570-102353-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.9, 1.37, "general"]], "response_mode": "offline", "answer": "[{\"word\":\"general\",\"start\":0.9,\"end\":1.37}]"} {"utterance_id": "4323-55228-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0039.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.41, 2.7, "five"]], "response_mode": "offline", "answer": "[{\"word\":\"five\",\"start\":2.41,\"end\":2.7}]"} {"utterance_id": "700-122867-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.14, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":0.65,\"end\":1.14}]"} {"utterance_id": "6467-62797-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0004.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[0.32, 0.65, "one"], [3.63, 4.15, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.32,\"end\":0.65},{\"word\":\"two\",\"start\":3.63,\"end\":4.15}]"} {"utterance_id": "1630-73710-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.03, 4.49, "mind"], [10.48, 10.91, "wing"]], "response_mode": "offline", "answer": "[{\"word\":\"mind\",\"start\":4.03,\"end\":4.49},{\"word\":\"wing\",\"start\":10.48,\"end\":10.91}]"} {"utterance_id": "8254-115543-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0034.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.38, "brahmin's"]], "response_mode": "offline", "answer": "[{\"word\":\"brahmin's\",\"start\":0.93,\"end\":1.38}]"} {"utterance_id": "1701-141760-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.03, 5.61, "grand duke's"]], "response_mode": "offline", "answer": "[{\"word\":\"grand duke's\",\"start\":5.03,\"end\":5.61}]"} {"utterance_id": "8173-294714-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0016.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.87, 2.31, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":1.87,\"end\":2.31}]"} {"utterance_id": "6841-88291-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.35, 2.68, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.35,\"end\":2.68}]"} {"utterance_id": "8288-274162-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0051.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.74, 1.12, "painter"]], "response_mode": "offline", "answer": "[{\"word\":\"painter\",\"start\":0.74,\"end\":1.12}]"} {"utterance_id": "6841-88291-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0018.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.89, "calf"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":0.6,\"end\":0.89}]"} {"utterance_id": "3660-172182-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0014.flac", "task_type": "lexical_detection", "query": "wounds", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wounds\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.45, 4.94, "wounds"]], "response_mode": "offline", "answer": "[{\"word\":\"wounds\",\"start\":4.45,\"end\":4.94}]"} {"utterance_id": "1630-102884-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0015.flac", "task_type": "lexical_detection", "query": "incapable", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"incapable\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 2.13, "incapable"]], "response_mode": "offline", "answer": "[{\"word\":\"incapable\",\"start\":1.25,\"end\":2.13}]"} {"utterance_id": "3660-6517-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0006.flac", "task_type": "lexical_detection", "query": "men's", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"men's\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.38, 0.63, "men's"]], "response_mode": "offline", "answer": "[{\"word\":\"men's\",\"start\":0.38,\"end\":0.63}]"} {"utterance_id": "4572-64670-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.27, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":3.82,\"end\":4.27}]"} {"utterance_id": "5849-50964-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.55, "nan"]], "response_mode": "offline", "answer": "[{\"word\":\"nan\",\"start\":2.26,\"end\":2.55}]"} {"utterance_id": "1585-131718-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0028.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.16, 13.46, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":13.16,\"end\":13.46}]"} {"utterance_id": "6467-97061-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0013.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.34, "apsaras"]], "response_mode": "offline", "answer": "[{\"word\":\"apsaras\",\"start\":0.75,\"end\":1.34}]"} {"utterance_id": "6455-67803-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.63, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":2.26,\"end\":2.63}]"} {"utterance_id": "2506-13150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0003.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "After the audio ends, extract all occurrences of a disease name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.15, 1.99, "sick sick"]], "response_mode": "offline", "answer": "[{\"word\":\"sick sick\",\"start\":0.15,\"end\":1.99}]"} {"utterance_id": "1701-141759-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0024.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.95, 4.3, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":3.95,\"end\":4.3}]"} {"utterance_id": "4323-13259-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.3, 0.64, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":0.3,\"end\":0.64}]"} {"utterance_id": "5849-50963-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.17, "aunt emily"], [4.35, 4.7, "nan"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt emily\",\"start\":0.6,\"end\":1.17},{\"word\":\"nan\",\"start\":4.35,\"end\":4.7}]"} {"utterance_id": "4831-25894-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0032.flac", "task_type": "lexical_detection", "query": "gingerbread", "prompt": "After listening to the whole recording, output every instance of the word \"gingerbread\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.47, 7.03, "gingerbread"]], "response_mode": "offline", "answer": "[{\"word\":\"gingerbread\",\"start\":6.47,\"end\":7.03}]"} {"utterance_id": "1585-157660-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.32, 0.85, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":0.32,\"end\":0.85}]"} {"utterance_id": "4323-55228-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.08, 0.46, "pete"], [5.18, 5.49, "pete's"]], "response_mode": "offline", "answer": "[{\"word\":\"pete\",\"start\":0.08,\"end\":0.46},{\"word\":\"pete's\",\"start\":5.18,\"end\":5.49}]"} {"utterance_id": "3660-6517-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.31, 1.74, "anderson's"]], "response_mode": "offline", "answer": "[{\"word\":\"anderson's\",\"start\":1.31,\"end\":1.74}]"} {"utterance_id": "1585-157660-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0015.flac", "task_type": "lexical_detection", "query": "invited", "prompt": "After the audio ends, extract all occurrences of the word \"invited\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.11, 6.62, "invited"]], "response_mode": "offline", "answer": "[{\"word\":\"invited\",\"start\":6.11,\"end\":6.62}]"} {"utterance_id": "2506-11278-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.71, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":2.39,\"end\":2.71}]"} {"utterance_id": "4572-64670-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0002.flac", "task_type": "lexical_detection", "query": "lead", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"lead\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.96, 1.2, "lead"]], "response_mode": "offline", "answer": "[{\"word\":\"lead\",\"start\":0.96,\"end\":1.2}]"} {"utterance_id": "6123-59186-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0004.flac", "task_type": "lexical_detection", "query": "writes", "prompt": "Process the full audio offline. Find every instance of the word \"writes\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.35, "writes"]], "response_mode": "offline", "answer": "[{\"word\":\"writes\",\"start\":0.96,\"end\":1.35}]"} {"utterance_id": "4323-55228-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0019.flac", "task_type": "lexical_detection", "query": "painting", "prompt": "Once the audio has finished, identify every instance of the word \"painting\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.02, 1.42, "painting"]], "response_mode": "offline", "answer": "[{\"word\":\"painting\",\"start\":1.02,\"end\":1.42}]"} {"utterance_id": "5849-50963-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0012.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.11, "clerks"]], "response_mode": "offline", "answer": "[{\"word\":\"clerks\",\"start\":0.73,\"end\":1.11}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[1.91, 2.24, "one"], [7.56, 7.95, "two"], [11.81, 12.28, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.91,\"end\":2.24},{\"word\":\"two\",\"start\":7.56,\"end\":7.95},{\"word\":\"three\",\"start\":11.81,\"end\":12.28}]"} {"utterance_id": "116-288048-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0022.flac", "task_type": "lexical_detection", "query": "maintained", "prompt": "After listening to the whole recording, output every instance of the word \"maintained\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.16, 3.72, "maintained"]], "response_mode": "offline", "answer": "[{\"word\":\"maintained\",\"start\":3.16,\"end\":3.72}]"} {"utterance_id": "4572-64670-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.71, "heart"], [4.17, 4.51, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":1.37,\"end\":1.71},{\"word\":\"lips\",\"start\":4.17,\"end\":4.51}]"} {"utterance_id": "6841-88291-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0052.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Once the audio has finished, list each occurrence of a body part as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.47, 2.02, "one"], [3.2, 3.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.47,\"end\":2.02},{\"word\":\"two\",\"start\":3.2,\"end\":3.42}]"} {"utterance_id": "8254-84205-0072", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0072.flac", "task_type": "lexical_detection", "query": "week", "prompt": "After listening to the whole recording, output every instance of the word \"week\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.69, 5.04, "week"]], "response_mode": "offline", "answer": "[{\"word\":\"week\",\"start\":4.69,\"end\":5.04}]"} {"utterance_id": "5543-27761-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0024.flac", "task_type": "lexical_detection", "query": "absorbed", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"absorbed\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.35, 1.9, "absorbed"]], "response_mode": "offline", "answer": "[{\"word\":\"absorbed\",\"start\":1.35,\"end\":1.9}]"} {"utterance_id": "700-122868-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0035.flac", "task_type": "lexical_detection", "query": "insulted", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"insulted\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.05, 2.47, "insulted"]], "response_mode": "offline", "answer": "[{\"word\":\"insulted\",\"start\":2.05,\"end\":2.47}]"} {"utterance_id": "1255-90413-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0006.flac", "task_type": "lexical_detection", "query": "determined", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"determined\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.17, "determined"]], "response_mode": "offline", "answer": "[{\"word\":\"determined\",\"start\":0.7,\"end\":1.17}]"} {"utterance_id": "2506-11278-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0028.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.89, 3.37, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":2.89,\"end\":3.37}]"} {"utterance_id": "1686-142278-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0007.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Process the full audio offline. Find every instance of a time of day and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.11, 5.58, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":5.11,\"end\":5.58}]"} {"utterance_id": "4572-64670-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.29, 9.78, "side"]], "response_mode": "offline", "answer": "[{\"word\":\"side\",\"start\":9.29,\"end\":9.78}]"} {"utterance_id": "6841-88291-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0035.flac", "task_type": "lexical_detection", "query": "business", "prompt": "After the audio ends, extract all occurrences of the word \"business\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.23, "business"]], "response_mode": "offline", "answer": "[{\"word\":\"business\",\"start\":0.8,\"end\":1.23}]"} {"utterance_id": "6455-67804-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0026.flac", "task_type": "lexical_detection", "query": "follow", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"follow\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.63, 3.98, "follow"]], "response_mode": "offline", "answer": "[{\"word\":\"follow\",\"start\":3.63,\"end\":3.98}]"} {"utterance_id": "1701-141760-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.83, 5.09, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":4.83,\"end\":5.09}]"} {"utterance_id": "1686-142278-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0047.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.29, "mamma"]], "response_mode": "offline", "answer": "[{\"word\":\"mamma\",\"start\":0.85,\"end\":1.29}]"} {"utterance_id": "6267-65525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.55, 3.47, "ellen douglas"]], "response_mode": "offline", "answer": "[{\"word\":\"ellen douglas\",\"start\":2.55,\"end\":3.47}]"} {"utterance_id": "116-288046-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0003.flac", "task_type": "lexical_counting", "query": "believe", "prompt": "Listen to the entire clip, then report the count sequence for the word \"believe\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.21, "one"], [2.69, 2.98, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.94,\"end\":1.21},{\"word\":\"two\",\"start\":2.69,\"end\":2.98}]"} {"utterance_id": "8254-115543-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.05, 2.44, "potter"], [4.61, 5.1, "fakir"]], "response_mode": "offline", "answer": "[{\"word\":\"potter\",\"start\":2.05,\"end\":2.44},{\"word\":\"fakir\",\"start\":4.61,\"end\":5.1}]"} {"utterance_id": "2506-11267-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0002.flac", "task_type": "lexical_detection", "query": "marching", "prompt": "After the audio ends, extract all occurrences of the word \"marching\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.69, 10.24, "marching"]], "response_mode": "offline", "answer": "[{\"word\":\"marching\",\"start\":9.69,\"end\":10.24}]"} {"utterance_id": "5849-50964-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0009.flac", "task_type": "lexical_detection", "query": "willing", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"willing\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.83, 3.13, "willing"]], "response_mode": "offline", "answer": "[{\"word\":\"willing\",\"start\":2.83,\"end\":3.13}]"} {"utterance_id": "4831-18525-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.77, 3.18, "urchin"]], "response_mode": "offline", "answer": "[{\"word\":\"urchin\",\"start\":2.77,\"end\":3.18}]"} {"utterance_id": "7697-105815-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.8, "maid"]], "response_mode": "offline", "answer": "[{\"word\":\"maid\",\"start\":1.45,\"end\":1.8}]"} {"utterance_id": "6123-59150-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0023.flac", "task_type": "lexical_detection", "query": "haven't", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"haven't\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.64, "haven't"]], "response_mode": "offline", "answer": "[{\"word\":\"haven't\",\"start\":0.32,\"end\":0.64}]"} {"utterance_id": "4831-25894-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.27, 0.58, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":0.27,\"end\":0.58}]"} {"utterance_id": "4572-112383-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0001.flac", "task_type": "lexical_detection", "query": "sita", "prompt": "After the audio ends, extract all occurrences of the word \"sita\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 2.13, "sita"]], "response_mode": "offline", "answer": "[{\"word\":\"sita\",\"start\":1.45,\"end\":2.13}]"} {"utterance_id": "1686-142278-0081", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0081.flac", "task_type": "semantic_counting", "query": "city", "prompt": "Once the audio has finished, list each occurrence of a city name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.85, 1.25, "one"], [9.94, 10.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.85,\"end\":1.25},{\"word\":\"two\",\"start\":9.94,\"end\":10.46}]"} {"utterance_id": "700-122866-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0005.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.13, 2.52, "avonlea"]], "response_mode": "offline", "answer": "[{\"word\":\"avonlea\",\"start\":2.13,\"end\":2.52}]"} {"utterance_id": "4515-11057-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0075.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.45, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":2.06,\"end\":2.45}]"} {"utterance_id": "3915-57461-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0009.flac", "task_type": "lexical_detection", "query": "travellers", "prompt": "After the audio ends, extract all occurrences of the word \"travellers\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.27, 3.95, "travellers"]], "response_mode": "offline", "answer": "[{\"word\":\"travellers\",\"start\":3.27,\"end\":3.95}]"} {"utterance_id": "116-288045-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0015.flac", "task_type": "lexical_detection", "query": "whispered", "prompt": "Once the audio has finished, identify every instance of the word \"whispered\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.32, 0.63, "whispered"]], "response_mode": "offline", "answer": "[{\"word\":\"whispered\",\"start\":0.32,\"end\":0.63}]"} {"utterance_id": "1650-167613-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0033.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.65, 5.3, "moscow"]], "response_mode": "offline", "answer": "[{\"word\":\"moscow\",\"start\":4.65,\"end\":5.3}]"} {"utterance_id": "4570-24733-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0008.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[7.35, 7.95, "one"], [10.99, 11.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.35,\"end\":7.95},{\"word\":\"two\",\"start\":10.99,\"end\":11.49}]"} {"utterance_id": "1686-142278-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0025.flac", "task_type": "lexical_detection", "query": "horror", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"horror\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.57, "horror"]], "response_mode": "offline", "answer": "[{\"word\":\"horror\",\"start\":3.14,\"end\":3.57}]"} {"utterance_id": "4831-18525-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.79, 2.15, "polly"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":1.79,\"end\":2.15}]"} {"utterance_id": "2506-11278-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0005.flac", "task_type": "lexical_detection", "query": "teller", "prompt": "When the utterance is complete, report all instances of the word \"teller\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.42, 6.85, "teller"]], "response_mode": "offline", "answer": "[{\"word\":\"teller\",\"start\":6.42,\"end\":6.85}]"} {"utterance_id": "6455-66379-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0011.flac", "task_type": "lexical_detection", "query": "kings", "prompt": "When the utterance is complete, report all instances of the word \"kings\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.82, "kings"]], "response_mode": "offline", "answer": "[{\"word\":\"kings\",\"start\":1.36,\"end\":1.82}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "lexical_detection", "query": "feed", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"feed\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.39, 4.8, "feed"]], "response_mode": "offline", "answer": "[{\"word\":\"feed\",\"start\":4.39,\"end\":4.8}]"} {"utterance_id": "5849-50964-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0000.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.66, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":4.23,\"end\":4.66}]"} {"utterance_id": "7601-291468-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0001.flac", "task_type": "lexical_detection", "query": "historic", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"historic\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.5, 3.22, "historic"]], "response_mode": "offline", "answer": "[{\"word\":\"historic\",\"start\":2.5,\"end\":3.22}]"} {"utterance_id": "8254-84205-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0033.flac", "task_type": "lexical_detection", "query": "standing", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"standing\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.68, "standing"]], "response_mode": "offline", "answer": "[{\"word\":\"standing\",\"start\":1.34,\"end\":1.68}]"} {"utterance_id": "4831-18525-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0024.flac", "task_type": "lexical_detection", "query": "try", "prompt": "After listening to the whole recording, output every instance of the word \"try\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.15, 4.36, "try"]], "response_mode": "offline", "answer": "[{\"word\":\"try\",\"start\":4.15,\"end\":4.36}]"} {"utterance_id": "1585-131718-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0014.flac", "task_type": "lexical_detection", "query": "bitter", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"bitter\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.71, "bitter"]], "response_mode": "offline", "answer": "[{\"word\":\"bitter\",\"start\":1.46,\"end\":1.71}]"} {"utterance_id": "4572-64670-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0022.flac", "task_type": "lexical_detection", "query": "ate", "prompt": "When the utterance is complete, report all instances of the word \"ate\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 3.96, "ate"]], "response_mode": "offline", "answer": "[{\"word\":\"ate\",\"start\":3.54,\"end\":3.96}]"} {"utterance_id": "4831-25894-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0020.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.73, 3.12, "one"], [5.75, 6.16, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.73,\"end\":3.12},{\"word\":\"two\",\"start\":5.75,\"end\":6.16}]"} {"utterance_id": "4570-102353-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0005.flac", "task_type": "lexical_detection", "query": "instructions", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"instructions\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.5, "instructions"]], "response_mode": "offline", "answer": "[{\"word\":\"instructions\",\"start\":0.86,\"end\":1.5}]"} {"utterance_id": "1585-131718-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0006.flac", "task_type": "lexical_detection", "query": "conditions", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"conditions\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.61, 2.57, "conditions"]], "response_mode": "offline", "answer": "[{\"word\":\"conditions\",\"start\":1.61,\"end\":2.57}]"} {"utterance_id": "4515-11057-0110", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0110.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "You will hear a complete utterance. When it ends, list every instance of a time of day you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.03, 1.37, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.03,\"end\":1.37}]"} {"utterance_id": "4570-102353-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0004.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "You will hear a complete utterance. When it ends, list every instance of an organization name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.7, 2.31, "regiments"], [7.26, 7.6, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"regiments\",\"start\":1.7,\"end\":2.31},{\"word\":\"army\",\"start\":7.26,\"end\":7.6}]"} {"utterance_id": "4515-11057-0083", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0083.flac", "task_type": "lexical_detection", "query": "faithful", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"faithful\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.08, 1.68, "faithful"]], "response_mode": "offline", "answer": "[{\"word\":\"faithful\",\"start\":1.08,\"end\":1.68}]"} {"utterance_id": "4153-186222-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0035.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.06, 4.56, "elmhurst"]], "response_mode": "offline", "answer": "[{\"word\":\"elmhurst\",\"start\":4.06,\"end\":4.56}]"} {"utterance_id": "700-122868-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.25, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":1.07,\"end\":1.25}]"} {"utterance_id": "4570-56594-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0005.flac", "task_type": "lexical_counting", "query": "lady", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"lady\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.08, 1.66, "one"], [4.61, 4.99, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.08,\"end\":1.66},{\"word\":\"two\",\"start\":4.61,\"end\":4.99}]"} {"utterance_id": "6467-62797-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0006.flac", "task_type": "lexical_detection", "query": "eating", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"eating\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.19, 5.48, "eating"]], "response_mode": "offline", "answer": "[{\"word\":\"eating\",\"start\":5.19,\"end\":5.48}]"} {"utterance_id": "1686-142278-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0042.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.27, 1.71, "bishop"]], "response_mode": "offline", "answer": "[{\"word\":\"bishop\",\"start\":1.27,\"end\":1.71}]"} {"utterance_id": "1585-131718-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0015.flac", "task_type": "lexical_detection", "query": "largest", "prompt": "After the audio ends, extract all occurrences of the word \"largest\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.1, 7.48, "largest"]], "response_mode": "offline", "answer": "[{\"word\":\"largest\",\"start\":7.1,\"end\":7.48}]"} {"utterance_id": "1650-167613-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0053.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.07, 2.45, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":2.07,\"end\":2.45}]"} {"utterance_id": "4831-18525-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0012.flac", "task_type": "lexical_detection", "query": "ensued", "prompt": "After listening to the whole recording, output every instance of the word \"ensued\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.93, 4.36, "ensued"]], "response_mode": "offline", "answer": "[{\"word\":\"ensued\",\"start\":3.93,\"end\":4.36}]"} {"utterance_id": "7697-105815-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0014.flac", "task_type": "lexical_detection", "query": "praise", "prompt": "After listening to the whole recording, output every instance of the word \"praise\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.61, 8.35, "praise"]], "response_mode": "offline", "answer": "[{\"word\":\"praise\",\"start\":7.61,\"end\":8.35}]"} {"utterance_id": "1585-131718-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0031.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.88, 7.76, "sixty five"]], "response_mode": "offline", "answer": "[{\"word\":\"sixty five\",\"start\":6.88,\"end\":7.76}]"} {"utterance_id": "4570-14911-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.24, "puppies"]], "response_mode": "offline", "answer": "[{\"word\":\"puppies\",\"start\":0.8,\"end\":1.24}]"} {"utterance_id": "6841-88294-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0046.flac", "task_type": "lexical_detection", "query": "better", "prompt": "After listening to the whole recording, output every instance of the word \"better\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.42, "better"]], "response_mode": "offline", "answer": "[{\"word\":\"better\",\"start\":2.16,\"end\":2.42}]"} {"utterance_id": "8173-294714-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0049.flac", "task_type": "lexical_detection", "query": "plan", "prompt": "Process the full audio offline. Find every instance of the word \"plan\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.6, 1.97, "plan"]], "response_mode": "offline", "answer": "[{\"word\":\"plan\",\"start\":1.6,\"end\":1.97}]"} {"utterance_id": "4515-11057-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0056.flac", "task_type": "lexical_detection", "query": "health", "prompt": "After listening to the whole recording, output every instance of the word \"health\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.25, "health"]], "response_mode": "offline", "answer": "[{\"word\":\"health\",\"start\":2.77,\"end\":3.25}]"} {"utterance_id": "6599-38590-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0003.flac", "task_type": "lexical_detection", "query": "patent", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"patent\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.84, 5.26, "patent"]], "response_mode": "offline", "answer": "[{\"word\":\"patent\",\"start\":4.84,\"end\":5.26}]"} {"utterance_id": "4323-55228-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0022.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.18, 3.4, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"voice\",\"start\":3.18,\"end\":3.4}]"} {"utterance_id": "6123-59186-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0004.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[3.64, 3.88, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.64,\"end\":3.88}]"} {"utterance_id": "4515-11057-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0027.flac", "task_type": "lexical_detection", "query": "silence", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"silence\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.83, 1.63, "silence"]], "response_mode": "offline", "answer": "[{\"word\":\"silence\",\"start\":0.83,\"end\":1.63}]"} {"utterance_id": "4153-186223-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0012.flac", "task_type": "lexical_detection", "query": "used", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"used\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.77, 1.08, "used"]], "response_mode": "offline", "answer": "[{\"word\":\"used\",\"start\":0.77,\"end\":1.08}]"} {"utterance_id": "4323-18416-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0002.flac", "task_type": "lexical_detection", "query": "snapped", "prompt": "After the audio ends, extract all occurrences of the word \"snapped\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.24, "snapped"]], "response_mode": "offline", "answer": "[{\"word\":\"snapped\",\"start\":2.75,\"end\":3.24}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.35, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":0.95,\"end\":1.35}]"} {"utterance_id": "6467-94831-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0023.flac", "task_type": "lexical_detection", "query": "rover", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"rover\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.56, "rover"]], "response_mode": "offline", "answer": "[{\"word\":\"rover\",\"start\":2.16,\"end\":2.56}]"} {"utterance_id": "6467-94831-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0035.flac", "task_type": "lexical_detection", "query": "fresh", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fresh\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.01, 1.37, "fresh"]], "response_mode": "offline", "answer": "[{\"word\":\"fresh\",\"start\":1.01,\"end\":1.37}]"} {"utterance_id": "8254-84205-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.76, 3.14, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":2.76,\"end\":3.14}]"} {"utterance_id": "1650-167613-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0014.flac", "task_type": "lexical_detection", "query": "russia", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"russia\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.63, 0.99, "russia"]], "response_mode": "offline", "answer": "[{\"word\":\"russia\",\"start\":0.63,\"end\":0.99}]"} {"utterance_id": "1686-142278-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0065.flac", "task_type": "lexical_detection", "query": "frederick", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"frederick\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.48, "frederick"]], "response_mode": "offline", "answer": "[{\"word\":\"frederick\",\"start\":1.95,\"end\":2.48}]"} {"utterance_id": "3660-172183-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0025.flac", "task_type": "lexical_detection", "query": "lord", "prompt": "After listening to the whole recording, output every instance of the word \"lord\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.4, "lord"]], "response_mode": "offline", "answer": "[{\"word\":\"lord\",\"start\":1.04,\"end\":1.4}]"} {"utterance_id": "5849-50873-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0011.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a profession (word, start, end). If none, reply [].", "labels": [[3.21, 3.74, "one"], [6.73, 7.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.21,\"end\":3.74},{\"word\":\"two\",\"start\":6.73,\"end\":7.28}]"} {"utterance_id": "4572-112375-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.02, 5.48, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":5.02,\"end\":5.48}]"} {"utterance_id": "1255-90407-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0013.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.94, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":1.56,\"end\":1.94}]"} {"utterance_id": "1650-167613-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0040.flac", "task_type": "lexical_detection", "query": "killed", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"killed\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.03, 3.43, "killed"]], "response_mode": "offline", "answer": "[{\"word\":\"killed\",\"start\":3.03,\"end\":3.43}]"} {"utterance_id": "4831-29134-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0003.flac", "task_type": "lexical_counting", "query": "sunday", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"sunday\" with word/start/end. If none, reply [].", "labels": [[0.91, 1.25, "one"], [6.79, 7.15, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.91,\"end\":1.25},{\"word\":\"two\",\"start\":6.79,\"end\":7.15}]"} {"utterance_id": "3915-98647-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0007.flac", "task_type": "lexical_detection", "query": "uncertain", "prompt": "When the utterance is complete, report all instances of the word \"uncertain\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.26, 4.95, "uncertain"]], "response_mode": "offline", "answer": "[{\"word\":\"uncertain\",\"start\":4.26,\"end\":4.95}]"} {"utterance_id": "3660-172182-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0003.flac", "task_type": "lexical_detection", "query": "beloved", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"beloved\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.44, 2.82, "beloved"]], "response_mode": "offline", "answer": "[{\"word\":\"beloved\",\"start\":2.44,\"end\":2.82}]"} {"utterance_id": "8173-294714-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0037.flac", "task_type": "lexical_detection", "query": "need", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"need\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.89, "need"]], "response_mode": "offline", "answer": "[{\"word\":\"need\",\"start\":0.63,\"end\":0.89}]"} {"utterance_id": "7641-96684-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.38, 1.73, "admetus admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus admetus\",\"start\":0.38,\"end\":1.73}]"} {"utterance_id": "1686-142278-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0002.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[10.42, 10.79, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":10.42,\"end\":10.79}]"} {"utterance_id": "1585-131718-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0015.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Do not answer until the audio ends. Then list each instance of a money or currency term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.96, 8.38, "income"]], "response_mode": "offline", "answer": "[{\"word\":\"income\",\"start\":7.96,\"end\":8.38}]"} {"utterance_id": "6841-88291-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0041.flac", "task_type": "lexical_detection", "query": "fro", "prompt": "When the utterance is complete, report all instances of the word \"fro\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.56, "fro"]], "response_mode": "offline", "answer": "[{\"word\":\"fro\",\"start\":1.18,\"end\":1.56}]"} {"utterance_id": "4515-11057-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0017.flac", "task_type": "lexical_detection", "query": "hopeless", "prompt": "Process the full audio offline. Find every instance of the word \"hopeless\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.32, 3.8, "hopeless"]], "response_mode": "offline", "answer": "[{\"word\":\"hopeless\",\"start\":3.32,\"end\":3.8}]"} {"utterance_id": "3660-172182-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0005.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.38, "one"], [5.52, 5.93, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.94,\"end\":4.38},{\"word\":\"two\",\"start\":5.52,\"end\":5.93}]"} {"utterance_id": "1686-142278-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0030.flac", "task_type": "lexical_detection", "query": "swallowed", "prompt": "After the audio ends, extract all occurrences of the word \"swallowed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.91, "swallowed"]], "response_mode": "offline", "answer": "[{\"word\":\"swallowed\",\"start\":0.46,\"end\":0.91}]"} {"utterance_id": "1686-142278-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0036.flac", "task_type": "lexical_detection", "query": "shook", "prompt": "Process the full audio offline. Find every instance of the word \"shook\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.86, "shook"]], "response_mode": "offline", "answer": "[{\"word\":\"shook\",\"start\":0.64,\"end\":0.86}]"} {"utterance_id": "4323-18416-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0015.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.13, "one"], [3.07, 3.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.79,\"end\":1.13},{\"word\":\"two\",\"start\":3.07,\"end\":3.56}]"} {"utterance_id": "8254-115543-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0019.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.97, "snake"]], "response_mode": "offline", "answer": "[{\"word\":\"snake\",\"start\":0.6,\"end\":0.97}]"} {"utterance_id": "6467-94831-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.86, 1.38, "thomas"], [4.93, 5.63, "george strong"]], "response_mode": "offline", "answer": "[{\"word\":\"thomas\",\"start\":0.86,\"end\":1.38},{\"word\":\"george strong\",\"start\":4.93,\"end\":5.63}]"} {"utterance_id": "1585-131718-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0035.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 0.85, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":0.65,\"end\":0.85}]"} {"utterance_id": "1585-131718-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0052.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen carefully to the entire clip, then report every instance of a disease name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 2.2, "pleurisy"]], "response_mode": "offline", "answer": "[{\"word\":\"pleurisy\",\"start\":1.58,\"end\":2.2}]"} {"utterance_id": "7641-96252-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0020.flac", "task_type": "lexical_detection", "query": "follows", "prompt": "After listening to the whole recording, output every instance of the word \"follows\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.65, 6.24, "follows"]], "response_mode": "offline", "answer": "[{\"word\":\"follows\",\"start\":5.65,\"end\":6.24}]"} {"utterance_id": "6467-94831-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0008.flac", "task_type": "lexical_detection", "query": "moaned", "prompt": "After the audio ends, extract all occurrences of the word \"moaned\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.7, "moaned"]], "response_mode": "offline", "answer": "[{\"word\":\"moaned\",\"start\":0.32,\"end\":0.7}]"} {"utterance_id": "3915-98647-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0006.flac", "task_type": "lexical_detection", "query": "brings", "prompt": "After listening to the whole recording, output every instance of the word \"brings\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.18, "brings"]], "response_mode": "offline", "answer": "[{\"word\":\"brings\",\"start\":0.91,\"end\":1.18}]"} {"utterance_id": "1701-141759-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0033.flac", "task_type": "lexical_detection", "query": "high", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"high\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.43, 2.8, "high"]], "response_mode": "offline", "answer": "[{\"word\":\"high\",\"start\":2.43,\"end\":2.8}]"} {"utterance_id": "4572-112383-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0006.flac", "task_type": "lexical_detection", "query": "saints", "prompt": "After listening to the whole recording, output every instance of the word \"saints\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.77, "saints"]], "response_mode": "offline", "answer": "[{\"word\":\"saints\",\"start\":1.25,\"end\":1.77}]"} {"utterance_id": "116-288045-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0016.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[1.8, 2.26, "one"], [6.72, 7.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.8,\"end\":2.26},{\"word\":\"two\",\"start\":6.72,\"end\":7.18}]"} {"utterance_id": "6599-38591-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0011.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Once the audio has finished, list each occurrence of a number as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.46, 0.84, "one"], [3.79, 3.96, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.46,\"end\":0.84},{\"word\":\"two\",\"start\":3.79,\"end\":3.96}]"} {"utterance_id": "4831-18525-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0023.flac", "task_type": "lexical_detection", "query": "pictures", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"pictures\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.64, "pictures"]], "response_mode": "offline", "answer": "[{\"word\":\"pictures\",\"start\":1.23,\"end\":1.64}]"} {"utterance_id": "6267-65525-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0056.flac", "task_type": "lexical_detection", "query": "dull", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"dull\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.56, "dull"]], "response_mode": "offline", "answer": "[{\"word\":\"dull\",\"start\":1.22,\"end\":1.56}]"} {"utterance_id": "1630-141772-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0021.flac", "task_type": "lexical_detection", "query": "peaceful", "prompt": "When the utterance is complete, report all instances of the word \"peaceful\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.51, 3.08, "peaceful"]], "response_mode": "offline", "answer": "[{\"word\":\"peaceful\",\"start\":2.51,\"end\":3.08}]"} {"utterance_id": "5849-50873-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0016.flac", "task_type": "lexical_detection", "query": "doing", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"doing\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.61, "doing"]], "response_mode": "offline", "answer": "[{\"word\":\"doing\",\"start\":1.27,\"end\":1.61}]"} {"utterance_id": "7601-175351-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.91, "johnson's"]], "response_mode": "offline", "answer": "[{\"word\":\"johnson's\",\"start\":2.22,\"end\":2.91}]"} {"utterance_id": "6841-88291-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.4, "j h"]], "response_mode": "offline", "answer": "[{\"word\":\"j h\",\"start\":2.95,\"end\":3.4}]"} {"utterance_id": "1255-90407-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.27, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":0.78,\"end\":1.27}]"} {"utterance_id": "6267-53049-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.65, "teacher"]], "response_mode": "offline", "answer": "[{\"word\":\"teacher\",\"start\":5.09,\"end\":5.65}]"} {"utterance_id": "7641-96252-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0003.flac", "task_type": "lexical_detection", "query": "die", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"die\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.07, "die"]], "response_mode": "offline", "answer": "[{\"word\":\"die\",\"start\":2.75,\"end\":3.07}]"} {"utterance_id": "4323-18416-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0014.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 1.19, "grandpapa"], [4.58, 5.13, "grandpapa"]], "response_mode": "offline", "answer": "[{\"word\":\"grandpapa\",\"start\":0.64,\"end\":1.19},{\"word\":\"grandpapa\",\"start\":4.58,\"end\":5.13}]"} {"utterance_id": "3660-172182-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0006.flac", "task_type": "lexical_detection", "query": "insult", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"insult\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.49, "insult"]], "response_mode": "offline", "answer": "[{\"word\":\"insult\",\"start\":2.12,\"end\":2.49}]"} {"utterance_id": "6467-94831-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0030.flac", "task_type": "lexical_detection", "query": "raise", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"raise\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.71, "raise"]], "response_mode": "offline", "answer": "[{\"word\":\"raise\",\"start\":0.46,\"end\":0.71}]"} {"utterance_id": "8254-84205-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0019.flac", "task_type": "lexical_detection", "query": "they'd", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"they'd\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.94, 3.06, "they'd"]], "response_mode": "offline", "answer": "[{\"word\":\"they'd\",\"start\":2.94,\"end\":3.06}]"} {"utterance_id": "4323-18416-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0034.flac", "task_type": "lexical_detection", "query": "dog", "prompt": "Process the full audio offline. Find every instance of the word \"dog\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.27, "dog"]], "response_mode": "offline", "answer": "[{\"word\":\"dog\",\"start\":0.99,\"end\":1.27}]"} {"utterance_id": "3660-6517-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0025.flac", "task_type": "lexical_detection", "query": "grapple", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"grapple\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.93, "grapple"]], "response_mode": "offline", "answer": "[{\"word\":\"grapple\",\"start\":1.56,\"end\":1.93}]"} {"utterance_id": "6841-88291-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.22, 0.65, "homer"], [2.01, 2.45, "john"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":0.22,\"end\":0.65},{\"word\":\"john\",\"start\":2.01,\"end\":2.45}]"} {"utterance_id": "1701-141760-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.8, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"rostov\",\"start\":1.38,\"end\":1.8}]"} {"utterance_id": "6267-65525-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0031.flac", "task_type": "lexical_detection", "query": "tossed", "prompt": "Process the full audio offline. Find every instance of the word \"tossed\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.27, "tossed"]], "response_mode": "offline", "answer": "[{\"word\":\"tossed\",\"start\":2.77,\"end\":3.27}]"} {"utterance_id": "4153-186222-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0030.flac", "task_type": "lexical_detection", "query": "nieces", "prompt": "After the audio ends, extract all occurrences of the word \"nieces\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.34, "nieces"]], "response_mode": "offline", "answer": "[{\"word\":\"nieces\",\"start\":3.94,\"end\":4.34}]"} {"utterance_id": "1255-138279-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0010.flac", "task_type": "lexical_detection", "query": "dozen", "prompt": "After listening to the whole recording, output every instance of the word \"dozen\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.91, 4.21, "dozen"]], "response_mode": "offline", "answer": "[{\"word\":\"dozen\",\"start\":3.91,\"end\":4.21}]"} {"utterance_id": "4323-13259-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0019.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Once the audio has finished, identify every instance of a time of day. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.34, 1.85, "afternoon"]], "response_mode": "offline", "answer": "[{\"word\":\"afternoon\",\"start\":1.34,\"end\":1.85}]"} {"utterance_id": "6467-97061-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0017.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Once the audio has finished, identify every instance of a color name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.89, 2.26, "silver"]], "response_mode": "offline", "answer": "[{\"word\":\"silver\",\"start\":1.89,\"end\":2.26}]"} {"utterance_id": "116-288048-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0020.flac", "task_type": "lexical_detection", "query": "silence", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"silence\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.38, "silence"]], "response_mode": "offline", "answer": "[{\"word\":\"silence\",\"start\":2.02,\"end\":2.38}]"} {"utterance_id": "4831-18525-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0027.flac", "task_type": "lexical_detection", "query": "adela", "prompt": "Process the full audio offline. Find every instance of the word \"adela\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.76, "adela"]], "response_mode": "offline", "answer": "[{\"word\":\"adela\",\"start\":1.46,\"end\":1.76}]"} {"utterance_id": "7641-96684-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0036.flac", "task_type": "lexical_detection", "query": "whom", "prompt": "When the utterance is complete, report all instances of the word \"whom\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.48, "whom"]], "response_mode": "offline", "answer": "[{\"word\":\"whom\",\"start\":1.2,\"end\":1.48}]"} {"utterance_id": "1686-142278-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0067.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.97, "frederick"], [2.31, 2.72, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"frederick\",\"start\":0.52,\"end\":0.97},{\"word\":\"margaret\",\"start\":2.31,\"end\":2.72}]"} {"utterance_id": "7641-96252-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0007.flac", "task_type": "lexical_detection", "query": "consequently", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"consequently\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.94, "consequently"]], "response_mode": "offline", "answer": "[{\"word\":\"consequently\",\"start\":3.22,\"end\":3.94}]"} {"utterance_id": "4515-11057-0103", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0103.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After the audio ends, extract all occurrences of a time of day. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.85, 2.19, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.85,\"end\":2.19}]"} {"utterance_id": "4323-55228-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0040.flac", "task_type": "lexical_detection", "query": "promptly", "prompt": "Process the full audio offline. Find every instance of the word \"promptly\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.33, 2.74, "promptly"]], "response_mode": "offline", "answer": "[{\"word\":\"promptly\",\"start\":2.33,\"end\":2.74}]"} {"utterance_id": "116-288048-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.22, 3.92, "peter paul john james judas"]], "response_mode": "offline", "answer": "[{\"word\":\"peter paul john james judas\",\"start\":0.22,\"end\":3.92}]"} {"utterance_id": "6455-67803-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0030.flac", "task_type": "lexical_detection", "query": "whole", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"whole\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.62, 4.9, "whole"]], "response_mode": "offline", "answer": "[{\"word\":\"whole\",\"start\":4.62,\"end\":4.9}]"} {"utterance_id": "700-122866-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0042.flac", "task_type": "lexical_detection", "query": "felt", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"felt\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.79, 2.07, "felt"]], "response_mode": "offline", "answer": "[{\"word\":\"felt\",\"start\":1.79,\"end\":2.07}]"} {"utterance_id": "8254-84205-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0032.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After listening to the whole recording, output every instance of a plant name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.46, "trees"]], "response_mode": "offline", "answer": "[{\"word\":\"trees\",\"start\":3.05,\"end\":3.46}]"} {"utterance_id": "4515-11057-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0098.flac", "task_type": "lexical_detection", "query": "months", "prompt": "After the audio ends, extract all occurrences of the word \"months\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.57, 6.06, "months"]], "response_mode": "offline", "answer": "[{\"word\":\"months\",\"start\":5.57,\"end\":6.06}]"} {"utterance_id": "4515-11057-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0011.flac", "task_type": "lexical_detection", "query": "poverty", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"poverty\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.54, 1.1, "poverty"]], "response_mode": "offline", "answer": "[{\"word\":\"poverty\",\"start\":0.54,\"end\":1.1}]"} {"utterance_id": "1650-167613-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0049.flac", "task_type": "lexical_detection", "query": "you're", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"you're\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.49, "you're"]], "response_mode": "offline", "answer": "[{\"word\":\"you're\",\"start\":1.26,\"end\":1.49}]"} {"utterance_id": "8254-84205-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0001.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "You will hear a complete utterance. When it ends, list every instance of a disease name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.09, 10.49, "wounds"]], "response_mode": "offline", "answer": "[{\"word\":\"wounds\",\"start\":10.09,\"end\":10.49}]"} {"utterance_id": "3663-172528-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0052.flac", "task_type": "lexical_detection", "query": "piece", "prompt": "Process the full audio offline. Find every instance of the word \"piece\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.36, 4.83, "piece"]], "response_mode": "offline", "answer": "[{\"word\":\"piece\",\"start\":4.36,\"end\":4.83}]"} {"utterance_id": "4831-25894-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0032.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.77, 6.06, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":5.77,\"end\":6.06}]"} {"utterance_id": "6455-67803-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0013.flac", "task_type": "lexical_detection", "query": "reasons", "prompt": "After listening to the whole recording, output every instance of the word \"reasons\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.67, "reasons"]], "response_mode": "offline", "answer": "[{\"word\":\"reasons\",\"start\":1.34,\"end\":1.67}]"} {"utterance_id": "8288-274162-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0035.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 4.0, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":3.6,\"end\":4.0}]"} {"utterance_id": "7641-96670-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0005.flac", "task_type": "lexical_counting", "query": "city", "prompt": "Process the full audio offline. For every occurrence of the word \"city\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.62, "one"], [6.65, 6.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.27,\"end\":1.62},{\"word\":\"two\",\"start\":6.65,\"end\":6.94}]"} {"utterance_id": "4323-18416-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.86, "mister king"], [2.9, 3.23, "polly's"], [5.05, 5.41, "charlotte"]], "response_mode": "offline", "answer": "[{\"word\":\"mister king\",\"start\":1.39,\"end\":1.86},{\"word\":\"polly's\",\"start\":2.9,\"end\":3.23},{\"word\":\"charlotte\",\"start\":5.05,\"end\":5.41}]"} {"utterance_id": "8288-274162-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0033.flac", "task_type": "lexical_detection", "query": "foolish", "prompt": "After listening to the whole recording, output every instance of the word \"foolish\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.11, 4.47, "foolish"]], "response_mode": "offline", "answer": "[{\"word\":\"foolish\",\"start\":4.11,\"end\":4.47}]"} {"utterance_id": "5849-50962-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0022.flac", "task_type": "lexical_detection", "query": "declare", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"declare\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.59, 1.17, "declare"]], "response_mode": "offline", "answer": "[{\"word\":\"declare\",\"start\":0.59,\"end\":1.17}]"} {"utterance_id": "6599-38591-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.92, 10.28, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":9.92,\"end\":10.28}]"} {"utterance_id": "4515-11057-0113", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0113.flac", "task_type": "lexical_detection", "query": "several", "prompt": "Process the full audio offline. Find every instance of the word \"several\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.12, 0.53, "several"]], "response_mode": "offline", "answer": "[{\"word\":\"several\",\"start\":0.12,\"end\":0.53}]"} {"utterance_id": "1686-142278-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0033.flac", "task_type": "lexical_detection", "query": "charged", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"charged\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.68, "charged"]], "response_mode": "offline", "answer": "[{\"word\":\"charged\",\"start\":1.33,\"end\":1.68}]"} {"utterance_id": "6267-65525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0000.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After the audio ends, extract all occurrences of a color name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.82, "blue"]], "response_mode": "offline", "answer": "[{\"word\":\"blue\",\"start\":1.47,\"end\":1.82}]"} {"utterance_id": "7697-245712-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.27, 0.77, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":0.27,\"end\":0.77}]"} {"utterance_id": "4831-25894-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0020.flac", "task_type": "lexical_detection", "query": "chuckling", "prompt": "After listening to the whole recording, output every instance of the word \"chuckling\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.81, 7.29, "chuckling"]], "response_mode": "offline", "answer": "[{\"word\":\"chuckling\",\"start\":6.81,\"end\":7.29}]"} {"utterance_id": "8254-84205-0074", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0074.flac", "task_type": "lexical_detection", "query": "settled", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"settled\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.87, 3.16, "settled"]], "response_mode": "offline", "answer": "[{\"word\":\"settled\",\"start\":2.87,\"end\":3.16}]"} {"utterance_id": "4831-29134-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0011.flac", "task_type": "lexical_detection", "query": "faces", "prompt": "After the audio ends, extract all occurrences of the word \"faces\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.02, 3.46, "faces"]], "response_mode": "offline", "answer": "[{\"word\":\"faces\",\"start\":3.02,\"end\":3.46}]"} {"utterance_id": "1630-73710-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0010.flac", "task_type": "lexical_detection", "query": "henrietta", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"henrietta\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.61, 4.48, "henrietta"]], "response_mode": "offline", "answer": "[{\"word\":\"henrietta\",\"start\":3.61,\"end\":4.48}]"} {"utterance_id": "8254-84205-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0040.flac", "task_type": "lexical_detection", "query": "redskins", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"redskins\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.29, "redskins"]], "response_mode": "offline", "answer": "[{\"word\":\"redskins\",\"start\":0.76,\"end\":1.29}]"} {"utterance_id": "4831-18525-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0027.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.86, 6.1, "eye"]], "response_mode": "offline", "answer": "[{\"word\":\"eye\",\"start\":5.86,\"end\":6.1}]"} {"utterance_id": "3915-57461-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0004.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Once the audio has finished, list each occurrence of a profession as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.55, 0.93, "one"], [2.94, 3.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.55,\"end\":0.93},{\"word\":\"two\",\"start\":2.94,\"end\":3.28}]"} {"utterance_id": "1701-141759-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.11, 4.52, "eyes"], [10.7, 11.18, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":4.11,\"end\":4.52},{\"word\":\"eyes\",\"start\":10.7,\"end\":11.18}]"} {"utterance_id": "8288-274162-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0007.flac", "task_type": "lexical_detection", "query": "deuce", "prompt": "Process the full audio offline. Find every instance of the word \"deuce\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.03, "deuce"]], "response_mode": "offline", "answer": "[{\"word\":\"deuce\",\"start\":0.73,\"end\":1.03}]"} {"utterance_id": "7641-96684-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.43, 0.62, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.43,\"end\":0.62}]"} {"utterance_id": "4153-185072-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0004.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.71, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":0.45,\"end\":0.71}]"} {"utterance_id": "4515-11057-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0045.flac", "task_type": "lexical_detection", "query": "began", "prompt": "After listening to the whole recording, output every instance of the word \"began\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 1.06, "began"]], "response_mode": "offline", "answer": "[{\"word\":\"began\",\"start\":0.66,\"end\":1.06}]"} {"utterance_id": "4153-186222-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0033.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.76, "elmhurst"]], "response_mode": "offline", "answer": "[{\"word\":\"elmhurst\",\"start\":1.34,\"end\":1.76}]"} {"utterance_id": "1630-96099-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0005.flac", "task_type": "lexical_detection", "query": "disgrace", "prompt": "When the utterance is complete, report all instances of the word \"disgrace\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.34, 3.98, "disgrace"]], "response_mode": "offline", "answer": "[{\"word\":\"disgrace\",\"start\":3.34,\"end\":3.98}]"} {"utterance_id": "6267-65525-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0037.flac", "task_type": "lexical_detection", "query": "evergreens", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"evergreens\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.65, 4.45, "evergreens"]], "response_mode": "offline", "answer": "[{\"word\":\"evergreens\",\"start\":3.65,\"end\":4.45}]"} {"utterance_id": "1650-173551-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0008.flac", "task_type": "lexical_detection", "query": "pushed", "prompt": "Process the full audio offline. Find every instance of the word \"pushed\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.74, 9.1, "pushed"]], "response_mode": "offline", "answer": "[{\"word\":\"pushed\",\"start\":8.74,\"end\":9.1}]"} {"utterance_id": "4570-14911-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0007.flac", "task_type": "lexical_detection", "query": "formed", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"formed\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.89, 9.36, "formed"]], "response_mode": "offline", "answer": "[{\"word\":\"formed\",\"start\":8.89,\"end\":9.36}]"} {"utterance_id": "4831-25894-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.63, 1.01, "tessa's"], [3.0, 3.5, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa's\",\"start\":0.63,\"end\":1.01},{\"word\":\"tommo\",\"start\":3.0,\"end\":3.5}]"} {"utterance_id": "7601-101622-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0002.flac", "task_type": "lexical_detection", "query": "send", "prompt": "After listening to the whole recording, output every instance of the word \"send\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.67, 5.02, "send"]], "response_mode": "offline", "answer": "[{\"word\":\"send\",\"start\":4.67,\"end\":5.02}]"} {"utterance_id": "4572-112375-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0005.flac", "task_type": "lexical_counting", "query": "thy", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"thy\" with word/start/end. If none, reply [].", "labels": [[0.87, 1.14, "one"], [3.62, 3.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.87,\"end\":1.14},{\"word\":\"two\",\"start\":3.62,\"end\":3.78}]"} {"utterance_id": "1585-131718-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0032.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.22, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.02,\"end\":1.22}]"} {"utterance_id": "4323-18416-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0020.flac", "task_type": "lexical_detection", "query": "slowly", "prompt": "When the utterance is complete, report all instances of the word \"slowly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 3.99, "slowly"]], "response_mode": "offline", "answer": "[{\"word\":\"slowly\",\"start\":3.6,\"end\":3.99}]"} {"utterance_id": "8173-294714-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0050.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Process the full audio offline. Find every instance of a money or currency term and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.02, 5.34, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":5.02,\"end\":5.34}]"} {"utterance_id": "1630-141772-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0005.flac", "task_type": "lexical_detection", "query": "wound", "prompt": "Once the audio has finished, identify every instance of the word \"wound\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[10.75, 11.1, "wound"]], "response_mode": "offline", "answer": "[{\"word\":\"wound\",\"start\":10.75,\"end\":11.1}]"} {"utterance_id": "1686-142278-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0041.flac", "task_type": "lexical_detection", "query": "yes", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"yes\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.99, "yes"]], "response_mode": "offline", "answer": "[{\"word\":\"yes\",\"start\":0.51,\"end\":0.99}]"} {"utterance_id": "1686-142278-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0062.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.96, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.5,\"end\":0.96}]"} {"utterance_id": "1686-142278-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0052.flac", "task_type": "lexical_detection", "query": "perceived", "prompt": "After listening to the whole recording, output every instance of the word \"perceived\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.03, 5.51, "perceived"]], "response_mode": "offline", "answer": "[{\"word\":\"perceived\",\"start\":5.03,\"end\":5.51}]"} {"utterance_id": "1686-142278-0095", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0095.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[2.57, 3.01, "child"]], "response_mode": "offline", "answer": "[{\"word\":\"child\",\"start\":2.57,\"end\":3.01}]"} {"utterance_id": "1650-167613-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0039.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Do not answer until the audio ends. Then list each instance of a money or currency term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.99, 4.45, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":3.99,\"end\":4.45}]"} {"utterance_id": "6841-88291-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0048.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 0.9, "noon"]], "response_mode": "offline", "answer": "[{\"word\":\"noon\",\"start\":0.7,\"end\":0.9}]"} {"utterance_id": "1585-131718-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0008.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "After listening to the whole recording, output timed count words for an organization name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.48, 3.05, "one"], [6.16, 6.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.48,\"end\":3.05},{\"word\":\"two\",\"start\":6.16,\"end\":6.72}]"} {"utterance_id": "2506-13150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.09, 8.78, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":8.09,\"end\":8.78}]"} {"utterance_id": "8288-274162-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0030.flac", "task_type": "lexical_detection", "query": "completely", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"completely\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.57, 1.07, "completely"]], "response_mode": "offline", "answer": "[{\"word\":\"completely\",\"start\":0.57,\"end\":1.07}]"} {"utterance_id": "4831-18525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 1.07, "tom selwyn"]], "response_mode": "offline", "answer": "[{\"word\":\"tom selwyn\",\"start\":0.37,\"end\":1.07}]"} {"utterance_id": "8173-294714-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0051.flac", "task_type": "lexical_detection", "query": "sadly", "prompt": "When the utterance is complete, report all instances of the word \"sadly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.7, "sadly"]], "response_mode": "offline", "answer": "[{\"word\":\"sadly\",\"start\":2.1,\"end\":2.7}]"} {"utterance_id": "700-122868-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0029.flac", "task_type": "lexical_detection", "query": "hesitated", "prompt": "When the utterance is complete, report all instances of the word \"hesitated\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.91, "hesitated"]], "response_mode": "offline", "answer": "[{\"word\":\"hesitated\",\"start\":1.25,\"end\":1.91}]"} {"utterance_id": "1650-167613-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0054.flac", "task_type": "lexical_detection", "query": "succeeded", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"succeeded\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.62, "succeeded"]], "response_mode": "offline", "answer": "[{\"word\":\"succeeded\",\"start\":0.96,\"end\":1.62}]"} {"utterance_id": "3663-172528-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0027.flac", "task_type": "lexical_detection", "query": "seated", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"seated\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.53, "seated"]], "response_mode": "offline", "answer": "[{\"word\":\"seated\",\"start\":1.09,\"end\":1.53}]"} {"utterance_id": "3663-172005-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0007.flac", "task_type": "lexical_detection", "query": "oars", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"oars\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.64, 5.08, "oars"]], "response_mode": "offline", "answer": "[{\"word\":\"oars\",\"start\":4.64,\"end\":5.08}]"} {"utterance_id": "116-288048-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0003.flac", "task_type": "lexical_counting", "query": "infant", "prompt": "After the audio ends, tally every instance of the word \"infant\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[5.6, 5.92, "one"], [12.35, 12.67, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.6,\"end\":5.92},{\"word\":\"two\",\"start\":12.35,\"end\":12.67}]"} {"utterance_id": "7601-175351-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.39, 1.04, "hatteras"], [6.26, 7.67, "sir edward belcher"]], "response_mode": "offline", "answer": "[{\"word\":\"hatteras\",\"start\":0.39,\"end\":1.04},{\"word\":\"sir edward belcher\",\"start\":6.26,\"end\":7.67}]"} {"utterance_id": "7697-105815-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0006.flac", "task_type": "lexical_detection", "query": "large", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"large\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 1.85, "large"]], "response_mode": "offline", "answer": "[{\"word\":\"large\",\"start\":1.52,\"end\":1.85}]"} {"utterance_id": "4831-25894-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0014.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.26, 4.63, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":4.26,\"end\":4.63}]"} {"utterance_id": "5849-50963-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0002.flac", "task_type": "lexical_detection", "query": "bar", "prompt": "Once the audio has finished, identify every instance of the word \"bar\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.69, 4.06, "bar"]], "response_mode": "offline", "answer": "[{\"word\":\"bar\",\"start\":3.69,\"end\":4.06}]"} {"utterance_id": "7697-105815-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0037.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.44, 1.69, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.44,\"end\":1.69}]"} {"utterance_id": "1686-142278-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0075.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.17, 4.61, "milton"]], "response_mode": "offline", "answer": "[{\"word\":\"milton\",\"start\":4.17,\"end\":4.61}]"} {"utterance_id": "5849-50873-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0010.flac", "task_type": "lexical_detection", "query": "promised", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"promised\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.0, 3.46, "promised"]], "response_mode": "offline", "answer": "[{\"word\":\"promised\",\"start\":3.0,\"end\":3.46}]"} {"utterance_id": "6841-88291-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0033.flac", "task_type": "lexical_detection", "query": "history", "prompt": "After the audio ends, extract all occurrences of the word \"history\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.23, "history"]], "response_mode": "offline", "answer": "[{\"word\":\"history\",\"start\":2.77,\"end\":3.23}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Process the full audio offline. Find every instance of a color name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.08, 12.49, "black"], [26.75, 27.18, "black"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":12.08,\"end\":12.49},{\"word\":\"black\",\"start\":26.75,\"end\":27.18}]"} {"utterance_id": "6841-88294-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0032.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.44, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":1.12,\"end\":1.44}]"} {"utterance_id": "4153-186222-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.93, "louise"]], "response_mode": "offline", "answer": "[{\"word\":\"louise\",\"start\":0.53,\"end\":0.93}]"} {"utterance_id": "1255-90407-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0026.flac", "task_type": "lexical_detection", "query": "temporary", "prompt": "Process the full audio offline. Find every instance of the word \"temporary\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.82, "temporary"]], "response_mode": "offline", "answer": "[{\"word\":\"temporary\",\"start\":1.41,\"end\":1.82}]"} {"utterance_id": "6267-65525-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0059.flac", "task_type": "lexical_detection", "query": "you'll", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"you'll\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.09, 1.35, "you'll"]], "response_mode": "offline", "answer": "[{\"word\":\"you'll\",\"start\":1.09,\"end\":1.35}]"} {"utterance_id": "1701-141759-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0006.flac", "task_type": "lexical_detection", "query": "sonya", "prompt": "After listening to the whole recording, output every instance of the word \"sonya\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.0, "sonya"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":1.55,\"end\":2.0}]"} {"utterance_id": "4572-112381-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0014.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.75, 6.54, "wives"]], "response_mode": "offline", "answer": "[{\"word\":\"wives\",\"start\":5.75,\"end\":6.54}]"} {"utterance_id": "4831-29134-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0012.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.34, "teachers"]], "response_mode": "offline", "answer": "[{\"word\":\"teachers\",\"start\":0.88,\"end\":1.34}]"} {"utterance_id": "8288-274162-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0004.flac", "task_type": "lexical_detection", "query": "recalled", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"recalled\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.93, "recalled"]], "response_mode": "offline", "answer": "[{\"word\":\"recalled\",\"start\":1.45,\"end\":1.93}]"} {"utterance_id": "6123-59150-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.55, 2.21, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":1.55,\"end\":2.21}]"} {"utterance_id": "8254-84205-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0067.flac", "task_type": "lexical_detection", "query": "griggs", "prompt": "After the audio ends, extract all occurrences of the word \"griggs\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.1, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.74,\"end\":2.1}]"} {"utterance_id": "3660-6517-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0028.flac", "task_type": "lexical_detection", "query": "gained", "prompt": "After listening to the whole recording, output every instance of the word \"gained\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.92, "gained"]], "response_mode": "offline", "answer": "[{\"word\":\"gained\",\"start\":0.64,\"end\":0.92}]"} {"utterance_id": "8288-274162-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0044.flac", "task_type": "lexical_detection", "query": "raoul", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"raoul\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.88, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":0.5,\"end\":0.88}]"} {"utterance_id": "6267-65525-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.55, 3.4, "paul moore"]], "response_mode": "offline", "answer": "[{\"word\":\"paul moore\",\"start\":2.55,\"end\":3.4}]"} {"utterance_id": "3915-57461-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0006.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 3.0, "cousin"]], "response_mode": "offline", "answer": "[{\"word\":\"cousin\",\"start\":2.57,\"end\":3.0}]"} {"utterance_id": "6841-88294-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0046.flac", "task_type": "lexical_counting", "query": "right", "prompt": "After the audio ends, tally every instance of the word \"right\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.92, 1.11, "one"], [2.68, 2.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.92,\"end\":1.11},{\"word\":\"two\",\"start\":2.68,\"end\":2.92}]"} {"utterance_id": "4831-18525-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.41, 4.83, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":4.41,\"end\":4.83}]"} {"utterance_id": "1255-90407-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0000.flac", "task_type": "lexical_detection", "query": "hoped", "prompt": "When the utterance is complete, report all instances of the word \"hoped\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.55, 0.78, "hoped"]], "response_mode": "offline", "answer": "[{\"word\":\"hoped\",\"start\":0.55,\"end\":0.78}]"} {"utterance_id": "7697-105815-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0036.flac", "task_type": "lexical_detection", "query": "lodge", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"lodge\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.66, 1.97, "lodge"]], "response_mode": "offline", "answer": "[{\"word\":\"lodge\",\"start\":1.66,\"end\":1.97}]"} {"utterance_id": "6841-88294-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0064.flac", "task_type": "lexical_detection", "query": "retorted", "prompt": "After the audio ends, extract all occurrences of the word \"retorted\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.47, 2.99, "retorted"]], "response_mode": "offline", "answer": "[{\"word\":\"retorted\",\"start\":2.47,\"end\":2.99}]"} {"utterance_id": "2506-169427-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.71, "eugenius"]], "response_mode": "offline", "answer": "[{\"word\":\"eugenius\",\"start\":1.17,\"end\":1.71}]"} {"utterance_id": "5849-50873-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0040.flac", "task_type": "lexical_detection", "query": "colony", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"colony\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 3.95, "colony"]], "response_mode": "offline", "answer": "[{\"word\":\"colony\",\"start\":3.52,\"end\":3.95}]"} {"utterance_id": "7601-101619-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0002.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.84, 4.17, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.84,\"end\":4.17}]"} {"utterance_id": "1255-90413-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0027.flac", "task_type": "lexical_counting", "query": "left", "prompt": "Once the audio has finished, list each occurrence of the word \"left\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.46, 1.88, "one"], [9.84, 10.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.46,\"end\":1.88},{\"word\":\"two\",\"start\":9.84,\"end\":10.18}]"} {"utterance_id": "7641-96670-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0023.flac", "task_type": "lexical_detection", "query": "yield", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"yield\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.58, 7.84, "yield"]], "response_mode": "offline", "answer": "[{\"word\":\"yield\",\"start\":7.58,\"end\":7.84}]"} {"utterance_id": "700-122868-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0000.flac", "task_type": "lexical_detection", "query": "elaine", "prompt": "Once the audio has finished, identify every instance of the word \"elaine\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.48, 1.83, "elaine"]], "response_mode": "offline", "answer": "[{\"word\":\"elaine\",\"start\":1.48,\"end\":1.83}]"} {"utterance_id": "8254-84205-0071", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0071.flac", "task_type": "lexical_detection", "query": "shouldn't", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"shouldn't\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.53, 2.78, "shouldn't"]], "response_mode": "offline", "answer": "[{\"word\":\"shouldn't\",\"start\":2.53,\"end\":2.78}]"} {"utterance_id": "4323-13259-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0022.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.67, 7.94, "liquor"]], "response_mode": "offline", "answer": "[{\"word\":\"liquor\",\"start\":7.67,\"end\":7.94}]"} {"utterance_id": "4323-55228-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0036.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Once the audio has finished, identify every instance of a date. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.73, 6.98, "morrow"]], "response_mode": "offline", "answer": "[{\"word\":\"morrow\",\"start\":6.73,\"end\":6.98}]"} {"utterance_id": "6267-65525-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0026.flac", "task_type": "lexical_detection", "query": "give", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"give\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.63, 2.93, "give"]], "response_mode": "offline", "answer": "[{\"word\":\"give\",\"start\":2.63,\"end\":2.93}]"} {"utterance_id": "700-122867-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0016.flac", "task_type": "lexical_detection", "query": "feels", "prompt": "After the audio ends, extract all occurrences of the word \"feels\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.19, "feels"]], "response_mode": "offline", "answer": "[{\"word\":\"feels\",\"start\":0.9,\"end\":1.19}]"} {"utterance_id": "3660-172183-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0010.flac", "task_type": "lexical_detection", "query": "enid", "prompt": "After listening to the whole recording, output every instance of the word \"enid\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.51, "enid"]], "response_mode": "offline", "answer": "[{\"word\":\"enid\",\"start\":1.23,\"end\":1.51}]"} {"utterance_id": "6841-88291-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0034.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.79, 9.09, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":8.79,\"end\":9.09}]"} {"utterance_id": "2506-11278-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0016.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.11, 8.53, "score"]], "response_mode": "offline", "answer": "[{\"word\":\"score\",\"start\":8.11,\"end\":8.53}]"} {"utterance_id": "4570-102353-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0004.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.35, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":0.72,\"end\":1.35}]"} {"utterance_id": "6841-88291-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0053.flac", "task_type": "lexical_detection", "query": "seventy", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"seventy\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.34, "seventy"]], "response_mode": "offline", "answer": "[{\"word\":\"seventy\",\"start\":1.97,\"end\":2.34}]"} {"utterance_id": "7641-96670-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.59, 3.0, "demeter"]], "response_mode": "offline", "answer": "[{\"word\":\"demeter\",\"start\":2.59,\"end\":3.0}]"} {"utterance_id": "6467-97061-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.28, 5.72, "prince"]], "response_mode": "offline", "answer": "[{\"word\":\"prince\",\"start\":5.28,\"end\":5.72}]"} {"utterance_id": "116-288045-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0026.flac", "task_type": "lexical_detection", "query": "turning", "prompt": "Once the audio has finished, identify every instance of the word \"turning\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.61, 2.91, "turning"]], "response_mode": "offline", "answer": "[{\"word\":\"turning\",\"start\":2.61,\"end\":2.91}]"} {"utterance_id": "4572-112381-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0008.flac", "task_type": "lexical_detection", "query": "dread", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"dread\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.22, 11.76, "dread"]], "response_mode": "offline", "answer": "[{\"word\":\"dread\",\"start\":11.22,\"end\":11.76}]"} {"utterance_id": "4515-11057-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0028.flac", "task_type": "lexical_detection", "query": "slight", "prompt": "Once the audio has finished, identify every instance of the word \"slight\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.67, 4.13, "slight"]], "response_mode": "offline", "answer": "[{\"word\":\"slight\",\"start\":3.67,\"end\":4.13}]"} {"utterance_id": "6123-59186-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0021.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.29, 6.79, "oxen"]], "response_mode": "offline", "answer": "[{\"word\":\"oxen\",\"start\":6.29,\"end\":6.79}]"} {"utterance_id": "6123-59150-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0018.flac", "task_type": "lexical_detection", "query": "artist", "prompt": "After listening to the whole recording, output every instance of the word \"artist\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 1.95, "artist"]], "response_mode": "offline", "answer": "[{\"word\":\"artist\",\"start\":1.54,\"end\":1.95}]"} {"utterance_id": "7601-291468-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0004.flac", "task_type": "lexical_detection", "query": "public", "prompt": "Once the audio has finished, identify every instance of the word \"public\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[20.65, 21.1, "public"]], "response_mode": "offline", "answer": "[{\"word\":\"public\",\"start\":20.65,\"end\":21.1}]"} {"utterance_id": "5849-50962-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0017.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.21, 2.61, "one"], [6.94, 7.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.21,\"end\":2.61},{\"word\":\"two\",\"start\":6.94,\"end\":7.28}]"} {"utterance_id": "1630-102884-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0014.flac", "task_type": "lexical_detection", "query": "realise", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"realise\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.08, 5.74, "realise"]], "response_mode": "offline", "answer": "[{\"word\":\"realise\",\"start\":5.08,\"end\":5.74}]"} {"utterance_id": "8254-84205-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0070.flac", "task_type": "lexical_detection", "query": "ejaculated", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"ejaculated\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.14, 0.73, "ejaculated"]], "response_mode": "offline", "answer": "[{\"word\":\"ejaculated\",\"start\":0.14,\"end\":0.73}]"} {"utterance_id": "8173-294714-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0000.flac", "task_type": "lexical_detection", "query": "prison", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"prison\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.53, 2.96, "prison"]], "response_mode": "offline", "answer": "[{\"word\":\"prison\",\"start\":2.53,\"end\":2.96}]"} {"utterance_id": "6841-88291-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0050.flac", "task_type": "lexical_detection", "query": "latter", "prompt": "When the utterance is complete, report all instances of the word \"latter\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 0.63, "latter"]], "response_mode": "offline", "answer": "[{\"word\":\"latter\",\"start\":0.35,\"end\":0.63}]"} {"utterance_id": "8254-84205-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0063.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.79, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.47,\"end\":1.79}]"} {"utterance_id": "5543-27761-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0010.flac", "task_type": "lexical_detection", "query": "prove", "prompt": "When the utterance is complete, report all instances of the word \"prove\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.49, "prove"]], "response_mode": "offline", "answer": "[{\"word\":\"prove\",\"start\":2.22,\"end\":2.49}]"} {"utterance_id": "700-122867-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0009.flac", "task_type": "lexical_detection", "query": "slid", "prompt": "Process the full audio offline. Find every instance of the word \"slid\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.2, "slid"]], "response_mode": "offline", "answer": "[{\"word\":\"slid\",\"start\":0.84,\"end\":1.2}]"} {"utterance_id": "7697-105815-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0023.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.85, 2.21, "zion"]], "response_mode": "offline", "answer": "[{\"word\":\"zion\",\"start\":1.85,\"end\":2.21}]"} {"utterance_id": "6467-97061-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0006.flac", "task_type": "lexical_detection", "query": "performed", "prompt": "After listening to the whole recording, output every instance of the word \"performed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.96, "performed"]], "response_mode": "offline", "answer": "[{\"word\":\"performed\",\"start\":3.38,\"end\":3.96}]"} {"utterance_id": "1701-141760-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0023.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.58, 1.91, "beast"]], "response_mode": "offline", "answer": "[{\"word\":\"beast\",\"start\":1.58,\"end\":1.91}]"} {"utterance_id": "8288-274162-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0056.flac", "task_type": "lexical_detection", "query": "woman", "prompt": "Once the audio has finished, identify every instance of the word \"woman\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.94, 2.44, "woman"]], "response_mode": "offline", "answer": "[{\"word\":\"woman\",\"start\":1.94,\"end\":2.44}]"} {"utterance_id": "8288-274150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 2.08, "artist"]], "response_mode": "offline", "answer": "[{\"word\":\"artist\",\"start\":1.52,\"end\":2.08}]"} {"utterance_id": "6267-65525-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.2, 0.58, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":0.2,\"end\":0.58}]"} {"utterance_id": "8254-84205-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0036.flac", "task_type": "lexical_detection", "query": "farther", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"farther\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 2.98, "farther"]], "response_mode": "offline", "answer": "[{\"word\":\"farther\",\"start\":2.7,\"end\":2.98}]"} {"utterance_id": "3663-172528-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0025.flac", "task_type": "lexical_detection", "query": "habit", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"habit\" with fields word, start, end (seconds). If none, reply [].", "labels": [[14.08, 14.57, "habit"]], "response_mode": "offline", "answer": "[{\"word\":\"habit\",\"start\":14.08,\"end\":14.57}]"} {"utterance_id": "3663-172528-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0004.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.87, 2.38, "twice"], [9.36, 9.52, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"twice\",\"start\":1.87,\"end\":2.38},{\"word\":\"one\",\"start\":9.36,\"end\":9.52}]"} {"utterance_id": "116-288045-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0032.flac", "task_type": "lexical_counting", "query": "faith", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"faith\" with word/start/end. If none, reply [].", "labels": [[1.46, 1.68, "one"], [4.6, 4.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.46,\"end\":1.68},{\"word\":\"two\",\"start\":4.6,\"end\":4.94}]"} {"utterance_id": "4570-102353-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0001.flac", "task_type": "lexical_detection", "query": "examine", "prompt": "After the audio ends, extract all occurrences of the word \"examine\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.37, 3.99, "examine"]], "response_mode": "offline", "answer": "[{\"word\":\"examine\",\"start\":3.37,\"end\":3.99}]"} {"utterance_id": "4323-13259-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.1, 4.43, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":4.1,\"end\":4.43}]"} {"utterance_id": "7641-96684-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0008.flac", "task_type": "lexical_counting", "query": "voice", "prompt": "Listen to the full audio. Count every occurrence of the word \"voice\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.35, "one"], [2.87, 3.23, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.06,\"end\":1.35},{\"word\":\"two\",\"start\":2.87,\"end\":3.23}]"} {"utterance_id": "4515-11057-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0067.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[3.0, 3.41, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":3.0,\"end\":3.41}]"} {"utterance_id": "116-288045-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0008.flac", "task_type": "lexical_detection", "query": "gods", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"gods\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.81, "gods"]], "response_mode": "offline", "answer": "[{\"word\":\"gods\",\"start\":1.25,\"end\":1.81}]"} {"utterance_id": "7601-175351-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0015.flac", "task_type": "lexical_detection", "query": "beginning", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"beginning\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.74, 1.25, "beginning"]], "response_mode": "offline", "answer": "[{\"word\":\"beginning\",\"start\":0.74,\"end\":1.25}]"} {"utterance_id": "6467-62797-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.99, 2.26, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.99,\"end\":2.26}]"} {"utterance_id": "6267-53049-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 3.41, "missus fairweather"], [5.83, 6.66, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"missus fairweather\",\"start\":1.79,\"end\":3.41},{\"word\":\"penelope\",\"start\":5.83,\"end\":6.66}]"} {"utterance_id": "3663-172528-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.51, 6.14, "giorgio"]], "response_mode": "offline", "answer": "[{\"word\":\"giorgio\",\"start\":5.51,\"end\":6.14}]"} {"utterance_id": "4515-11057-0084", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0084.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.1, 4.55, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":4.1,\"end\":4.55}]"} {"utterance_id": "4831-18525-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0010.flac", "task_type": "lexical_detection", "query": "joel's", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"joel's\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.12, 5.61, "joel's"]], "response_mode": "offline", "answer": "[{\"word\":\"joel's\",\"start\":5.12,\"end\":5.61}]"} {"utterance_id": "6123-59150-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.21, 2.74, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":2.21,\"end\":2.74}]"} {"utterance_id": "1585-131718-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0039.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[3.1, 3.53, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":3.1,\"end\":3.53}]"} {"utterance_id": "3660-172182-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0027.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.43, 5.74, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":5.43,\"end\":5.74}]"} {"utterance_id": "1585-157660-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0001.flac", "task_type": "lexical_detection", "query": "faults", "prompt": "Process the full audio offline. Find every instance of the word \"faults\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.64, 4.21, "faults"]], "response_mode": "offline", "answer": "[{\"word\":\"faults\",\"start\":3.64,\"end\":4.21}]"} {"utterance_id": "1585-131718-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0047.flac", "task_type": "lexical_detection", "query": "amusement", "prompt": "When the utterance is complete, report all instances of the word \"amusement\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.85, "amusement"]], "response_mode": "offline", "answer": "[{\"word\":\"amusement\",\"start\":3.33,\"end\":3.85}]"} {"utterance_id": "6841-88294-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0004.flac", "task_type": "lexical_detection", "query": "sharp", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sharp\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.95, "sharp"]], "response_mode": "offline", "answer": "[{\"word\":\"sharp\",\"start\":1.56,\"end\":1.95}]"} {"utterance_id": "1585-131718-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0045.flac", "task_type": "lexical_detection", "query": "twelve", "prompt": "After listening to the whole recording, output every instance of the word \"twelve\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.44, "twelve"]], "response_mode": "offline", "answer": "[{\"word\":\"twelve\",\"start\":3.14,\"end\":3.44}]"} {"utterance_id": "6455-67804-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.73, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":1.19,\"end\":1.73}]"} {"utterance_id": "4572-112381-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0007.flac", "task_type": "lexical_detection", "query": "sugriva", "prompt": "Once the audio has finished, identify every instance of the word \"sugriva\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.01, 4.68, "sugriva"]], "response_mode": "offline", "answer": "[{\"word\":\"sugriva\",\"start\":4.01,\"end\":4.68}]"} {"utterance_id": "6841-88291-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0056.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a profession with word/start/end. If none, reply [].", "labels": [[3.46, 4.1, "one"], [6.26, 6.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.46,\"end\":4.1},{\"word\":\"two\",\"start\":6.26,\"end\":6.87}]"} {"utterance_id": "7601-101619-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0002.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.54, 1.31, "one"], [15.76, 16.75, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.54,\"end\":1.31},{\"word\":\"two\",\"start\":15.76,\"end\":16.75}]"} {"utterance_id": "6455-67804-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.94, 3.5, "mister binnie"]], "response_mode": "offline", "answer": "[{\"word\":\"mister binnie\",\"start\":2.94,\"end\":3.5}]"} {"utterance_id": "700-122866-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.26, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":0.85,\"end\":1.26}]"} {"utterance_id": "4323-18416-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.7, 4.26, "cousin eunice"], [6.02, 6.47, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"cousin eunice\",\"start\":3.7,\"end\":4.26},{\"word\":\"phronsie\",\"start\":6.02,\"end\":6.47}]"} {"utterance_id": "1630-73710-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0017.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.41, 6.86, "papa"]], "response_mode": "offline", "answer": "[{\"word\":\"papa\",\"start\":6.41,\"end\":6.86}]"} {"utterance_id": "6123-59186-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0017.flac", "task_type": "lexical_detection", "query": "mighty", "prompt": "When the utterance is complete, report all instances of the word \"mighty\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.39, 5.83, "mighty"]], "response_mode": "offline", "answer": "[{\"word\":\"mighty\",\"start\":5.39,\"end\":5.83}]"} {"utterance_id": "1650-157641-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.07, 2.73, "three score"]], "response_mode": "offline", "answer": "[{\"word\":\"three score\",\"start\":2.07,\"end\":2.73}]"} {"utterance_id": "4570-14911-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.64, 6.06, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.64,\"end\":6.06}]"} {"utterance_id": "7641-96684-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0016.flac", "task_type": "lexical_detection", "query": "man's", "prompt": "Process the full audio offline. Find every instance of the word \"man's\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.71, "man's"]], "response_mode": "offline", "answer": "[{\"word\":\"man's\",\"start\":1.4,\"end\":1.71}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "lexical_detection", "query": "annoyed", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"annoyed\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.78, 7.34, "annoyed"]], "response_mode": "offline", "answer": "[{\"word\":\"annoyed\",\"start\":6.78,\"end\":7.34}]"} {"utterance_id": "6455-67803-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0023.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.24, "one"], [3.73, 4.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.82,\"end\":2.24},{\"word\":\"two\",\"start\":3.73,\"end\":4.18}]"} {"utterance_id": "7641-96684-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.65, "apollo"], [4.34, 4.9, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":2.24,\"end\":2.65},{\"word\":\"admetus\",\"start\":4.34,\"end\":4.9}]"} {"utterance_id": "3915-57461-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.59, 9.06, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":8.59,\"end\":9.06}]"} {"utterance_id": "3915-57461-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.56, 2.89, "ears"]], "response_mode": "offline", "answer": "[{\"word\":\"ears\",\"start\":2.56,\"end\":2.89}]"} {"utterance_id": "3660-6517-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0021.flac", "task_type": "lexical_detection", "query": "occasionally", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"occasionally\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.93, 5.42, "occasionally"]], "response_mode": "offline", "answer": "[{\"word\":\"occasionally\",\"start\":4.93,\"end\":5.42}]"} {"utterance_id": "116-288048-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.63, 6.36, "celsus"]], "response_mode": "offline", "answer": "[{\"word\":\"celsus\",\"start\":5.63,\"end\":6.36}]"} {"utterance_id": "1585-131718-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0028.flac", "task_type": "lexical_detection", "query": "factory", "prompt": "Process the full audio offline. Find every instance of the word \"factory\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.33, 6.97, "factory"]], "response_mode": "offline", "answer": "[{\"word\":\"factory\",\"start\":6.33,\"end\":6.97}]"} {"utterance_id": "1585-131718-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0028.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.11, 7.73, "harrison"]], "response_mode": "offline", "answer": "[{\"word\":\"harrison\",\"start\":7.11,\"end\":7.73}]"} {"utterance_id": "6467-97061-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0008.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "After the audio ends, tally every instance of a family relation. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.51, 2.02, "one"], [7.82, 8.23, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.51,\"end\":2.02},{\"word\":\"two\",\"start\":7.82,\"end\":8.23}]"} {"utterance_id": "1650-167613-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0002.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.81, 6.52, "moscow"]], "response_mode": "offline", "answer": "[{\"word\":\"moscow\",\"start\":5.81,\"end\":6.52}]"} {"utterance_id": "6599-38591-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0004.flac", "task_type": "lexical_counting", "query": "spanish", "prompt": "Listen to the entire clip, then report the count sequence for the word \"spanish\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.41, "one"], [4.48, 4.95, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.93,\"end\":3.41},{\"word\":\"two\",\"start\":4.48,\"end\":4.95}]"} {"utterance_id": "6841-88294-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0007.flac", "task_type": "lexical_detection", "query": "days", "prompt": "After the audio ends, extract all occurrences of the word \"days\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.36, 2.85, "days"]], "response_mode": "offline", "answer": "[{\"word\":\"days\",\"start\":2.36,\"end\":2.85}]"} {"utterance_id": "7697-105817-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0007.flac", "task_type": "lexical_detection", "query": "hopeful", "prompt": "Once the audio has finished, identify every instance of the word \"hopeful\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.31, 0.84, "hopeful"]], "response_mode": "offline", "answer": "[{\"word\":\"hopeful\",\"start\":0.31,\"end\":0.84}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen carefully to the entire clip, then report every instance of a disease name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.81, 5.32, "plague"], [27.75, 28.21, "plague"]], "response_mode": "offline", "answer": "[{\"word\":\"plague\",\"start\":4.81,\"end\":5.32},{\"word\":\"plague\",\"start\":27.75,\"end\":28.21}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[10.59, 10.91, "porter"]], "response_mode": "offline", "answer": "[{\"word\":\"porter\",\"start\":10.59,\"end\":10.91}]"} {"utterance_id": "4570-102353-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.64, 1.61, "quartermaster general"]], "response_mode": "offline", "answer": "[{\"word\":\"quartermaster general\",\"start\":0.64,\"end\":1.61}]"} {"utterance_id": "1255-90407-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0020.flac", "task_type": "lexical_detection", "query": "encircling", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"encircling\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.85, 2.35, "encircling"]], "response_mode": "offline", "answer": "[{\"word\":\"encircling\",\"start\":1.85,\"end\":2.35}]"} {"utterance_id": "7641-96252-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0021.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.09, 8.4, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":8.09,\"end\":8.4}]"} {"utterance_id": "1686-142278-0082", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0082.flac", "task_type": "lexical_detection", "query": "remind", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"remind\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[15.88, 16.35, "remind"]], "response_mode": "offline", "answer": "[{\"word\":\"remind\",\"start\":15.88,\"end\":16.35}]"} {"utterance_id": "4515-11057-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0059.flac", "task_type": "lexical_detection", "query": "occurred", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"occurred\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.76, "occurred"]], "response_mode": "offline", "answer": "[{\"word\":\"occurred\",\"start\":1.34,\"end\":1.76}]"} {"utterance_id": "6267-53049-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0031.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.34, 7.66, "eye"]], "response_mode": "offline", "answer": "[{\"word\":\"eye\",\"start\":7.34,\"end\":7.66}]"} {"utterance_id": "4323-13259-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0018.flac", "task_type": "lexical_counting", "query": "seems", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"seems\" (word, start, end). If none, reply [].", "labels": [[2.86, 3.09, "one"], [5.2, 5.43, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.86,\"end\":3.09},{\"word\":\"two\",\"start\":5.2,\"end\":5.43}]"} {"utterance_id": "8173-294714-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0014.flac", "task_type": "lexical_detection", "query": "both", "prompt": "After the audio ends, extract all occurrences of the word \"both\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.43, "both"]], "response_mode": "offline", "answer": "[{\"word\":\"both\",\"start\":2.03,\"end\":2.43}]"} {"utterance_id": "5849-50873-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0041.flac", "task_type": "lexical_detection", "query": "colonies", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"colonies\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.4, 4.07, "colonies"]], "response_mode": "offline", "answer": "[{\"word\":\"colonies\",\"start\":3.4,\"end\":4.07}]"} {"utterance_id": "3915-57461-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0023.flac", "task_type": "lexical_detection", "query": "ways", "prompt": "Once the audio has finished, identify every instance of the word \"ways\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.72, 3.12, "ways"]], "response_mode": "offline", "answer": "[{\"word\":\"ways\",\"start\":2.72,\"end\":3.12}]"} {"utterance_id": "4515-11057-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0004.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.47, 5.31, "regent's park"]], "response_mode": "offline", "answer": "[{\"word\":\"regent's park\",\"start\":4.47,\"end\":5.31}]"} {"utterance_id": "8288-274162-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0022.flac", "task_type": "lexical_detection", "query": "says", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"says\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.11, "says"]], "response_mode": "offline", "answer": "[{\"word\":\"says\",\"start\":0.79,\"end\":1.11}]"} {"utterance_id": "1255-90407-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0022.flac", "task_type": "lexical_detection", "query": "admission", "prompt": "Process the full audio offline. Find every instance of the word \"admission\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.85, "admission"]], "response_mode": "offline", "answer": "[{\"word\":\"admission\",\"start\":1.4,\"end\":1.85}]"} {"utterance_id": "7641-96684-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.85, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":1.3,\"end\":1.85}]"} {"utterance_id": "1701-141760-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0022.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.51, 0.76, "one"], [10.68, 11.82, "two"], [14.26, 14.73, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.51,\"end\":0.76},{\"word\":\"two\",\"start\":10.68,\"end\":11.82},{\"word\":\"three\",\"start\":14.26,\"end\":14.73}]"} {"utterance_id": "1630-96099-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.95, 1.22, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.95,\"end\":1.22}]"} {"utterance_id": "4323-13259-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0025.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.56, 0.7, "one"], [2.62, 2.78, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.56,\"end\":0.7},{\"word\":\"one\",\"start\":2.62,\"end\":2.78}]"} {"utterance_id": "4831-18525-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.56, "polly"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":3.1,\"end\":3.56}]"} {"utterance_id": "4572-112383-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.83, 10.33, "sita"]], "response_mode": "offline", "answer": "[{\"word\":\"sita\",\"start\":9.83,\"end\":10.33}]"} {"utterance_id": "6267-65525-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0049.flac", "task_type": "lexical_detection", "query": "four", "prompt": "When the utterance is complete, report all instances of the word \"four\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.07, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":0.77,\"end\":1.07}]"} {"utterance_id": "7697-105815-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0009.flac", "task_type": "lexical_detection", "query": "tomb", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"tomb\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.6, 6.03, "tomb"]], "response_mode": "offline", "answer": "[{\"word\":\"tomb\",\"start\":5.6,\"end\":6.03}]"} {"utterance_id": "4831-25894-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0007.flac", "task_type": "lexical_detection", "query": "songs", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"songs\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.58, 8.96, "songs"]], "response_mode": "offline", "answer": "[{\"word\":\"songs\",\"start\":8.58,\"end\":8.96}]"} {"utterance_id": "8173-294714-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0041.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.46, 2.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.46,\"end\":2.72}]"} {"utterance_id": "7697-105815-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0049.flac", "task_type": "lexical_detection", "query": "nor", "prompt": "When the utterance is complete, report all instances of the word \"nor\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.04, "nor"]], "response_mode": "offline", "answer": "[{\"word\":\"nor\",\"start\":1.68,\"end\":2.04}]"} {"utterance_id": "1255-90407-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0011.flac", "task_type": "lexical_detection", "query": "doesn't", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"doesn't\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.59, "doesn't"]], "response_mode": "offline", "answer": "[{\"word\":\"doesn't\",\"start\":1.28,\"end\":1.59}]"} {"utterance_id": "8288-274162-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0025.flac", "task_type": "lexical_detection", "query": "curiosity", "prompt": "After the audio ends, extract all occurrences of the word \"curiosity\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.84, "curiosity"]], "response_mode": "offline", "answer": "[{\"word\":\"curiosity\",\"start\":1.16,\"end\":1.84}]"} {"utterance_id": "4515-11057-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0009.flac", "task_type": "lexical_detection", "query": "deuce", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"deuce\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.73, "deuce"]], "response_mode": "offline", "answer": "[{\"word\":\"deuce\",\"start\":1.36,\"end\":1.73}]"} {"utterance_id": "3660-172183-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.92, 2.28, "geraint"], [4.0, 4.3, "enid"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":1.92,\"end\":2.28},{\"word\":\"enid\",\"start\":4.0,\"end\":4.3}]"} {"utterance_id": "8288-274150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0008.flac", "task_type": "lexical_detection", "query": "knees", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"knees\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.17, 3.56, "knees"]], "response_mode": "offline", "answer": "[{\"word\":\"knees\",\"start\":3.17,\"end\":3.56}]"} {"utterance_id": "6267-53049-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.26, 0.78, "doris"]], "response_mode": "offline", "answer": "[{\"word\":\"doris\",\"start\":0.26,\"end\":0.78}]"} {"utterance_id": "1651-136854-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0024.flac", "task_type": "lexical_detection", "query": "pouring", "prompt": "Once the audio has finished, identify every instance of the word \"pouring\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.02, 5.44, "pouring"]], "response_mode": "offline", "answer": "[{\"word\":\"pouring\",\"start\":5.02,\"end\":5.44}]"} {"utterance_id": "1650-167613-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.96, 4.44, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":3.96,\"end\":4.44}]"} {"utterance_id": "1630-102884-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0005.flac", "task_type": "lexical_detection", "query": "builds", "prompt": "After the audio ends, extract all occurrences of the word \"builds\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.95, 6.4, "builds"], [9.04, 9.61, "builds"]], "response_mode": "offline", "answer": "[{\"word\":\"builds\",\"start\":5.95,\"end\":6.4},{\"word\":\"builds\",\"start\":9.04,\"end\":9.61}]"} {"utterance_id": "1686-142278-0076", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0076.flac", "task_type": "lexical_detection", "query": "strength", "prompt": "Process the full audio offline. Find every instance of the word \"strength\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.29, "strength"]], "response_mode": "offline", "answer": "[{\"word\":\"strength\",\"start\":1.84,\"end\":2.29}]"} {"utterance_id": "6467-94831-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.9, "sam"]], "response_mode": "offline", "answer": "[{\"word\":\"sam\",\"start\":0.53,\"end\":0.9}]"} {"utterance_id": "8173-294714-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0035.flac", "task_type": "lexical_detection", "query": "brandt", "prompt": "Process the full audio offline. Find every instance of the word \"brandt\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.21, 4.6, "brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"brandt\",\"start\":4.21,\"end\":4.6}]"} {"utterance_id": "6123-59186-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.42, 2.86, "browning"]], "response_mode": "offline", "answer": "[{\"word\":\"browning\",\"start\":2.42,\"end\":2.86}]"} {"utterance_id": "3660-172183-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0006.flac", "task_type": "lexical_detection", "query": "went", "prompt": "After the audio ends, extract all occurrences of the word \"went\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.3, "went"]], "response_mode": "offline", "answer": "[{\"word\":\"went\",\"start\":1.1,\"end\":1.3}]"} {"utterance_id": "6455-67803-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.11, 7.67, "mister binnie"]], "response_mode": "offline", "answer": "[{\"word\":\"mister binnie\",\"start\":7.11,\"end\":7.67}]"} {"utterance_id": "3915-98647-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0020.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 1.04, "body"], [5.22, 5.65, "chin"], [6.66, 7.15, "knees"]], "response_mode": "offline", "answer": "[{\"word\":\"body\",\"start\":0.56,\"end\":1.04},{\"word\":\"chin\",\"start\":5.22,\"end\":5.65},{\"word\":\"knees\",\"start\":6.66,\"end\":7.15}]"} {"utterance_id": "4572-112375-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0006.flac", "task_type": "lexical_counting", "query": "thus", "prompt": "After listening to the whole recording, output timed count words for the word \"thus\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[3.82, 4.21, "one"], [13.45, 13.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.82,\"end\":4.21},{\"word\":\"two\",\"start\":13.45,\"end\":13.72}]"} {"utterance_id": "116-288045-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0016.flac", "task_type": "lexical_detection", "query": "radiant", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"radiant\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.66, 6.08, "radiant"]], "response_mode": "offline", "answer": "[{\"word\":\"radiant\",\"start\":5.66,\"end\":6.08}]"} {"utterance_id": "1255-138279-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0016.flac", "task_type": "lexical_detection", "query": "tremulously", "prompt": "After listening to the whole recording, output every instance of the word \"tremulously\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.99, "tremulously"]], "response_mode": "offline", "answer": "[{\"word\":\"tremulously\",\"start\":2.22,\"end\":2.99}]"} {"utterance_id": "6467-94831-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0026.flac", "task_type": "lexical_detection", "query": "tiny", "prompt": "Once the audio has finished, identify every instance of the word \"tiny\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.08, 1.36, "tiny"]], "response_mode": "offline", "answer": "[{\"word\":\"tiny\",\"start\":1.08,\"end\":1.36}]"} {"utterance_id": "3915-57461-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0019.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.7, 2.12, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":1.7,\"end\":2.12}]"} {"utterance_id": "1701-141760-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0024.flac", "task_type": "lexical_detection", "query": "wine", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"wine\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.85, 2.26, "wine"]], "response_mode": "offline", "answer": "[{\"word\":\"wine\",\"start\":1.85,\"end\":2.26}]"} {"utterance_id": "8288-274162-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0013.flac", "task_type": "lexical_detection", "query": "ask", "prompt": "Process the full audio offline. Find every instance of the word \"ask\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.15, "ask"]], "response_mode": "offline", "answer": "[{\"word\":\"ask\",\"start\":2.98,\"end\":3.15}]"} {"utterance_id": "4515-11057-0108", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0108.flac", "task_type": "lexical_detection", "query": "recognised", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"recognised\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.85, 3.58, "recognised"]], "response_mode": "offline", "answer": "[{\"word\":\"recognised\",\"start\":2.85,\"end\":3.58}]"} {"utterance_id": "6455-67803-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.42, 7.96, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":7.42,\"end\":7.96}]"} {"utterance_id": "7641-96684-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 2.79, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":2.46,\"end\":2.79}]"} {"utterance_id": "3660-6517-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0008.flac", "task_type": "lexical_detection", "query": "entrance", "prompt": "Process the full audio offline. Find every instance of the word \"entrance\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.47, 4.78, "entrance"]], "response_mode": "offline", "answer": "[{\"word\":\"entrance\",\"start\":4.47,\"end\":4.78}]"} {"utterance_id": "4515-11057-0101", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0101.flac", "task_type": "lexical_detection", "query": "delayed", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"delayed\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.22, "delayed"]], "response_mode": "offline", "answer": "[{\"word\":\"delayed\",\"start\":0.77,\"end\":1.22}]"} {"utterance_id": "5543-27761-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0034.flac", "task_type": "lexical_detection", "query": "kingdom", "prompt": "After listening to the whole recording, output every instance of the word \"kingdom\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.72, 2.31, "kingdom"]], "response_mode": "offline", "answer": "[{\"word\":\"kingdom\",\"start\":1.72,\"end\":2.31}]"} {"utterance_id": "7601-101622-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0004.flac", "task_type": "lexical_detection", "query": "fathers", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fathers\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.86, 6.61, "fathers"]], "response_mode": "offline", "answer": "[{\"word\":\"fathers\",\"start\":5.86,\"end\":6.61}]"} {"utterance_id": "7697-105815-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.61, 4.91, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":4.61,\"end\":4.91}]"} {"utterance_id": "4572-112383-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0002.flac", "task_type": "lexical_detection", "query": "scorched", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"scorched\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.43, 1.18, "scorched"]], "response_mode": "offline", "answer": "[{\"word\":\"scorched\",\"start\":0.43,\"end\":1.18}]"} {"utterance_id": "6455-67803-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0008.flac", "task_type": "lexical_detection", "query": "attractive", "prompt": "After the audio ends, extract all occurrences of the word \"attractive\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.75, 4.39, "attractive"]], "response_mode": "offline", "answer": "[{\"word\":\"attractive\",\"start\":3.75,\"end\":4.39}]"} {"utterance_id": "4831-18525-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0003.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.9, 2.25, "one"], [5.12, 5.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.9,\"end\":2.25},{\"word\":\"two\",\"start\":5.12,\"end\":5.56}]"} {"utterance_id": "8288-274150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 3.11, "malicorne s"], [6.18, 6.91, "saint aignan"]], "response_mode": "offline", "answer": "[{\"word\":\"malicorne s\",\"start\":2.52,\"end\":3.11},{\"word\":\"saint aignan\",\"start\":6.18,\"end\":6.91}]"} {"utterance_id": "6841-88291-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0051.flac", "task_type": "lexical_detection", "query": "seventy", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"seventy\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.47, "seventy"]], "response_mode": "offline", "answer": "[{\"word\":\"seventy\",\"start\":1.02,\"end\":1.47}]"} {"utterance_id": "3915-98647-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 5.37, "queen althea"]], "response_mode": "offline", "answer": "[{\"word\":\"queen althea\",\"start\":4.52,\"end\":5.37}]"} {"utterance_id": "3660-172183-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0005.flac", "task_type": "lexical_counting", "query": "thou", "prompt": "After the audio ends, tally every instance of the word \"thou\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[2.48, 2.67, "one"], [3.79, 3.98, "two"], [7.28, 7.5, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.48,\"end\":2.67},{\"word\":\"two\",\"start\":3.79,\"end\":3.98},{\"word\":\"three\",\"start\":7.28,\"end\":7.5}]"} {"utterance_id": "5543-27761-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0037.flac", "task_type": "lexical_counting", "query": "does", "prompt": "Once the audio has finished, list each occurrence of the word \"does\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.39, 1.64, "one"], [4.36, 4.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.39,\"end\":1.64},{\"word\":\"two\",\"start\":4.36,\"end\":4.66}]"} {"utterance_id": "7697-245712-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0019.flac", "task_type": "lexical_detection", "query": "directing", "prompt": "After listening to the whole recording, output every instance of the word \"directing\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.5, 8.01, "directing"]], "response_mode": "offline", "answer": "[{\"word\":\"directing\",\"start\":7.5,\"end\":8.01}]"} {"utterance_id": "4323-18416-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.28, 8.53, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":8.28,\"end\":8.53}]"} {"utterance_id": "1701-141760-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0052.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.43, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":0.99,\"end\":1.43}]"} {"utterance_id": "7641-96670-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0006.flac", "task_type": "lexical_detection", "query": "heroes", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"heroes\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.85, 2.28, "heroes"]], "response_mode": "offline", "answer": "[{\"word\":\"heroes\",\"start\":1.85,\"end\":2.28}]"} {"utterance_id": "700-122867-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0039.flac", "task_type": "lexical_detection", "query": "sounds", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sounds\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.15, 2.44, "sounds"]], "response_mode": "offline", "answer": "[{\"word\":\"sounds\",\"start\":2.15,\"end\":2.44}]"} {"utterance_id": "4323-55228-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.77, 3.04, "pete"]], "response_mode": "offline", "answer": "[{\"word\":\"pete\",\"start\":2.77,\"end\":3.04}]"} {"utterance_id": "5543-27761-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0024.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.66, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.37,\"end\":0.66}]"} {"utterance_id": "8173-294714-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0010.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.23, 3.57, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":3.23,\"end\":3.57}]"} {"utterance_id": "4831-29134-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0003.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a date with fields word, start, end (seconds). If none, reply [].", "labels": [[0.91, 1.25, "sunday"], [6.79, 7.15, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":0.91,\"end\":1.25},{\"word\":\"sunday\",\"start\":6.79,\"end\":7.15}]"} {"utterance_id": "116-288047-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0018.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[4.24, 4.74, "mind"]], "response_mode": "offline", "answer": "[{\"word\":\"mind\",\"start\":4.24,\"end\":4.74}]"} {"utterance_id": "6455-66379-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0010.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, list every instance of a plant name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.3, "wood"]], "response_mode": "offline", "answer": "[{\"word\":\"wood\",\"start\":3.14,\"end\":3.3}]"} {"utterance_id": "4572-64670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.6, 5.93, "rama"]], "response_mode": "offline", "answer": "[{\"word\":\"rama\",\"start\":5.6,\"end\":5.93}]"} {"utterance_id": "4323-55228-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0003.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.45, "one"], [3.93, 4.22, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.24,\"end\":2.45},{\"word\":\"two\",\"start\":3.93,\"end\":4.22}]"} {"utterance_id": "8173-294714-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0008.flac", "task_type": "lexical_detection", "query": "shame", "prompt": "Once the audio has finished, identify every instance of the word \"shame\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.0, 1.45, "shame"]], "response_mode": "offline", "answer": "[{\"word\":\"shame\",\"start\":1.0,\"end\":1.45}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[4.42, 4.92, "one"], [6.37, 6.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.42,\"end\":4.92},{\"word\":\"two\",\"start\":6.37,\"end\":6.7}]"} {"utterance_id": "8254-84205-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0046.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.6, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.19,\"end\":1.6}]"} {"utterance_id": "4831-18525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0007.flac", "task_type": "lexical_detection", "query": "ring", "prompt": "When the utterance is complete, report all instances of the word \"ring\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.02, 6.23, "ring"]], "response_mode": "offline", "answer": "[{\"word\":\"ring\",\"start\":6.02,\"end\":6.23}]"} {"utterance_id": "4153-61735-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0012.flac", "task_type": "lexical_detection", "query": "contained", "prompt": "Once the audio has finished, identify every instance of the word \"contained\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[10.14, 10.7, "contained"]], "response_mode": "offline", "answer": "[{\"word\":\"contained\",\"start\":10.14,\"end\":10.7}]"} {"utterance_id": "116-288048-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0000.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "When the utterance is complete, report all instances of an organization name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 3.09, "christian church"]], "response_mode": "offline", "answer": "[{\"word\":\"christian church\",\"start\":2.27,\"end\":3.09}]"} {"utterance_id": "8173-294714-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0040.flac", "task_type": "lexical_detection", "query": "necessity", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"necessity\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.83, "necessity"]], "response_mode": "offline", "answer": "[{\"word\":\"necessity\",\"start\":1.05,\"end\":1.83}]"} {"utterance_id": "7697-105815-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0020.flac", "task_type": "lexical_detection", "query": "foot", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"foot\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.16, 3.33, "foot"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":3.16,\"end\":3.33}]"} {"utterance_id": "1630-96099-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0015.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 2.03, "eleventh"]], "response_mode": "offline", "answer": "[{\"word\":\"eleventh\",\"start\":1.56,\"end\":2.03}]"} {"utterance_id": "4153-186222-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0011.flac", "task_type": "lexical_detection", "query": "provided", "prompt": "Once the audio has finished, identify every instance of the word \"provided\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.96, 2.45, "provided"]], "response_mode": "offline", "answer": "[{\"word\":\"provided\",\"start\":1.96,\"end\":2.45}]"} {"utterance_id": "7697-245712-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0020.flac", "task_type": "lexical_detection", "query": "reasons", "prompt": "When the utterance is complete, report all instances of the word \"reasons\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.85, 5.4, "reasons"]], "response_mode": "offline", "answer": "[{\"word\":\"reasons\",\"start\":4.85,\"end\":5.4}]"} {"utterance_id": "4572-112383-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0000.flac", "task_type": "lexical_detection", "query": "ceased", "prompt": "When the utterance is complete, report all instances of the word \"ceased\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.39, "ceased"]], "response_mode": "offline", "answer": "[{\"word\":\"ceased\",\"start\":0.76,\"end\":1.39}]"} {"utterance_id": "6841-88291-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0023.flac", "task_type": "lexical_detection", "query": "scorched", "prompt": "After the audio ends, extract all occurrences of the word \"scorched\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.69, "scorched"]], "response_mode": "offline", "answer": "[{\"word\":\"scorched\",\"start\":2.1,\"end\":2.69}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, list every instance of a plant name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[22.24, 22.66, "rose"]], "response_mode": "offline", "answer": "[{\"word\":\"rose\",\"start\":22.24,\"end\":22.66}]"} {"utterance_id": "6841-88291-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0055.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a time of day with fields word, start, end (seconds). If none, reply [].", "labels": [[2.19, 2.65, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":2.19,\"end\":2.65}]"} {"utterance_id": "1686-142278-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 2.87, "mister hale's"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale's\",\"start\":2.3,\"end\":2.87}]"} {"utterance_id": "700-122868-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0005.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "After the audio ends, tally every instance of an animal name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.55, 1.96, "one"], [7.55, 7.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.55,\"end\":1.96},{\"word\":\"two\",\"start\":7.55,\"end\":7.78}]"} {"utterance_id": "4323-55228-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0048.flac", "task_type": "lexical_detection", "query": "it'll", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"it'll\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.58, 0.77, "it'll"]], "response_mode": "offline", "answer": "[{\"word\":\"it'll\",\"start\":0.58,\"end\":0.77}]"} {"utterance_id": "6467-94831-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0040.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.28, "donkey"]], "response_mode": "offline", "answer": "[{\"word\":\"donkey\",\"start\":0.7,\"end\":1.28}]"} {"utterance_id": "2506-11267-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0000.flac", "task_type": "lexical_detection", "query": "along", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"along\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.31, 6.56, "along"]], "response_mode": "offline", "answer": "[{\"word\":\"along\",\"start\":6.31,\"end\":6.56}]"} {"utterance_id": "6455-66379-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0002.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.5, "supper"]], "response_mode": "offline", "answer": "[{\"word\":\"supper\",\"start\":0.21,\"end\":0.5}]"} {"utterance_id": "116-288048-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 1.01, "peter"]], "response_mode": "offline", "answer": "[{\"word\":\"peter\",\"start\":0.68,\"end\":1.01}]"} {"utterance_id": "7697-105815-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0046.flac", "task_type": "lexical_detection", "query": "dwell", "prompt": "Process the full audio offline. Find every instance of the word \"dwell\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.8, 4.12, "dwell"]], "response_mode": "offline", "answer": "[{\"word\":\"dwell\",\"start\":3.8,\"end\":4.12}]"} {"utterance_id": "4323-55228-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0016.flac", "task_type": "lexical_detection", "query": "thankful", "prompt": "Process the full audio offline. Find every instance of the word \"thankful\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.65, 5.03, "thankful"]], "response_mode": "offline", "answer": "[{\"word\":\"thankful\",\"start\":4.65,\"end\":5.03}]"} {"utterance_id": "1650-157641-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.53, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":1.1,\"end\":1.53}]"} {"utterance_id": "116-288047-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0016.flac", "task_type": "lexical_detection", "query": "science", "prompt": "After the audio ends, extract all occurrences of the word \"science\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 2.19, "science"]], "response_mode": "offline", "answer": "[{\"word\":\"science\",\"start\":1.5,\"end\":2.19}]"} {"utterance_id": "4153-186223-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.56, 7.02, "merrick"]], "response_mode": "offline", "answer": "[{\"word\":\"merrick\",\"start\":6.56,\"end\":7.02}]"} {"utterance_id": "5543-27761-0087", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0087.flac", "task_type": "lexical_detection", "query": "met", "prompt": "After listening to the whole recording, output every instance of the word \"met\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 0.89, "met"]], "response_mode": "offline", "answer": "[{\"word\":\"met\",\"start\":0.67,\"end\":0.89}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.39, 4.8, "feed"]], "response_mode": "offline", "answer": "[{\"word\":\"feed\",\"start\":4.39,\"end\":4.8}]"} {"utterance_id": "1255-138279-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.14, "frank"]], "response_mode": "offline", "answer": "[{\"word\":\"frank\",\"start\":0.86,\"end\":1.14}]"} {"utterance_id": "4515-11057-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0065.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 2.9, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.48,\"end\":2.9}]"} {"utterance_id": "4323-55228-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0024.flac", "task_type": "lexical_detection", "query": "shifted", "prompt": "Once the audio has finished, identify every instance of the word \"shifted\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.7, 2.03, "shifted"]], "response_mode": "offline", "answer": "[{\"word\":\"shifted\",\"start\":1.7,\"end\":2.03}]"} {"utterance_id": "8288-274162-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0015.flac", "task_type": "lexical_detection", "query": "unhappy", "prompt": "Process the full audio offline. Find every instance of the word \"unhappy\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 1.0, "unhappy"]], "response_mode": "offline", "answer": "[{\"word\":\"unhappy\",\"start\":0.56,\"end\":1.0}]"} {"utterance_id": "3660-172182-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.51, 3.88, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":3.51,\"end\":3.88}]"} {"utterance_id": "6841-88291-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0015.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[1.57, 1.97, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.57,\"end\":1.97}]"} {"utterance_id": "3663-172528-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0053.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "When the utterance is complete, report how a number accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 2.88, "one"], [7.04, 7.68, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.61,\"end\":2.88},{\"word\":\"two\",\"start\":7.04,\"end\":7.68}]"} {"utterance_id": "4153-186222-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.39, "beth"]], "response_mode": "offline", "answer": "[{\"word\":\"beth\",\"start\":1.09,\"end\":1.39}]"} {"utterance_id": "6841-88294-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0038.flac", "task_type": "lexical_detection", "query": "flat", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"flat\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.77, 1.07, "flat"]], "response_mode": "offline", "answer": "[{\"word\":\"flat\",\"start\":0.77,\"end\":1.07}]"} {"utterance_id": "700-122868-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 0.99, "elaine"]], "response_mode": "offline", "answer": "[{\"word\":\"elaine\",\"start\":0.76,\"end\":0.99}]"} {"utterance_id": "8173-294714-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0022.flac", "task_type": "lexical_detection", "query": "creature", "prompt": "Once the audio has finished, identify every instance of the word \"creature\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.19, 1.58, "creature"]], "response_mode": "offline", "answer": "[{\"word\":\"creature\",\"start\":1.19,\"end\":1.58}]"} {"utterance_id": "7697-105815-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0006.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.75, 6.56, "interpreter"]], "response_mode": "offline", "answer": "[{\"word\":\"interpreter\",\"start\":5.75,\"end\":6.56}]"} {"utterance_id": "4515-11057-0084", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0084.flac", "task_type": "lexical_detection", "query": "beneath", "prompt": "After listening to the whole recording, output every instance of the word \"beneath\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.07, 8.45, "beneath"]], "response_mode": "offline", "answer": "[{\"word\":\"beneath\",\"start\":8.07,\"end\":8.45}]"} {"utterance_id": "8254-115543-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0035.flac", "task_type": "lexical_detection", "query": "haughty", "prompt": "When the utterance is complete, report all instances of the word \"haughty\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.82, "haughty"]], "response_mode": "offline", "answer": "[{\"word\":\"haughty\",\"start\":1.47,\"end\":1.82}]"} {"utterance_id": "1630-96099-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0011.flac", "task_type": "lexical_detection", "query": "stuffed", "prompt": "After the audio ends, extract all occurrences of the word \"stuffed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.92, 4.32, "stuffed"]], "response_mode": "offline", "answer": "[{\"word\":\"stuffed\",\"start\":3.92,\"end\":4.32}]"} {"utterance_id": "6267-65525-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0036.flac", "task_type": "lexical_detection", "query": "mistake", "prompt": "After the audio ends, extract all occurrences of the word \"mistake\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.71, "mistake"]], "response_mode": "offline", "answer": "[{\"word\":\"mistake\",\"start\":2.0,\"end\":2.71}]"} {"utterance_id": "6841-88291-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0037.flac", "task_type": "lexical_detection", "query": "happened", "prompt": "After the audio ends, extract all occurrences of the word \"happened\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.51, "happened"]], "response_mode": "offline", "answer": "[{\"word\":\"happened\",\"start\":1.15,\"end\":1.51}]"} {"utterance_id": "7697-105815-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.98, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.46,\"end\":0.98}]"} {"utterance_id": "6841-88294-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0023.flac", "task_type": "lexical_detection", "query": "begged", "prompt": "After the audio ends, extract all occurrences of the word \"begged\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.41, 2.81, "begged"]], "response_mode": "offline", "answer": "[{\"word\":\"begged\",\"start\":2.41,\"end\":2.81}]"} {"utterance_id": "116-288048-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0014.flac", "task_type": "lexical_detection", "query": "peter", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"peter\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 1.01, "peter"]], "response_mode": "offline", "answer": "[{\"word\":\"peter\",\"start\":0.68,\"end\":1.01}]"} {"utterance_id": "4515-11057-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0029.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.28, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":1.88,\"end\":2.28}]"} {"utterance_id": "6123-59186-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0018.flac", "task_type": "lexical_detection", "query": "beauty", "prompt": "After the audio ends, extract all occurrences of the word \"beauty\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.48, 6.92, "beauty"]], "response_mode": "offline", "answer": "[{\"word\":\"beauty\",\"start\":6.48,\"end\":6.92}]"} {"utterance_id": "6841-88291-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0008.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.29, 4.68, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":4.29,\"end\":4.68}]"} {"utterance_id": "116-288045-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0029.flac", "task_type": "lexical_detection", "query": "breath", "prompt": "When the utterance is complete, report all instances of the word \"breath\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.7, 8.23, "breath"]], "response_mode": "offline", "answer": "[{\"word\":\"breath\",\"start\":7.7,\"end\":8.23}]"} {"utterance_id": "7697-105815-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0047.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[0.44, 0.89, "one"], [3.34, 3.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.44,\"end\":0.89},{\"word\":\"two\",\"start\":3.34,\"end\":3.92}]"} {"utterance_id": "4831-25894-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0012.flac", "task_type": "lexical_detection", "query": "laugh", "prompt": "When the utterance is complete, report all instances of the word \"laugh\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.68, 5.07, "laugh"]], "response_mode": "offline", "answer": "[{\"word\":\"laugh\",\"start\":4.68,\"end\":5.07}]"} {"utterance_id": "6123-59150-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0034.flac", "task_type": "lexical_detection", "query": "measure", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"measure\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.24, 1.68, "measure"]], "response_mode": "offline", "answer": "[{\"word\":\"measure\",\"start\":1.24,\"end\":1.68}]"} {"utterance_id": "4831-25894-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0023.flac", "task_type": "lexical_detection", "query": "sang", "prompt": "When the utterance is complete, report all instances of the word \"sang\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.55, "sang"]], "response_mode": "offline", "answer": "[{\"word\":\"sang\",\"start\":3.33,\"end\":3.55}]"} {"utterance_id": "1650-173551-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0003.flac", "task_type": "lexical_detection", "query": "landing", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"landing\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.23, "landing"]], "response_mode": "offline", "answer": "[{\"word\":\"landing\",\"start\":0.84,\"end\":1.23}]"} {"utterance_id": "4572-112381-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.88, 11.44, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":10.88,\"end\":11.44}]"} {"utterance_id": "6467-94831-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.68, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":2.35,\"end\":2.68}]"} {"utterance_id": "7697-105817-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0003.flac", "task_type": "lexical_detection", "query": "lost", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"lost\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.38, 1.64, "lost"]], "response_mode": "offline", "answer": "[{\"word\":\"lost\",\"start\":1.38,\"end\":1.64}]"} {"utterance_id": "4831-29134-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[3.07, 3.53, "senator"]], "response_mode": "offline", "answer": "[{\"word\":\"senator\",\"start\":3.07,\"end\":3.53}]"} {"utterance_id": "1651-136854-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0023.flac", "task_type": "lexical_detection", "query": "cases", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"cases\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.68, "cases"]], "response_mode": "offline", "answer": "[{\"word\":\"cases\",\"start\":2.14,\"end\":2.68}]"} {"utterance_id": "6841-88294-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0017.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Process the full audio offline. Find every instance of a color name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.48, "black"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":2.13,\"end\":2.48}]"} {"utterance_id": "116-288047-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0007.flac", "task_type": "lexical_detection", "query": "fiction", "prompt": "After listening to the whole recording, output every instance of the word \"fiction\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.66, 2.19, "fiction"]], "response_mode": "offline", "answer": "[{\"word\":\"fiction\",\"start\":1.66,\"end\":2.19}]"} {"utterance_id": "8288-274162-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0018.flac", "task_type": "lexical_detection", "query": "astonished", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"astonished\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 2.06, "astonished"]], "response_mode": "offline", "answer": "[{\"word\":\"astonished\",\"start\":1.49,\"end\":2.06}]"} {"utterance_id": "4323-55228-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.89, 6.1, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":5.89,\"end\":6.1}]"} {"utterance_id": "1686-142278-0082", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0082.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After listening to the whole recording, output every instance of a country name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.48, 5.95, "england"]], "response_mode": "offline", "answer": "[{\"word\":\"england\",\"start\":5.48,\"end\":5.95}]"} {"utterance_id": "5849-50873-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0021.flac", "task_type": "lexical_detection", "query": "wadsworth", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wadsworth\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.49, "wadsworth"]], "response_mode": "offline", "answer": "[{\"word\":\"wadsworth\",\"start\":3.94,\"end\":4.49}]"} {"utterance_id": "4831-25894-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.73, 3.12, "tommo"], [5.75, 6.16, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":2.73,\"end\":3.12},{\"word\":\"tessa\",\"start\":5.75,\"end\":6.16}]"} {"utterance_id": "1686-142278-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0034.flac", "task_type": "lexical_detection", "query": "resolution", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"resolution\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.54, 5.17, "resolution"]], "response_mode": "offline", "answer": "[{\"word\":\"resolution\",\"start\":4.54,\"end\":5.17}]"} {"utterance_id": "1686-142278-0084", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0084.flac", "task_type": "lexical_detection", "query": "exactly", "prompt": "When the utterance is complete, report all instances of the word \"exactly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.88, "exactly"]], "response_mode": "offline", "answer": "[{\"word\":\"exactly\",\"start\":1.14,\"end\":1.88}]"} {"utterance_id": "4323-55228-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0026.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.48, 2.96, "eyelids"]], "response_mode": "offline", "answer": "[{\"word\":\"eyelids\",\"start\":2.48,\"end\":2.96}]"} {"utterance_id": "1255-74899-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0012.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an ordinal number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.66, "foremost"]], "response_mode": "offline", "answer": "[{\"word\":\"foremost\",\"start\":1.18,\"end\":1.66}]"} {"utterance_id": "5849-50873-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.6, 2.48, "charles stevens"]], "response_mode": "offline", "answer": "[{\"word\":\"charles stevens\",\"start\":1.6,\"end\":2.48}]"} {"utterance_id": "4831-18525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0021.flac", "task_type": "lexical_detection", "query": "tear", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"tear\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.4, 3.64, "tear"]], "response_mode": "offline", "answer": "[{\"word\":\"tear\",\"start\":3.4,\"end\":3.64}]"} {"utterance_id": "6467-62797-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0005.flac", "task_type": "lexical_detection", "query": "breast", "prompt": "Process the full audio offline. Find every instance of the word \"breast\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.41, "breast"]], "response_mode": "offline", "answer": "[{\"word\":\"breast\",\"start\":1.95,\"end\":2.41}]"} {"utterance_id": "1630-96099-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0023.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.21, 8.79, "devil"]], "response_mode": "offline", "answer": "[{\"word\":\"devil\",\"start\":8.21,\"end\":8.79}]"} {"utterance_id": "6267-65525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0021.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[10.63, 11.22, "plainfield"]], "response_mode": "offline", "answer": "[{\"word\":\"plainfield\",\"start\":10.63,\"end\":11.22}]"} {"utterance_id": "2506-11278-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0015.flac", "task_type": "lexical_detection", "query": "landlady", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"landlady\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.58, 6.3, "landlady"]], "response_mode": "offline", "answer": "[{\"word\":\"landlady\",\"start\":5.58,\"end\":6.3}]"} {"utterance_id": "2506-13150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0008.flac", "task_type": "lexical_detection", "query": "ten", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ten\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.14, 2.42, "ten"]], "response_mode": "offline", "answer": "[{\"word\":\"ten\",\"start\":2.14,\"end\":2.42}]"} {"utterance_id": "4831-25894-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0027.flac", "task_type": "lexical_detection", "query": "brothers", "prompt": "When the utterance is complete, report all instances of the word \"brothers\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.37, "brothers"]], "response_mode": "offline", "answer": "[{\"word\":\"brothers\",\"start\":0.98,\"end\":1.37}]"} {"utterance_id": "7697-105815-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0039.flac", "task_type": "lexical_counting", "query": "went", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"went\" (word, start, end). If none, reply [].", "labels": [[2.59, 2.86, "one"], [9.87, 10.1, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.59,\"end\":2.86},{\"word\":\"two\",\"start\":9.87,\"end\":10.1}]"} {"utterance_id": "8254-84205-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0025.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.13, "one"], [3.25, 3.57, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.75,\"end\":2.13},{\"word\":\"two\",\"start\":3.25,\"end\":3.57}]"} {"utterance_id": "1686-142278-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0024.flac", "task_type": "lexical_detection", "query": "slightest", "prompt": "After listening to the whole recording, output every instance of the word \"slightest\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.83, 3.38, "slightest"]], "response_mode": "offline", "answer": "[{\"word\":\"slightest\",\"start\":2.83,\"end\":3.38}]"} {"utterance_id": "1701-141759-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0032.flac", "task_type": "lexical_detection", "query": "reading", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"reading\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.35, "reading"]], "response_mode": "offline", "answer": "[{\"word\":\"reading\",\"start\":3.05,\"end\":3.35}]"} {"utterance_id": "2506-169427-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0005.flac", "task_type": "lexical_detection", "query": "yorick", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"yorick\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.5, 0.95, "yorick"]], "response_mode": "offline", "answer": "[{\"word\":\"yorick\",\"start\":0.5,\"end\":0.95}]"} {"utterance_id": "1585-157660-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0016.flac", "task_type": "lexical_detection", "query": "james", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"james\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.8, "james"]], "response_mode": "offline", "answer": "[{\"word\":\"james\",\"start\":0.42,\"end\":0.8}]"} {"utterance_id": "6123-59186-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0030.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.39, "wives"]], "response_mode": "offline", "answer": "[{\"word\":\"wives\",\"start\":0.81,\"end\":1.39}]"} {"utterance_id": "1701-141759-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0004.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 3.96, "mamma"]], "response_mode": "offline", "answer": "[{\"word\":\"mamma\",\"start\":3.56,\"end\":3.96}]"} {"utterance_id": "7641-96252-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0022.flac", "task_type": "lexical_detection", "query": "weeps", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"weeps\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.43, 1.82, "weeps"], [3.64, 3.95, "weeps"]], "response_mode": "offline", "answer": "[{\"word\":\"weeps\",\"start\":1.43,\"end\":1.82},{\"word\":\"weeps\",\"start\":3.64,\"end\":3.95}]"} {"utterance_id": "3660-172182-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.71, 1.32, "guenever"]], "response_mode": "offline", "answer": "[{\"word\":\"guenever\",\"start\":0.71,\"end\":1.32}]"} {"utterance_id": "4515-11057-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0062.flac", "task_type": "lexical_detection", "query": "listened", "prompt": "When the utterance is complete, report all instances of the word \"listened\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.19, "listened"]], "response_mode": "offline", "answer": "[{\"word\":\"listened\",\"start\":0.78,\"end\":1.19}]"} {"utterance_id": "2506-169427-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.91, "eugenius"]], "response_mode": "offline", "answer": "[{\"word\":\"eugenius\",\"start\":0.37,\"end\":0.91}]"} {"utterance_id": "3660-172183-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0020.flac", "task_type": "lexical_detection", "query": "whence", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"whence\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.55, 3.9, "whence"]], "response_mode": "offline", "answer": "[{\"word\":\"whence\",\"start\":3.55,\"end\":3.9}]"} {"utterance_id": "1630-141772-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.88, "prince andrew"], [7.02, 8.1, "princess mary"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":2.02,\"end\":2.88},{\"word\":\"princess mary\",\"start\":7.02,\"end\":8.1}]"} {"utterance_id": "4831-29134-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0010.flac", "task_type": "lexical_counting", "query": "school", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"school\" with word/start/end. If none, reply [].", "labels": [[5.47, 5.89, "one"], [7.45, 7.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.47,\"end\":5.89},{\"word\":\"two\",\"start\":7.45,\"end\":7.87}]"} {"utterance_id": "4323-18416-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0017.flac", "task_type": "lexical_detection", "query": "repeated", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"repeated\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 1.96, "repeated"]], "response_mode": "offline", "answer": "[{\"word\":\"repeated\",\"start\":1.57,\"end\":1.96}]"} {"utterance_id": "8254-115543-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0023.flac", "task_type": "lexical_counting", "query": "big", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"big\" (word, start, end). If none, reply [].", "labels": [[3.65, 3.85, "one"], [9.63, 9.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.65,\"end\":3.85},{\"word\":\"two\",\"start\":9.63,\"end\":9.87}]"} {"utterance_id": "8288-274162-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0000.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the full audio. Count every occurrence of a body part. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.79, 6.27, "one"], [8.65, 9.35, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.79,\"end\":6.27},{\"word\":\"two\",\"start\":8.65,\"end\":9.35}]"} {"utterance_id": "4153-185072-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0011.flac", "task_type": "lexical_detection", "query": "sufficient", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"sufficient\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.65, 5.18, "sufficient"]], "response_mode": "offline", "answer": "[{\"word\":\"sufficient\",\"start\":4.65,\"end\":5.18}]"} {"utterance_id": "7641-96670-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0019.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the full audio. After it finishes, find every instance of a color name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.16, "dark"], [3.38, 3.64, "golden"]], "response_mode": "offline", "answer": "[{\"word\":\"dark\",\"start\":0.92,\"end\":1.16},{\"word\":\"golden\",\"start\":3.38,\"end\":3.64}]"} {"utterance_id": "1255-74899-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0019.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.92, 3.3, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":2.92,\"end\":3.3}]"} {"utterance_id": "4153-186222-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0007.flac", "task_type": "lexical_detection", "query": "mean", "prompt": "After the audio ends, extract all occurrences of the word \"mean\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.88, "mean"]], "response_mode": "offline", "answer": "[{\"word\":\"mean\",\"start\":0.68,\"end\":0.88}]"} {"utterance_id": "6467-97061-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0023.flac", "task_type": "lexical_detection", "query": "enemy", "prompt": "After the audio ends, extract all occurrences of the word \"enemy\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.75, "enemy"]], "response_mode": "offline", "answer": "[{\"word\":\"enemy\",\"start\":3.39,\"end\":3.75}]"} {"utterance_id": "7641-96684-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0019.flac", "task_type": "lexical_detection", "query": "footsteps", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"footsteps\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.89, 1.67, "footsteps"]], "response_mode": "offline", "answer": "[{\"word\":\"footsteps\",\"start\":0.89,\"end\":1.67}]"} {"utterance_id": "8173-294714-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.65, 2.28, "van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"van brandt\",\"start\":1.65,\"end\":2.28}]"} {"utterance_id": "8288-274162-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0029.flac", "task_type": "lexical_detection", "query": "brain", "prompt": "Process the full audio offline. Find every instance of the word \"brain\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.01, 3.49, "brain"]], "response_mode": "offline", "answer": "[{\"word\":\"brain\",\"start\":3.01,\"end\":3.49}]"} {"utterance_id": "4831-25894-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0024.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.02, "mamma"]], "response_mode": "offline", "answer": "[{\"word\":\"mamma\",\"start\":0.7,\"end\":1.02}]"} {"utterance_id": "4831-25894-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0011.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.66, 4.94, "brown"]], "response_mode": "offline", "answer": "[{\"word\":\"brown\",\"start\":4.66,\"end\":4.94}]"} {"utterance_id": "4153-186223-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0013.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.3, "mare"]], "response_mode": "offline", "answer": "[{\"word\":\"mare\",\"start\":2.01,\"end\":2.3}]"} {"utterance_id": "2506-11278-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0028.flac", "task_type": "lexical_detection", "query": "silly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"silly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.96, "silly"]], "response_mode": "offline", "answer": "[{\"word\":\"silly\",\"start\":1.59,\"end\":1.96}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Once the audio has finished, list each occurrence of a number as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.29, 2.95, "one"], [5.12, 5.47, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.29,\"end\":2.95},{\"word\":\"two\",\"start\":5.12,\"end\":5.47}]"} {"utterance_id": "4515-11057-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.65, "mister yule"]], "response_mode": "offline", "answer": "[{\"word\":\"mister yule\",\"start\":0.95,\"end\":1.65}]"} {"utterance_id": "4323-13259-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0025.flac", "task_type": "lexical_detection", "query": "witnessed", "prompt": "After the audio ends, extract all occurrences of the word \"witnessed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.34, 4.89, "witnessed"]], "response_mode": "offline", "answer": "[{\"word\":\"witnessed\",\"start\":4.34,\"end\":4.89}]"} {"utterance_id": "3660-172183-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0017.flac", "task_type": "lexical_detection", "query": "horses", "prompt": "After the audio ends, extract all occurrences of the word \"horses\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.44, 5.84, "horses"]], "response_mode": "offline", "answer": "[{\"word\":\"horses\",\"start\":5.44,\"end\":5.84}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a country name with fields word, start, end (seconds). If none, reply [].", "labels": [[14.5, 15.11, "american"]], "response_mode": "offline", "answer": "[{\"word\":\"american\",\"start\":14.5,\"end\":15.11}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.88, "one"], [6.11, 6.79, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.21,\"end\":0.88},{\"word\":\"two\",\"start\":6.11,\"end\":6.79}]"} {"utterance_id": "4323-55228-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.86, "bertram's"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram's\",\"start\":1.47,\"end\":1.86}]"} {"utterance_id": "4515-11057-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0037.flac", "task_type": "lexical_detection", "query": "pleaded", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pleaded\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.5, 4.05, "pleaded"]], "response_mode": "offline", "answer": "[{\"word\":\"pleaded\",\"start\":3.5,\"end\":4.05}]"} {"utterance_id": "1650-173552-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0001.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.67, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.22,\"end\":1.67}]"} {"utterance_id": "2506-11267-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0018.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an ordinal number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.95, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":0.66,\"end\":0.95}]"} {"utterance_id": "7641-96684-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0037.flac", "task_type": "lexical_detection", "query": "pangs", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"pangs\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.25, 3.66, "pangs"]], "response_mode": "offline", "answer": "[{\"word\":\"pangs\",\"start\":3.25,\"end\":3.66}]"} {"utterance_id": "116-288045-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0009.flac", "task_type": "lexical_detection", "query": "low", "prompt": "Process the full audio offline. Find every instance of the word \"low\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.24, "low"]], "response_mode": "offline", "answer": "[{\"word\":\"low\",\"start\":1.99,\"end\":2.24}]"} {"utterance_id": "1585-131718-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0034.flac", "task_type": "lexical_detection", "query": "charge", "prompt": "After the audio ends, extract all occurrences of the word \"charge\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.47, 3.9, "charge"]], "response_mode": "offline", "answer": "[{\"word\":\"charge\",\"start\":3.47,\"end\":3.9}]"} {"utterance_id": "4831-29134-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0001.flac", "task_type": "semantic_detection", "query": "country", "prompt": "You will hear a complete utterance. When it ends, list every instance of a country name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 3.07, "united states"]], "response_mode": "offline", "answer": "[{\"word\":\"united states\",\"start\":2.3,\"end\":3.07}]"} {"utterance_id": "3660-172182-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0023.flac", "task_type": "lexical_detection", "query": "welcome", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"welcome\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.02, "welcome"]], "response_mode": "offline", "answer": "[{\"word\":\"welcome\",\"start\":2.7,\"end\":3.02}]"} {"utterance_id": "4831-25894-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.37, 4.88, "fingers"]], "response_mode": "offline", "answer": "[{\"word\":\"fingers\",\"start\":4.37,\"end\":4.88}]"} {"utterance_id": "3663-172528-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0028.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen to the full audio. After it finishes, find every instance of a disease name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.87, 13.57, "illness"]], "response_mode": "offline", "answer": "[{\"word\":\"illness\",\"start\":12.87,\"end\":13.57}]"} {"utterance_id": "6841-88291-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0055.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[1.0, 1.47, "wrestler's"]], "response_mode": "offline", "answer": "[{\"word\":\"wrestler's\",\"start\":1.0,\"end\":1.47}]"} {"utterance_id": "4153-61735-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0004.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.73, 5.86, "casa del corvo"]], "response_mode": "offline", "answer": "[{\"word\":\"casa del corvo\",\"start\":4.73,\"end\":5.86}]"} {"utterance_id": "1650-167613-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0024.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.05, 2.99, "austerlitz"]], "response_mode": "offline", "answer": "[{\"word\":\"austerlitz\",\"start\":2.05,\"end\":2.99}]"} {"utterance_id": "1686-142278-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0061.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.63, 1.79, "one"], [4.01, 4.38, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.63,\"end\":1.79},{\"word\":\"two\",\"start\":4.01,\"end\":4.38}]"} {"utterance_id": "1585-157660-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 1.01, "devil's"]], "response_mode": "offline", "answer": "[{\"word\":\"devil's\",\"start\":0.62,\"end\":1.01}]"} {"utterance_id": "6841-88291-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0049.flac", "task_type": "lexical_detection", "query": "finally", "prompt": "When the utterance is complete, report all instances of the word \"finally\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.25, 0.77, "finally"]], "response_mode": "offline", "answer": "[{\"word\":\"finally\",\"start\":0.25,\"end\":0.77}]"} {"utterance_id": "4515-11057-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0021.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 1.96, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.43,\"end\":1.96}]"} {"utterance_id": "4323-13259-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0022.flac", "task_type": "lexical_counting", "query": "spoke", "prompt": "Process the full audio offline. For every occurrence of the word \"spoke\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.4, "one"], [4.56, 4.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.11,\"end\":1.4},{\"word\":\"two\",\"start\":4.56,\"end\":4.87}]"} {"utterance_id": "2506-11267-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0004.flac", "task_type": "lexical_detection", "query": "hunting", "prompt": "Process the full audio offline. Find every instance of the word \"hunting\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.39, "hunting"]], "response_mode": "offline", "answer": "[{\"word\":\"hunting\",\"start\":3.05,\"end\":3.39}]"} {"utterance_id": "1701-141760-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0046.flac", "task_type": "lexical_detection", "query": "slightly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"slightly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 0.74, "slightly"]], "response_mode": "offline", "answer": "[{\"word\":\"slightly\",\"start\":0.36,\"end\":0.74}]"} {"utterance_id": "7697-105817-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0009.flac", "task_type": "lexical_detection", "query": "despair", "prompt": "Once the audio has finished, identify every instance of the word \"despair\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.28, 1.81, "despair"]], "response_mode": "offline", "answer": "[{\"word\":\"despair\",\"start\":1.28,\"end\":1.81}]"} {"utterance_id": "7601-101619-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0005.flac", "task_type": "lexical_detection", "query": "dives", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"dives\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.95, 1.4, "dives"]], "response_mode": "offline", "answer": "[{\"word\":\"dives\",\"start\":0.95,\"end\":1.4}]"} {"utterance_id": "6267-65525-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0039.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[2.47, 3.0, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":2.47,\"end\":3.0}]"} {"utterance_id": "3660-172182-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0026.flac", "task_type": "lexical_detection", "query": "abode", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"abode\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.23, 1.51, "abode"]], "response_mode": "offline", "answer": "[{\"word\":\"abode\",\"start\":1.23,\"end\":1.51}]"} {"utterance_id": "8254-115543-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0022.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "When the utterance is complete, report all instances of an ordinal number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.93, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.51,\"end\":0.93}]"} {"utterance_id": "4572-64670-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0008.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.0, 4.34, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.0,\"end\":4.34}]"} {"utterance_id": "4323-13259-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0018.flac", "task_type": "lexical_detection", "query": "circumstances", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"circumstances\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.6, 1.43, "circumstances"]], "response_mode": "offline", "answer": "[{\"word\":\"circumstances\",\"start\":0.6,\"end\":1.43}]"} {"utterance_id": "3660-172183-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.18, 3.37, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.18,\"end\":3.37}]"} {"utterance_id": "4570-56594-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0002.flac", "task_type": "lexical_detection", "query": "ain't", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ain't\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.9, 1.04, "ain't"]], "response_mode": "offline", "answer": "[{\"word\":\"ain't\",\"start\":0.9,\"end\":1.04}]"} {"utterance_id": "3660-6517-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0013.flac", "task_type": "lexical_detection", "query": "wholly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"wholly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.04, 1.45, "wholly"]], "response_mode": "offline", "answer": "[{\"word\":\"wholly\",\"start\":1.04,\"end\":1.45}]"} {"utterance_id": "8254-84205-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0002.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.72, 3.15, "animal"]], "response_mode": "offline", "answer": "[{\"word\":\"animal\",\"start\":2.72,\"end\":3.15}]"} {"utterance_id": "5849-50873-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0040.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "After the audio ends, tally every instance of an organization name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.92, 2.5, "one"], [10.18, 11.38, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.92,\"end\":2.5},{\"word\":\"two\",\"start\":10.18,\"end\":11.38}]"} {"utterance_id": "4323-55228-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0034.flac", "task_type": "lexical_detection", "query": "another", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"another\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.91, "another"]], "response_mode": "offline", "answer": "[{\"word\":\"another\",\"start\":0.63,\"end\":0.91}]"} {"utterance_id": "1686-142278-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0031.flac", "task_type": "lexical_detection", "query": "putting", "prompt": "After the audio ends, extract all occurrences of the word \"putting\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.31, 3.6, "putting"]], "response_mode": "offline", "answer": "[{\"word\":\"putting\",\"start\":3.31,\"end\":3.6}]"} {"utterance_id": "4153-186223-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0001.flac", "task_type": "lexical_detection", "query": "neither", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"neither\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.87, 4.18, "neither"]], "response_mode": "offline", "answer": "[{\"word\":\"neither\",\"start\":3.87,\"end\":4.18}]"} {"utterance_id": "6455-66379-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0005.flac", "task_type": "lexical_detection", "query": "learn", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"learn\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.91, 1.25, "learn"], [2.36, 2.6, "learn"]], "response_mode": "offline", "answer": "[{\"word\":\"learn\",\"start\":0.91,\"end\":1.25},{\"word\":\"learn\",\"start\":2.36,\"end\":2.6}]"} {"utterance_id": "3660-172182-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0013.flac", "task_type": "lexical_detection", "query": "morrow", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"morrow\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.89, "morrow"]], "response_mode": "offline", "answer": "[{\"word\":\"morrow\",\"start\":0.6,\"end\":0.89}]"} {"utterance_id": "700-122868-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0006.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.02, "one"], [2.11, 2.52, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.77,\"end\":1.02},{\"word\":\"two\",\"start\":2.11,\"end\":2.52}]"} {"utterance_id": "1650-167613-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0053.flac", "task_type": "lexical_detection", "query": "meeting", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"meeting\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.45, 2.94, "meeting"]], "response_mode": "offline", "answer": "[{\"word\":\"meeting\",\"start\":2.45,\"end\":2.94}]"} {"utterance_id": "5543-27761-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0005.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.45, 2.01, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":1.45,\"end\":2.01}]"} {"utterance_id": "1701-141760-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0037.flac", "task_type": "lexical_detection", "query": "smiling", "prompt": "After the audio ends, extract all occurrences of the word \"smiling\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.19, 2.7, "smiling"]], "response_mode": "offline", "answer": "[{\"word\":\"smiling\",\"start\":2.19,\"end\":2.7}]"} {"utterance_id": "8173-294714-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0049.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.79, 3.72, "mister van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"mister van brandt\",\"start\":2.79,\"end\":3.72}]"} {"utterance_id": "6467-94831-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0030.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.48, 5.01, "statesmen"]], "response_mode": "offline", "answer": "[{\"word\":\"statesmen\",\"start\":4.48,\"end\":5.01}]"} {"utterance_id": "1255-74899-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.05, 2.59, "mister thwaite"]], "response_mode": "offline", "answer": "[{\"word\":\"mister thwaite\",\"start\":2.05,\"end\":2.59}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.11, "father"], [7.95, 8.31, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":0.81,\"end\":1.11},{\"word\":\"mother\",\"start\":7.95,\"end\":8.31}]"} {"utterance_id": "8173-294714-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0011.flac", "task_type": "lexical_counting", "query": "misfortune", "prompt": "After listening to the whole recording, output timed count words for the word \"misfortune\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.22, 2.77, "one"], [3.77, 4.39, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.22,\"end\":2.77},{\"word\":\"two\",\"start\":3.77,\"end\":4.39}]"} {"utterance_id": "6267-65525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0021.flac", "task_type": "lexical_detection", "query": "remote", "prompt": "When the utterance is complete, report all instances of the word \"remote\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.91, 8.39, "remote"]], "response_mode": "offline", "answer": "[{\"word\":\"remote\",\"start\":7.91,\"end\":8.39}]"} {"utterance_id": "116-288046-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0000.flac", "task_type": "lexical_detection", "query": "communicate", "prompt": "When the utterance is complete, report all instances of the word \"communicate\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.94, 6.52, "communicate"]], "response_mode": "offline", "answer": "[{\"word\":\"communicate\",\"start\":5.94,\"end\":6.52}]"} {"utterance_id": "7697-105815-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.02, "timorous"]], "response_mode": "offline", "answer": "[{\"word\":\"timorous\",\"start\":0.5,\"end\":1.02}]"} {"utterance_id": "8288-274162-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0038.flac", "task_type": "lexical_detection", "query": "how", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"how\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.56, "how"]], "response_mode": "offline", "answer": "[{\"word\":\"how\",\"start\":0.26,\"end\":0.56}]"} {"utterance_id": "4515-11057-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0004.flac", "task_type": "lexical_detection", "query": "unable", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"unable\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.91, "unable"]], "response_mode": "offline", "answer": "[{\"word\":\"unable\",\"start\":0.39,\"end\":0.91}]"} {"utterance_id": "2506-11278-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0014.flac", "task_type": "lexical_detection", "query": "taken", "prompt": "Process the full audio offline. Find every instance of the word \"taken\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.48, "taken"]], "response_mode": "offline", "answer": "[{\"word\":\"taken\",\"start\":1.14,\"end\":1.48}]"} {"utterance_id": "5849-50962-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0016.flac", "task_type": "lexical_detection", "query": "arms", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"arms\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.76, 6.07, "arms"]], "response_mode": "offline", "answer": "[{\"word\":\"arms\",\"start\":5.76,\"end\":6.07}]"} {"utterance_id": "8254-115543-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0016.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.52, "dogs"], [4.72, 5.02, "dogs"]], "response_mode": "offline", "answer": "[{\"word\":\"dogs\",\"start\":1.07,\"end\":1.52},{\"word\":\"dogs\",\"start\":4.72,\"end\":5.02}]"} {"utterance_id": "6841-88294-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0022.flac", "task_type": "lexical_detection", "query": "soon", "prompt": "Once the audio has finished, identify every instance of the word \"soon\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.59, 3.77, "soon"]], "response_mode": "offline", "answer": "[{\"word\":\"soon\",\"start\":3.59,\"end\":3.77}]"} {"utterance_id": "3915-98647-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0022.flac", "task_type": "lexical_detection", "query": "rule", "prompt": "Process the full audio offline. Find every instance of the word \"rule\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 1.88, "rule"]], "response_mode": "offline", "answer": "[{\"word\":\"rule\",\"start\":1.71,\"end\":1.88}]"} {"utterance_id": "5849-50963-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0012.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "When the utterance is complete, report all instances of an organization name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.12, 4.41, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"army\",\"start\":4.12,\"end\":4.41}]"} {"utterance_id": "5849-50962-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.9, "hal"], [2.27, 2.66, "nan"]], "response_mode": "offline", "answer": "[{\"word\":\"hal\",\"start\":0.63,\"end\":0.9},{\"word\":\"nan\",\"start\":2.27,\"end\":2.66}]"} {"utterance_id": "7697-105815-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0051.flac", "task_type": "lexical_detection", "query": "ass", "prompt": "Process the full audio offline. Find every instance of the word \"ass\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.5, "ass"]], "response_mode": "offline", "answer": "[{\"word\":\"ass\",\"start\":2.18,\"end\":2.5}]"} {"utterance_id": "4570-14911-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0001.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.44, 4.02, "sixteen"], [11.04, 11.4, "twenty"], [13.33, 13.79, "nineteen"]], "response_mode": "offline", "answer": "[{\"word\":\"sixteen\",\"start\":3.44,\"end\":4.02},{\"word\":\"twenty\",\"start\":11.04,\"end\":11.4},{\"word\":\"nineteen\",\"start\":13.33,\"end\":13.79}]"} {"utterance_id": "1585-131718-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0000.flac", "task_type": "lexical_detection", "query": "necessary", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"necessary\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.67, "necessary"]], "response_mode": "offline", "answer": "[{\"word\":\"necessary\",\"start\":0.99,\"end\":1.67}]"} {"utterance_id": "7641-96670-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.02, "heart"], [3.77, 4.07, "eyes"], [5.12, 5.59, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":0.77,\"end\":1.02},{\"word\":\"eyes\",\"start\":3.77,\"end\":4.07},{\"word\":\"eyes\",\"start\":5.12,\"end\":5.59}]"} {"utterance_id": "116-288045-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0012.flac", "task_type": "lexical_detection", "query": "fault", "prompt": "Once the audio has finished, identify every instance of the word \"fault\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.34, 1.78, "fault"]], "response_mode": "offline", "answer": "[{\"word\":\"fault\",\"start\":1.34,\"end\":1.78}]"} {"utterance_id": "116-288045-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0004.flac", "task_type": "lexical_detection", "query": "guide", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"guide\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.01, 3.39, "guide"]], "response_mode": "offline", "answer": "[{\"word\":\"guide\",\"start\":3.01,\"end\":3.39}]"} {"utterance_id": "4515-11057-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[5.52, 6.1, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":5.52,\"end\":6.1}]"} {"utterance_id": "4323-13259-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0007.flac", "task_type": "lexical_detection", "query": "yielded", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"yielded\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.64, "yielded"]], "response_mode": "offline", "answer": "[{\"word\":\"yielded\",\"start\":1.3,\"end\":1.64}]"} {"utterance_id": "3915-98647-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0022.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.1, 6.49, "hag"]], "response_mode": "offline", "answer": "[{\"word\":\"hag\",\"start\":6.1,\"end\":6.49}]"} {"utterance_id": "6841-88294-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0065.flac", "task_type": "lexical_detection", "query": "talked", "prompt": "Once the audio has finished, identify every instance of the word \"talked\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.55, 0.8, "talked"]], "response_mode": "offline", "answer": "[{\"word\":\"talked\",\"start\":0.55,\"end\":0.8}]"} {"utterance_id": "6123-59150-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0018.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "After listening to the whole recording, output timed count words for a number (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.52, 0.71, "one"], [3.11, 3.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.52,\"end\":0.71},{\"word\":\"two\",\"start\":3.11,\"end\":3.46}]"} {"utterance_id": "8254-115543-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0014.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.38, 7.72, "alms"]], "response_mode": "offline", "answer": "[{\"word\":\"alms\",\"start\":7.38,\"end\":7.72}]"} {"utterance_id": "4323-13259-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0006.flac", "task_type": "lexical_detection", "query": "ago", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ago\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.72, 3.11, "ago"]], "response_mode": "offline", "answer": "[{\"word\":\"ago\",\"start\":2.72,\"end\":3.11}]"} {"utterance_id": "4831-29134-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0001.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[4.28, 4.54, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":4.28,\"end\":4.54}]"} {"utterance_id": "1701-141760-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0035.flac", "task_type": "lexical_counting", "query": "day", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"day\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.69, 0.88, "one"], [2.74, 3.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.69,\"end\":0.88},{\"word\":\"two\",\"start\":2.74,\"end\":3.03}]"} {"utterance_id": "5543-27761-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0039.flac", "task_type": "lexical_detection", "query": "doll", "prompt": "After listening to the whole recording, output every instance of the word \"doll\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.71, 5.18, "doll"]], "response_mode": "offline", "answer": "[{\"word\":\"doll\",\"start\":4.71,\"end\":5.18}]"} {"utterance_id": "4831-29134-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.23, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.98,\"end\":3.23}]"} {"utterance_id": "8254-115543-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0014.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Once the audio has finished, list each occurrence of a number as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.48, 0.75, "one"], [4.63, 4.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.48,\"end\":0.75},{\"word\":\"two\",\"start\":4.63,\"end\":4.87}]"} {"utterance_id": "7641-96670-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0019.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Process the full audio offline. For every occurrence of a body part, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.47, "one"], [3.64, 4.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.16,\"end\":1.47},{\"word\":\"two\",\"start\":3.64,\"end\":4.03}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "You will hear a complete utterance. When it ends, list every instance of a time of day you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.55, 11.04, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":10.55,\"end\":11.04}]"} {"utterance_id": "6123-59150-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0035.flac", "task_type": "lexical_detection", "query": "deny", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"deny\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.65, 2.07, "deny"]], "response_mode": "offline", "answer": "[{\"word\":\"deny\",\"start\":1.65,\"end\":2.07}]"} {"utterance_id": "6841-88291-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0054.flac", "task_type": "lexical_detection", "query": "pocket", "prompt": "After the audio ends, extract all occurrences of the word \"pocket\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.23, 2.68, "pocket"]], "response_mode": "offline", "answer": "[{\"word\":\"pocket\",\"start\":2.23,\"end\":2.68}]"} {"utterance_id": "700-122867-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0003.flac", "task_type": "lexical_detection", "query": "marilla", "prompt": "Once the audio has finished, identify every instance of the word \"marilla\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.62, 3.09, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":2.62,\"end\":3.09}]"} {"utterance_id": "3660-172182-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 2.26, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.81,\"end\":2.26}]"} {"utterance_id": "4831-29134-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0003.flac", "task_type": "lexical_detection", "query": "hurried", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"hurried\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.36, 2.73, "hurried"]], "response_mode": "offline", "answer": "[{\"word\":\"hurried\",\"start\":2.36,\"end\":2.73}]"} {"utterance_id": "700-122868-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.63, 0.91, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":0.63,\"end\":0.91}]"} {"utterance_id": "2506-11267-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0014.flac", "task_type": "lexical_detection", "query": "yonder", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"yonder\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.54, 1.01, "yonder"]], "response_mode": "offline", "answer": "[{\"word\":\"yonder\",\"start\":0.54,\"end\":1.01}]"} {"utterance_id": "4323-13259-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0026.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.42, "hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"hundred\",\"start\":1.06,\"end\":1.42}]"} {"utterance_id": "2506-169427-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0006.flac", "task_type": "lexical_detection", "query": "yorick", "prompt": "After the audio ends, extract all occurrences of the word \"yorick\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.29, "yorick"]], "response_mode": "offline", "answer": "[{\"word\":\"yorick\",\"start\":1.68,\"end\":2.29}]"} {"utterance_id": "3663-172528-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[16.34, 16.78, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":16.34,\"end\":16.78}]"} {"utterance_id": "4323-13259-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0020.flac", "task_type": "lexical_detection", "query": "impressive", "prompt": "When the utterance is complete, report all instances of the word \"impressive\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.48, "impressive"]], "response_mode": "offline", "answer": "[{\"word\":\"impressive\",\"start\":3.94,\"end\":4.48}]"} {"utterance_id": "1701-141760-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0044.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After listening to the whole recording, output every instance of an organization name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.59, 8.14, "companies"]], "response_mode": "offline", "answer": "[{\"word\":\"companies\",\"start\":7.59,\"end\":8.14}]"} {"utterance_id": "4831-18525-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0016.flac", "task_type": "lexical_counting", "query": "know", "prompt": "When the utterance is complete, report how the word \"know\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 0.87, "one"], [7.54, 7.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.75,\"end\":0.87},{\"word\":\"two\",\"start\":7.54,\"end\":7.76}]"} {"utterance_id": "1686-142278-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.11, "mister hale"], [3.64, 4.1, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale\",\"start\":0.5,\"end\":1.11},{\"word\":\"margaret\",\"start\":3.64,\"end\":4.1}]"} {"utterance_id": "1630-141772-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0018.flac", "task_type": "lexical_detection", "query": "calm", "prompt": "Process the full audio offline. Find every instance of the word \"calm\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.35, 1.69, "calm"]], "response_mode": "offline", "answer": "[{\"word\":\"calm\",\"start\":1.35,\"end\":1.69}]"} {"utterance_id": "700-122867-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0025.flac", "task_type": "lexical_detection", "query": "having", "prompt": "Once the audio has finished, identify every instance of the word \"having\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.01, 2.31, "having"]], "response_mode": "offline", "answer": "[{\"word\":\"having\",\"start\":2.01,\"end\":2.31}]"} {"utterance_id": "700-122868-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0034.flac", "task_type": "lexical_detection", "query": "either", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"either\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.29, 3.71, "either"]], "response_mode": "offline", "answer": "[{\"word\":\"either\",\"start\":3.29,\"end\":3.71}]"} {"utterance_id": "5849-50873-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0024.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 0.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":0.7,\"end\":0.86}]"} {"utterance_id": "5543-27761-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0098.flac", "task_type": "lexical_detection", "query": "walked", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"walked\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.13, 1.42, "walked"]], "response_mode": "offline", "answer": "[{\"word\":\"walked\",\"start\":1.13,\"end\":1.42}]"} {"utterance_id": "8254-84205-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0004.flac", "task_type": "lexical_detection", "query": "regularly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"regularly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.75, 3.34, "regularly"]], "response_mode": "offline", "answer": "[{\"word\":\"regularly\",\"start\":2.75,\"end\":3.34}]"} {"utterance_id": "4515-11057-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0001.flac", "task_type": "lexical_detection", "query": "due", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"due\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.15, 2.42, "due"]], "response_mode": "offline", "answer": "[{\"word\":\"due\",\"start\":2.15,\"end\":2.42}]"} {"utterance_id": "5849-50963-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0000.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Listen carefully to the entire clip, then report every instance of a weather term as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.78, 3.36, "storm"]], "response_mode": "offline", "answer": "[{\"word\":\"storm\",\"start\":2.78,\"end\":3.36}]"} {"utterance_id": "1630-73710-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.29, 4.95, "ferdinand"]], "response_mode": "offline", "answer": "[{\"word\":\"ferdinand\",\"start\":4.29,\"end\":4.95}]"} {"utterance_id": "116-288048-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0009.flac", "task_type": "lexical_detection", "query": "argue", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"argue\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.2, 3.62, "argue"]], "response_mode": "offline", "answer": "[{\"word\":\"argue\",\"start\":3.2,\"end\":3.62}]"} {"utterance_id": "1701-141760-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0002.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Once the audio has finished, identify every instance of an organization name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.64, 1.14, "regiments"], [3.22, 3.6, "bands"], [11.19, 11.54, "guards"]], "response_mode": "offline", "answer": "[{\"word\":\"regiments\",\"start\":0.64,\"end\":1.14},{\"word\":\"bands\",\"start\":3.22,\"end\":3.6},{\"word\":\"guards\",\"start\":11.19,\"end\":11.54}]"} {"utterance_id": "8254-115543-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0009.flac", "task_type": "lexical_detection", "query": "oxen", "prompt": "When the utterance is complete, report all instances of the word \"oxen\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.24, 3.66, "oxen"]], "response_mode": "offline", "answer": "[{\"word\":\"oxen\",\"start\":3.24,\"end\":3.66}]"} {"utterance_id": "6841-88294-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0035.flac", "task_type": "lexical_detection", "query": "afterwards", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"afterwards\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.26, 4.73, "afterwards"]], "response_mode": "offline", "answer": "[{\"word\":\"afterwards\",\"start\":4.26,\"end\":4.73}]"} {"utterance_id": "6599-38590-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0001.flac", "task_type": "lexical_detection", "query": "shame", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"shame\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.79, 10.24, "shame"]], "response_mode": "offline", "answer": "[{\"word\":\"shame\",\"start\":9.79,\"end\":10.24}]"} {"utterance_id": "3915-98647-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0000.flac", "task_type": "lexical_detection", "query": "shrine", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"shrine\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.4, 4.84, "shrine"]], "response_mode": "offline", "answer": "[{\"word\":\"shrine\",\"start\":4.4,\"end\":4.84}]"} {"utterance_id": "6841-88294-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0008.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.94, "wine"], [4.05, 4.57, "syrup"]], "response_mode": "offline", "answer": "[{\"word\":\"wine\",\"start\":0.68,\"end\":0.94},{\"word\":\"syrup\",\"start\":4.05,\"end\":4.57}]"} {"utterance_id": "7641-96252-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.45, 4.9, "bosom"]], "response_mode": "offline", "answer": "[{\"word\":\"bosom\",\"start\":4.45,\"end\":4.9}]"} {"utterance_id": "1701-141760-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.6, 3.0, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"rostov\",\"start\":2.6,\"end\":3.0}]"} {"utterance_id": "6841-88291-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0047.flac", "task_type": "lexical_detection", "query": "reckon", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"reckon\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.03, "reckon"]], "response_mode": "offline", "answer": "[{\"word\":\"reckon\",\"start\":2.7,\"end\":3.03}]"} {"utterance_id": "7697-105815-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0047.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.28, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.88,\"end\":2.28}]"} {"utterance_id": "7697-105817-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0005.flac", "task_type": "lexical_detection", "query": "field", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"field\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.93, "field"]], "response_mode": "offline", "answer": "[{\"word\":\"field\",\"start\":1.59,\"end\":1.93}]"} {"utterance_id": "1650-167613-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0045.flac", "task_type": "lexical_detection", "query": "everyone", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"everyone\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.28, 5.98, "everyone"]], "response_mode": "offline", "answer": "[{\"word\":\"everyone\",\"start\":5.28,\"end\":5.98}]"} {"utterance_id": "2506-169427-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0004.flac", "task_type": "lexical_detection", "query": "inclined", "prompt": "Once the audio has finished, identify every instance of the word \"inclined\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.7, 1.17, "inclined"]], "response_mode": "offline", "answer": "[{\"word\":\"inclined\",\"start\":0.7,\"end\":1.17}]"} {"utterance_id": "700-122867-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0007.flac", "task_type": "lexical_detection", "query": "get", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"get\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.72, "get"]], "response_mode": "offline", "answer": "[{\"word\":\"get\",\"start\":0.46,\"end\":0.72}]"} {"utterance_id": "6841-88291-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0042.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.28, 1.69, "calves"]], "response_mode": "offline", "answer": "[{\"word\":\"calves\",\"start\":1.28,\"end\":1.69}]"} {"utterance_id": "6467-56885-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.01, 5.24, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.01,\"end\":5.24}]"} {"utterance_id": "8254-115543-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0008.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "After the audio ends, tally every instance of a profession. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[2.05, 2.44, "one"], [4.61, 5.1, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.05,\"end\":2.44},{\"word\":\"two\",\"start\":4.61,\"end\":5.1}]"} {"utterance_id": "8288-274150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0003.flac", "task_type": "lexical_counting", "query": "nor", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"nor\" (word, start, end). If none, reply [].", "labels": [[3.47, 3.73, "one"], [5.08, 5.35, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.47,\"end\":3.73},{\"word\":\"two\",\"start\":5.08,\"end\":5.35}]"} {"utterance_id": "700-122868-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0003.flac", "task_type": "lexical_detection", "query": "lily", "prompt": "After listening to the whole recording, output every instance of the word \"lily\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.32, 1.59, "lily"]], "response_mode": "offline", "answer": "[{\"word\":\"lily\",\"start\":1.32,\"end\":1.59}]"} {"utterance_id": "6123-59186-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0032.flac", "task_type": "lexical_counting", "query": "world", "prompt": "Once the audio has finished, list each occurrence of the word \"world\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.92, 1.24, "one"], [2.38, 2.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.92,\"end\":1.24},{\"word\":\"two\",\"start\":2.38,\"end\":2.63}]"} {"utterance_id": "8288-274162-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0059.flac", "task_type": "lexical_detection", "query": "admit", "prompt": "After listening to the whole recording, output every instance of the word \"admit\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.25, "admit"]], "response_mode": "offline", "answer": "[{\"word\":\"admit\",\"start\":0.93,\"end\":1.25}]"} {"utterance_id": "1630-73710-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0003.flac", "task_type": "lexical_detection", "query": "changes", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"changes\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.72, 2.24, "changes"]], "response_mode": "offline", "answer": "[{\"word\":\"changes\",\"start\":1.72,\"end\":2.24}]"} {"utterance_id": "6841-88294-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0020.flac", "task_type": "lexical_detection", "query": "long", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"long\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.13, "long"]], "response_mode": "offline", "answer": "[{\"word\":\"long\",\"start\":1.84,\"end\":2.13}]"} {"utterance_id": "3663-172528-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0023.flac", "task_type": "lexical_detection", "query": "dearest", "prompt": "Process the full audio offline. Find every instance of the word \"dearest\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.35, 7.89, "dearest"]], "response_mode": "offline", "answer": "[{\"word\":\"dearest\",\"start\":7.35,\"end\":7.89}]"} {"utterance_id": "116-288045-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0001.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.41, "black"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":3.05,\"end\":3.41}]"} {"utterance_id": "8173-294714-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0034.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.5, 1.04, "doctor's"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor's\",\"start\":0.5,\"end\":1.04}]"} {"utterance_id": "6467-94831-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0014.flac", "task_type": "lexical_detection", "query": "feelings", "prompt": "Process the full audio offline. Find every instance of the word \"feelings\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.62, "feelings"]], "response_mode": "offline", "answer": "[{\"word\":\"feelings\",\"start\":3.06,\"end\":3.62}]"} {"utterance_id": "3663-172528-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0013.flac", "task_type": "lexical_detection", "query": "drunk", "prompt": "Once the audio has finished, identify every instance of the word \"drunk\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.21, 4.61, "drunk"]], "response_mode": "offline", "answer": "[{\"word\":\"drunk\",\"start\":4.21,\"end\":4.61}]"} {"utterance_id": "6455-67803-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0020.flac", "task_type": "lexical_detection", "query": "fast", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"fast\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.51, "fast"]], "response_mode": "offline", "answer": "[{\"word\":\"fast\",\"start\":1.11,\"end\":1.51}]"} {"utterance_id": "4570-14911-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0001.flac", "task_type": "lexical_counting", "query": "inches", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"inches\" with word/start/end. If none, reply [].", "labels": [[4.02, 4.31, "one"], [11.4, 11.88, "two"], [13.79, 13.96, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.02,\"end\":4.31},{\"word\":\"two\",\"start\":11.4,\"end\":11.88},{\"word\":\"three\",\"start\":13.79,\"end\":13.96}]"} {"utterance_id": "4570-14911-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0008.flac", "task_type": "semantic_detection", "query": "country", "prompt": "You will hear a complete utterance. When it ends, list every instance of a country name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 2.77, "country"]], "response_mode": "offline", "answer": "[{\"word\":\"country\",\"start\":2.34,\"end\":2.77}]"} {"utterance_id": "3660-6517-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0034.flac", "task_type": "lexical_detection", "query": "measure", "prompt": "Once the audio has finished, identify every instance of the word \"measure\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[10.05, 10.35, "measure"]], "response_mode": "offline", "answer": "[{\"word\":\"measure\",\"start\":10.05,\"end\":10.35}]"} {"utterance_id": "6467-97061-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0000.flac", "task_type": "lexical_detection", "query": "sleeping", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"sleeping\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.17, "sleeping"]], "response_mode": "offline", "answer": "[{\"word\":\"sleeping\",\"start\":1.77,\"end\":2.17}]"} {"utterance_id": "8254-115543-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.5, "neck"]], "response_mode": "offline", "answer": "[{\"word\":\"neck\",\"start\":1.28,\"end\":1.5}]"} {"utterance_id": "7641-96252-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0012.flac", "task_type": "lexical_detection", "query": "anxiety", "prompt": "When the utterance is complete, report all instances of the word \"anxiety\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.99, 5.71, "anxiety"]], "response_mode": "offline", "answer": "[{\"word\":\"anxiety\",\"start\":4.99,\"end\":5.71}]"} {"utterance_id": "7601-175351-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0001.flac", "task_type": "lexical_detection", "query": "forward", "prompt": "Once the audio has finished, identify every instance of the word \"forward\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.1, 5.71, "forward"]], "response_mode": "offline", "answer": "[{\"word\":\"forward\",\"start\":5.1,\"end\":5.71}]"} {"utterance_id": "4831-25894-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0030.flac", "task_type": "lexical_counting", "query": "dress", "prompt": "Listen to the entire clip, then report the count sequence for the word \"dress\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.59, 5.03, "one"], [8.12, 8.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.59,\"end\":5.03},{\"word\":\"two\",\"start\":8.12,\"end\":8.53}]"} {"utterance_id": "5543-27761-0077", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0077.flac", "task_type": "lexical_detection", "query": "fever", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"fever\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 0.72, "fever"]], "response_mode": "offline", "answer": "[{\"word\":\"fever\",\"start\":0.36,\"end\":0.72}]"} {"utterance_id": "8288-274162-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0033.flac", "task_type": "lexical_counting", "query": "enough", "prompt": "After listening to the whole recording, output timed count words for the word \"enough\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.62, 2.93, "one"], [4.47, 4.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.62,\"end\":2.93},{\"word\":\"two\",\"start\":4.47,\"end\":4.78}]"} {"utterance_id": "700-122868-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0015.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.91, 1.21, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.91,\"end\":1.21}]"} {"utterance_id": "4323-18416-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0010.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.64, 4.94, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":4.64,\"end\":4.94}]"} {"utterance_id": "4153-186222-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0010.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Once the audio has finished, identify every instance of a money or currency term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.69, 4.09, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":3.69,\"end\":4.09}]"} {"utterance_id": "5849-50873-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0003.flac", "task_type": "lexical_detection", "query": "refused", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"refused\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 2.02, "refused"]], "response_mode": "offline", "answer": "[{\"word\":\"refused\",\"start\":1.48,\"end\":2.02}]"} {"utterance_id": "4831-25894-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0012.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.16, 2.86, "one"], [5.44, 5.77, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.16,\"end\":2.86},{\"word\":\"two\",\"start\":5.44,\"end\":5.77}]"} {"utterance_id": "1630-96099-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0001.flac", "task_type": "lexical_detection", "query": "special", "prompt": "When the utterance is complete, report all instances of the word \"special\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.83, "special"]], "response_mode": "offline", "answer": "[{\"word\":\"special\",\"start\":2.38,\"end\":2.83}]"} {"utterance_id": "6455-67803-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0012.flac", "task_type": "lexical_detection", "query": "smiled", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"smiled\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.55, 0.85, "smiled"]], "response_mode": "offline", "answer": "[{\"word\":\"smiled\",\"start\":0.55,\"end\":0.85}]"} {"utterance_id": "3663-172005-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0000.flac", "task_type": "lexical_detection", "query": "fawn", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fawn\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.85, "fawn"]], "response_mode": "offline", "answer": "[{\"word\":\"fawn\",\"start\":0.47,\"end\":0.85}]"} {"utterance_id": "6467-94831-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.5, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":3.1,\"end\":3.5}]"} {"utterance_id": "6467-94831-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0019.flac", "task_type": "lexical_detection", "query": "trouble", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"trouble\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.4, "trouble"]], "response_mode": "offline", "answer": "[{\"word\":\"trouble\",\"start\":0.95,\"end\":1.4}]"} {"utterance_id": "4515-11057-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0034.flac", "task_type": "lexical_detection", "query": "paper", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"paper\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 2.01, "paper"]], "response_mode": "offline", "answer": "[{\"word\":\"paper\",\"start\":1.52,\"end\":2.01}]"} {"utterance_id": "1585-157660-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0011.flac", "task_type": "lexical_detection", "query": "jack", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"jack\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.37, "jack"]], "response_mode": "offline", "answer": "[{\"word\":\"jack\",\"start\":0.97,\"end\":1.37}]"} {"utterance_id": "1255-138279-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0014.flac", "task_type": "lexical_detection", "query": "window", "prompt": "After the audio ends, extract all occurrences of the word \"window\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 0.84, "window"]], "response_mode": "offline", "answer": "[{\"word\":\"window\",\"start\":0.61,\"end\":0.84}]"} {"utterance_id": "4831-25894-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0016.flac", "task_type": "semantic_detection", "query": "money", "prompt": "You will hear a complete utterance. When it ends, list every instance of a money or currency term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.0, 11.47, "dollars"]], "response_mode": "offline", "answer": "[{\"word\":\"dollars\",\"start\":11.0,\"end\":11.47}]"} {"utterance_id": "3663-172528-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.53, 3.07, "maestro agostino"]], "response_mode": "offline", "answer": "[{\"word\":\"maestro agostino\",\"start\":1.53,\"end\":3.07}]"} {"utterance_id": "1701-141759-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0026.flac", "task_type": "lexical_detection", "query": "ear", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"ear\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.66, "ear"]], "response_mode": "offline", "answer": "[{\"word\":\"ear\",\"start\":1.45,\"end\":1.66}]"} {"utterance_id": "4570-102353-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0007.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "You will hear a complete utterance. When it ends, produce a running count of an organization name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[11.2, 11.56, "one"], [14.65, 15.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":11.2,\"end\":11.56},{\"word\":\"two\",\"start\":14.65,\"end\":15.36}]"} {"utterance_id": "3915-57461-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0010.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.85, 8.17, "ronin"]], "response_mode": "offline", "answer": "[{\"word\":\"ronin\",\"start\":7.85,\"end\":8.17}]"} {"utterance_id": "4323-55228-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0029.flac", "task_type": "lexical_detection", "query": "see", "prompt": "After listening to the whole recording, output every instance of the word \"see\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 1.05, "see"]], "response_mode": "offline", "answer": "[{\"word\":\"see\",\"start\":0.62,\"end\":1.05}]"} {"utterance_id": "2506-11267-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0012.flac", "task_type": "lexical_detection", "query": "himself", "prompt": "After listening to the whole recording, output every instance of the word \"himself\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.36, "himself"]], "response_mode": "offline", "answer": "[{\"word\":\"himself\",\"start\":0.95,\"end\":1.36}]"} {"utterance_id": "6841-88294-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0062.flac", "task_type": "lexical_detection", "query": "fellow", "prompt": "After listening to the whole recording, output every instance of the word \"fellow\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.13, 3.38, "fellow"]], "response_mode": "offline", "answer": "[{\"word\":\"fellow\",\"start\":3.13,\"end\":3.38}]"} {"utterance_id": "6467-94831-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0032.flac", "task_type": "lexical_detection", "query": "process", "prompt": "Once the audio has finished, identify every instance of the word \"process\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.13, 2.75, "process"]], "response_mode": "offline", "answer": "[{\"word\":\"process\",\"start\":2.13,\"end\":2.75}]"} {"utterance_id": "1686-142278-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0053.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.02, "milton"]], "response_mode": "offline", "answer": "[{\"word\":\"milton\",\"start\":0.5,\"end\":1.02}]"} {"utterance_id": "4831-29134-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 3.98, "minister"], [8.62, 9.4, "superintendent"]], "response_mode": "offline", "answer": "[{\"word\":\"minister\",\"start\":3.52,\"end\":3.98},{\"word\":\"superintendent\",\"start\":8.62,\"end\":9.4}]"} {"utterance_id": "4831-18525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0007.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a color name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.07, 2.27, "blue"]], "response_mode": "offline", "answer": "[{\"word\":\"blue\",\"start\":2.07,\"end\":2.27}]"} {"utterance_id": "6467-94831-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0037.flac", "task_type": "lexical_detection", "query": "sam", "prompt": "Process the full audio offline. Find every instance of the word \"sam\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.61, "sam"]], "response_mode": "offline", "answer": "[{\"word\":\"sam\",\"start\":0.21,\"end\":0.61}]"} {"utterance_id": "6267-65525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0022.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a plant name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.84, 1.85, "evergreens"]], "response_mode": "offline", "answer": "[{\"word\":\"evergreens\",\"start\":0.84,\"end\":1.85}]"} {"utterance_id": "7697-105815-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0035.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.44, 2.83, "beasts"]], "response_mode": "offline", "answer": "[{\"word\":\"beasts\",\"start\":2.44,\"end\":2.83}]"} {"utterance_id": "7601-175351-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 4.09, "warren"]], "response_mode": "offline", "answer": "[{\"word\":\"warren\",\"start\":3.56,\"end\":4.09}]"} {"utterance_id": "8288-274162-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0043.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.55, 4.9, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":4.55,\"end\":4.9}]"} {"utterance_id": "1255-74899-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0006.flac", "task_type": "lexical_detection", "query": "weak", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"weak\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 1.86, "weak"]], "response_mode": "offline", "answer": "[{\"word\":\"weak\",\"start\":1.57,\"end\":1.86}]"} {"utterance_id": "8254-84205-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0031.flac", "task_type": "lexical_detection", "query": "don't", "prompt": "After the audio ends, extract all occurrences of the word \"don't\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.65, "don't"]], "response_mode": "offline", "answer": "[{\"word\":\"don't\",\"start\":0.32,\"end\":0.65}]"} {"utterance_id": "5849-50962-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0020.flac", "task_type": "lexical_detection", "query": "guess", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"guess\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.3, 0.53, "guess"]], "response_mode": "offline", "answer": "[{\"word\":\"guess\",\"start\":0.3,\"end\":0.53}]"} {"utterance_id": "3663-172528-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0029.flac", "task_type": "lexical_detection", "query": "agostino", "prompt": "After listening to the whole recording, output every instance of the word \"agostino\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 3.07, "agostino"]], "response_mode": "offline", "answer": "[{\"word\":\"agostino\",\"start\":2.09,\"end\":3.07}]"} {"utterance_id": "4572-112381-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0019.flac", "task_type": "lexical_detection", "query": "loving", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"loving\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.74, 1.08, "loving"]], "response_mode": "offline", "answer": "[{\"word\":\"loving\",\"start\":0.74,\"end\":1.08}]"} {"utterance_id": "6455-67803-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0028.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.21, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":0.91,\"end\":1.21}]"} {"utterance_id": "8288-274162-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0027.flac", "task_type": "lexical_detection", "query": "louise", "prompt": "Once the audio has finished, identify every instance of the word \"louise\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.61, 1.99, "louise"]], "response_mode": "offline", "answer": "[{\"word\":\"louise\",\"start\":1.61,\"end\":1.99}]"} {"utterance_id": "8173-294714-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0053.flac", "task_type": "lexical_detection", "query": "trustworthy", "prompt": "When the utterance is complete, report all instances of the word \"trustworthy\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.24, "trustworthy"]], "response_mode": "offline", "answer": "[{\"word\":\"trustworthy\",\"start\":0.5,\"end\":1.24}]"} {"utterance_id": "4831-25894-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0010.flac", "task_type": "lexical_detection", "query": "tessa's", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"tessa's\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 1.01, "tessa's"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa's\",\"start\":0.63,\"end\":1.01}]"} {"utterance_id": "6599-38590-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0003.flac", "task_type": "lexical_counting", "query": "give", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"give\" with word/start/end. If none, reply [].", "labels": [[4.45, 4.65, "one"], [11.88, 12.06, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.45,\"end\":4.65},{\"word\":\"two\",\"start\":11.88,\"end\":12.06}]"} {"utterance_id": "1585-157660-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0007.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.93, 2.17, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":1.93,\"end\":2.17}]"} {"utterance_id": "1585-131718-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0030.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 2.02, "one"], [3.55, 4.1, "two"], [8.5, 8.93, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.57,\"end\":2.02},{\"word\":\"two\",\"start\":3.55,\"end\":4.1},{\"word\":\"three\",\"start\":8.5,\"end\":8.93}]"} {"utterance_id": "6123-59186-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0037.flac", "task_type": "lexical_detection", "query": "mournful", "prompt": "When the utterance is complete, report all instances of the word \"mournful\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.83, 3.26, "mournful"]], "response_mode": "offline", "answer": "[{\"word\":\"mournful\",\"start\":2.83,\"end\":3.26}]"} {"utterance_id": "4153-185072-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.13, "allen"]], "response_mode": "offline", "answer": "[{\"word\":\"allen\",\"start\":4.82,\"end\":5.13}]"} {"utterance_id": "8288-274162-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0006.flac", "task_type": "lexical_detection", "query": "pale", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"pale\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.58, "pale"]], "response_mode": "offline", "answer": "[{\"word\":\"pale\",\"start\":1.23,\"end\":1.58}]"} {"utterance_id": "4323-18416-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0025.flac", "task_type": "lexical_detection", "query": "nice", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"nice\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.74, 1.02, "nice"]], "response_mode": "offline", "answer": "[{\"word\":\"nice\",\"start\":0.74,\"end\":1.02}]"} {"utterance_id": "4831-25894-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0003.flac", "task_type": "semantic_detection", "query": "date", "prompt": "You will hear a complete utterance. When it ends, list every instance of a date you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.57, 4.49, "merry christmas"]], "response_mode": "offline", "answer": "[{\"word\":\"merry christmas\",\"start\":3.57,\"end\":4.49}]"} {"utterance_id": "1651-136854-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0010.flac", "task_type": "lexical_detection", "query": "pursued", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"pursued\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.76, 2.38, "pursued"]], "response_mode": "offline", "answer": "[{\"word\":\"pursued\",\"start\":1.76,\"end\":2.38}]"} {"utterance_id": "3915-98647-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0031.flac", "task_type": "lexical_detection", "query": "deceased", "prompt": "After the audio ends, extract all occurrences of the word \"deceased\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.82, 6.42, "deceased"]], "response_mode": "offline", "answer": "[{\"word\":\"deceased\",\"start\":5.82,\"end\":6.42}]"} {"utterance_id": "700-122867-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0002.flac", "task_type": "lexical_detection", "query": "real", "prompt": "After listening to the whole recording, output every instance of the word \"real\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.87, 6.13, "real"]], "response_mode": "offline", "answer": "[{\"word\":\"real\",\"start\":5.87,\"end\":6.13}]"} {"utterance_id": "1255-138279-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0011.flac", "task_type": "lexical_detection", "query": "smaller", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"smaller\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.49, 2.97, "smaller"]], "response_mode": "offline", "answer": "[{\"word\":\"smaller\",\"start\":2.49,\"end\":2.97}]"} {"utterance_id": "4572-64670-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0001.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "After listening to the whole recording, output timed count words for a family relation (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.21, 1.69, "one"], [2.78, 3.23, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.21,\"end\":1.69},{\"word\":\"two\",\"start\":2.78,\"end\":3.23}]"} {"utterance_id": "4153-186223-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0000.flac", "task_type": "lexical_detection", "query": "lap", "prompt": "Once the audio has finished, identify every instance of the word \"lap\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.12, 7.37, "lap"]], "response_mode": "offline", "answer": "[{\"word\":\"lap\",\"start\":7.12,\"end\":7.37}]"} {"utterance_id": "4572-64670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0024.flac", "task_type": "lexical_detection", "query": "grass", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"grass\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.36, 3.85, "grass"]], "response_mode": "offline", "answer": "[{\"word\":\"grass\",\"start\":3.36,\"end\":3.85}]"} {"utterance_id": "6123-59186-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0000.flac", "task_type": "lexical_detection", "query": "temperament", "prompt": "Process the full audio offline. Find every instance of the word \"temperament\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.24, 1.92, "temperament"]], "response_mode": "offline", "answer": "[{\"word\":\"temperament\",\"start\":1.24,\"end\":1.92}]"} {"utterance_id": "116-288045-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.86, 5.32, "jesus"], [6.35, 7.02, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":4.86,\"end\":5.32},{\"word\":\"apollo\",\"start\":6.35,\"end\":7.02}]"} {"utterance_id": "5849-50873-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0038.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "You will hear a complete utterance. When it ends, list every instance of a time of day you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.76, 5.29, "midnight"]], "response_mode": "offline", "answer": "[{\"word\":\"midnight\",\"start\":4.76,\"end\":5.29}]"} {"utterance_id": "700-122868-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.07, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":2.81,\"end\":3.07}]"} {"utterance_id": "6841-88291-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0005.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.2, 4.65, "two hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"two hundred\",\"start\":4.2,\"end\":4.65}]"} {"utterance_id": "8288-274162-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0028.flac", "task_type": "lexical_detection", "query": "suppose", "prompt": "Once the audio has finished, identify every instance of the word \"suppose\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.83, 1.29, "suppose"]], "response_mode": "offline", "answer": "[{\"word\":\"suppose\",\"start\":0.83,\"end\":1.29}]"} {"utterance_id": "1255-90413-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.46, "jocelyn's"]], "response_mode": "offline", "answer": "[{\"word\":\"jocelyn's\",\"start\":1.04,\"end\":1.46}]"} {"utterance_id": "6267-65525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0015.flac", "task_type": "lexical_detection", "query": "interesting", "prompt": "When the utterance is complete, report all instances of the word \"interesting\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.87, 5.6, "interesting"]], "response_mode": "offline", "answer": "[{\"word\":\"interesting\",\"start\":4.87,\"end\":5.6}]"} {"utterance_id": "700-122867-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0036.flac", "task_type": "lexical_detection", "query": "better", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"better\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.61, "better"]], "response_mode": "offline", "answer": "[{\"word\":\"better\",\"start\":1.34,\"end\":1.61}]"} {"utterance_id": "6267-65525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0001.flac", "task_type": "lexical_detection", "query": "lamp", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"lamp\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.61, 3.97, "lamp"]], "response_mode": "offline", "answer": "[{\"word\":\"lamp\",\"start\":3.61,\"end\":3.97}]"} {"utterance_id": "6267-65525-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0038.flac", "task_type": "lexical_detection", "query": "understand", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"understand\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.43, "understand"]], "response_mode": "offline", "answer": "[{\"word\":\"understand\",\"start\":0.82,\"end\":1.43}]"} {"utterance_id": "116-288048-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0032.flac", "task_type": "lexical_detection", "query": "ignorance", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"ignorance\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.13, 3.75, "ignorance"]], "response_mode": "offline", "answer": "[{\"word\":\"ignorance\",\"start\":3.13,\"end\":3.75}]"} {"utterance_id": "6841-88291-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0044.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.51, "calves"]], "response_mode": "offline", "answer": "[{\"word\":\"calves\",\"start\":1.14,\"end\":1.51}]"} {"utterance_id": "4323-55228-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0051.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.56, 2.91, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":2.56,\"end\":2.91}]"} {"utterance_id": "3915-98647-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0023.flac", "task_type": "lexical_detection", "query": "bags", "prompt": "After listening to the whole recording, output every instance of the word \"bags\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.86, 4.19, "bags"]], "response_mode": "offline", "answer": "[{\"word\":\"bags\",\"start\":3.86,\"end\":4.19}]"} {"utterance_id": "1255-90407-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0017.flac", "task_type": "lexical_detection", "query": "except", "prompt": "Once the audio has finished, identify every instance of the word \"except\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.88, 3.51, "except"]], "response_mode": "offline", "answer": "[{\"word\":\"except\",\"start\":2.88,\"end\":3.51}]"} {"utterance_id": "8288-274162-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.88, 4.84, "monsieur d'artagnan"]], "response_mode": "offline", "answer": "[{\"word\":\"monsieur d'artagnan\",\"start\":3.88,\"end\":4.84}]"} {"utterance_id": "8288-274150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0000.flac", "task_type": "lexical_detection", "query": "malicorne", "prompt": "Once the audio has finished, identify every instance of the word \"malicorne\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.52, 3.05, "malicorne"]], "response_mode": "offline", "answer": "[{\"word\":\"malicorne\",\"start\":2.52,\"end\":3.05}]"} {"utterance_id": "4323-18416-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.19, 3.58, "charlotte"], [8.49, 8.85, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"charlotte\",\"start\":3.19,\"end\":3.58},{\"word\":\"phronsie\",\"start\":8.49,\"end\":8.85}]"} {"utterance_id": "4831-29134-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0015.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After the audio ends, extract all occurrences of a country name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 2.21, "united states"]], "response_mode": "offline", "answer": "[{\"word\":\"united states\",\"start\":1.25,\"end\":2.21}]"} {"utterance_id": "1255-74899-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0005.flac", "task_type": "lexical_detection", "query": "required", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"required\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.35, "required"]], "response_mode": "offline", "answer": "[{\"word\":\"required\",\"start\":0.91,\"end\":1.35}]"} {"utterance_id": "7641-96670-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0021.flac", "task_type": "lexical_detection", "query": "seed", "prompt": "After listening to the whole recording, output every instance of the word \"seed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.42, "seed"]], "response_mode": "offline", "answer": "[{\"word\":\"seed\",\"start\":2.07,\"end\":2.42}]"} {"utterance_id": "1630-102884-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0006.flac", "task_type": "lexical_detection", "query": "true", "prompt": "When the utterance is complete, report all instances of the word \"true\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.14, "true"]], "response_mode": "offline", "answer": "[{\"word\":\"true\",\"start\":1.88,\"end\":2.14}]"} {"utterance_id": "6841-88291-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0042.flac", "task_type": "lexical_detection", "query": "calves", "prompt": "After listening to the whole recording, output every instance of the word \"calves\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.69, "calves"]], "response_mode": "offline", "answer": "[{\"word\":\"calves\",\"start\":1.28,\"end\":1.69}]"} {"utterance_id": "700-122866-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0018.flac", "task_type": "lexical_detection", "query": "monday", "prompt": "After the audio ends, extract all occurrences of the word \"monday\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.45, "monday"]], "response_mode": "offline", "answer": "[{\"word\":\"monday\",\"start\":1.14,\"end\":1.45}]"} {"utterance_id": "1686-142278-0097", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0097.flac", "task_type": "lexical_detection", "query": "murmur", "prompt": "After listening to the whole recording, output every instance of the word \"murmur\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.67, "murmur"]], "response_mode": "offline", "answer": "[{\"word\":\"murmur\",\"start\":1.28,\"end\":1.67}]"} {"utterance_id": "6123-59186-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0020.flac", "task_type": "lexical_counting", "query": "part", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"part\" with word/start/end. If none, reply [].", "labels": [[2.7, 3.03, "one"], [4.13, 4.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.7,\"end\":3.03},{\"word\":\"two\",\"start\":4.13,\"end\":4.49}]"} {"utterance_id": "1651-136854-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0013.flac", "task_type": "lexical_detection", "query": "friendship", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"friendship\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.94, 1.44, "friendship"]], "response_mode": "offline", "answer": "[{\"word\":\"friendship\",\"start\":0.94,\"end\":1.44}]"} {"utterance_id": "4515-11057-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0013.flac", "task_type": "lexical_detection", "query": "poor", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"poor\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.2, 2.78, "poor"]], "response_mode": "offline", "answer": "[{\"word\":\"poor\",\"start\":2.2,\"end\":2.78}]"} {"utterance_id": "7641-96670-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0017.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After listening to the whole recording, output every instance of a color name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.32, 2.62, "golden"]], "response_mode": "offline", "answer": "[{\"word\":\"golden\",\"start\":2.32,\"end\":2.62}]"} {"utterance_id": "4570-14911-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0008.flac", "task_type": "lexical_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of the word \"country\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.34, 2.77, "country"]], "response_mode": "offline", "answer": "[{\"word\":\"country\",\"start\":2.34,\"end\":2.77}]"} {"utterance_id": "6455-67804-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0030.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.6, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":1.22,\"end\":1.6}]"} {"utterance_id": "1686-142278-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0059.flac", "task_type": "lexical_detection", "query": "dislike", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"dislike\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.77, 1.25, "dislike"]], "response_mode": "offline", "answer": "[{\"word\":\"dislike\",\"start\":0.77,\"end\":1.25}]"} {"utterance_id": "6467-97061-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0003.flac", "task_type": "lexical_detection", "query": "fearful", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"fearful\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.64, 3.08, "fearful"]], "response_mode": "offline", "answer": "[{\"word\":\"fearful\",\"start\":2.64,\"end\":3.08}]"} {"utterance_id": "1701-141760-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.37, 1.78, "parents"], [10.58, 10.91, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"parents\",\"start\":1.37,\"end\":1.78},{\"word\":\"son\",\"start\":10.58,\"end\":10.91}]"} {"utterance_id": "1650-173552-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0005.flac", "task_type": "lexical_detection", "query": "kremlin", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"kremlin\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.48, 2.15, "kremlin"], [3.98, 4.63, "kremlin"]], "response_mode": "offline", "answer": "[{\"word\":\"kremlin\",\"start\":1.48,\"end\":2.15},{\"word\":\"kremlin\",\"start\":3.98,\"end\":4.63}]"} {"utterance_id": "8173-294714-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0039.flac", "task_type": "lexical_detection", "query": "plain", "prompt": "Once the audio has finished, identify every instance of the word \"plain\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.08, 2.37, "plain"]], "response_mode": "offline", "answer": "[{\"word\":\"plain\",\"start\":2.08,\"end\":2.37}]"} {"utterance_id": "6123-59186-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.46, 7.0, "poet"]], "response_mode": "offline", "answer": "[{\"word\":\"poet\",\"start\":6.46,\"end\":7.0}]"} {"utterance_id": "4153-186222-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0031.flac", "task_type": "lexical_detection", "query": "pause", "prompt": "After listening to the whole recording, output every instance of the word \"pause\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.61, 1.94, "pause"]], "response_mode": "offline", "answer": "[{\"word\":\"pause\",\"start\":1.61,\"end\":1.94}]"} {"utterance_id": "3663-172528-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0013.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.93, 5.48, "water"]], "response_mode": "offline", "answer": "[{\"word\":\"water\",\"start\":4.93,\"end\":5.48}]"} {"utterance_id": "116-288048-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0025.flac", "task_type": "lexical_detection", "query": "wars", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wars\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[16.55, 17.08, "wars"]], "response_mode": "offline", "answer": "[{\"word\":\"wars\",\"start\":16.55,\"end\":17.08}]"} {"utterance_id": "7601-291468-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0002.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of a country name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.32, 3.93, "countries"], [12.94, 13.58, "nations"]], "response_mode": "offline", "answer": "[{\"word\":\"countries\",\"start\":3.32,\"end\":3.93},{\"word\":\"nations\",\"start\":12.94,\"end\":13.58}]"} {"utterance_id": "1701-141760-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0018.flac", "task_type": "lexical_detection", "query": "pillow", "prompt": "When the utterance is complete, report all instances of the word \"pillow\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.26, 3.66, "pillow"]], "response_mode": "offline", "answer": "[{\"word\":\"pillow\",\"start\":3.26,\"end\":3.66}]"} {"utterance_id": "5849-50964-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0012.flac", "task_type": "lexical_detection", "query": "downy", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"downy\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.61, 4.14, "downy"]], "response_mode": "offline", "answer": "[{\"word\":\"downy\",\"start\":3.61,\"end\":4.14}]"} {"utterance_id": "3660-172182-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0022.flac", "task_type": "lexical_detection", "query": "dismounted", "prompt": "When the utterance is complete, report all instances of the word \"dismounted\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.95, "dismounted"]], "response_mode": "offline", "answer": "[{\"word\":\"dismounted\",\"start\":1.28,\"end\":1.95}]"} {"utterance_id": "4323-13259-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0024.flac", "task_type": "lexical_detection", "query": "queer", "prompt": "When the utterance is complete, report all instances of the word \"queer\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.12, "queer"]], "response_mode": "offline", "answer": "[{\"word\":\"queer\",\"start\":1.88,\"end\":2.12}]"} {"utterance_id": "5543-27761-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0063.flac", "task_type": "lexical_detection", "query": "eagerly", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"eagerly\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.0, "eagerly"]], "response_mode": "offline", "answer": "[{\"word\":\"eagerly\",\"start\":0.52,\"end\":1.0}]"} {"utterance_id": "1585-131718-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0031.flac", "task_type": "semantic_counting", "query": "ordinal", "prompt": "You will hear a complete utterance. When it ends, produce a running count of an ordinal number: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[7.96, 8.28, "one"], [9.68, 10.06, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.96,\"end\":8.28},{\"word\":\"two\",\"start\":9.68,\"end\":10.06}]"} {"utterance_id": "4570-24733-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0001.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[11.52, 11.9, "one"], [26.97, 27.57, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":11.52,\"end\":11.9},{\"word\":\"two\",\"start\":26.97,\"end\":27.57}]"} {"utterance_id": "1630-141772-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.93, 10.31, "neck"]], "response_mode": "offline", "answer": "[{\"word\":\"neck\",\"start\":9.93,\"end\":10.31}]"} {"utterance_id": "4831-25894-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0023.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.4, 0.74, "one"], [2.69, 3.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.4,\"end\":0.74},{\"word\":\"two\",\"start\":2.69,\"end\":3.13}]"} {"utterance_id": "1585-131718-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0008.flac", "task_type": "lexical_counting", "query": "companies", "prompt": "When the utterance is complete, report how the word \"companies\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 3.05, "one"], [6.16, 6.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.48,\"end\":3.05},{\"word\":\"two\",\"start\":6.16,\"end\":6.72}]"} {"utterance_id": "4323-13259-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0025.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a number with word/start/end. If none, reply [].", "labels": [[0.56, 0.7, "one"], [2.62, 2.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.56,\"end\":0.7},{\"word\":\"two\",\"start\":2.62,\"end\":2.78}]"} {"utterance_id": "4572-112383-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0008.flac", "task_type": "lexical_detection", "query": "deadly", "prompt": "After listening to the whole recording, output every instance of the word \"deadly\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.75, "deadly"]], "response_mode": "offline", "answer": "[{\"word\":\"deadly\",\"start\":1.36,\"end\":1.75}]"} {"utterance_id": "1255-74899-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0010.flac", "task_type": "lexical_detection", "query": "merit", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"merit\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.01, "merit"]], "response_mode": "offline", "answer": "[{\"word\":\"merit\",\"start\":1.68,\"end\":2.01}]"} {"utterance_id": "116-288045-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.83, 4.36, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"voice\",\"start\":3.83,\"end\":4.36}]"} {"utterance_id": "6267-65525-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0052.flac", "task_type": "lexical_detection", "query": "manly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"manly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.56, "manly"]], "response_mode": "offline", "answer": "[{\"word\":\"manly\",\"start\":2.0,\"end\":2.56}]"} {"utterance_id": "6267-65525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0001.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.84, 2.24, "one"], [4.36, 4.79, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.84,\"end\":2.24},{\"word\":\"two\",\"start\":4.36,\"end\":4.79}]"} {"utterance_id": "3660-6517-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0000.flac", "task_type": "lexical_detection", "query": "period", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"period\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.32, "period"]], "response_mode": "offline", "answer": "[{\"word\":\"period\",\"start\":2.93,\"end\":3.32}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.51, 7.17, "athens"]], "response_mode": "offline", "answer": "[{\"word\":\"athens\",\"start\":6.51,\"end\":7.17}]"} {"utterance_id": "4323-18416-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0020.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Process the full audio offline. For every occurrence of a family relation, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 1.98, "one"], [7.81, 8.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.67,\"end\":1.98},{\"word\":\"two\",\"start\":7.81,\"end\":8.36}]"} {"utterance_id": "4515-11057-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0025.flac", "task_type": "lexical_detection", "query": "paused", "prompt": "When the utterance is complete, report all instances of the word \"paused\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.27, "paused"]], "response_mode": "offline", "answer": "[{\"word\":\"paused\",\"start\":0.73,\"end\":1.27}]"} {"utterance_id": "116-288047-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0015.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.96, 2.42, "fairies"]], "response_mode": "offline", "answer": "[{\"word\":\"fairies\",\"start\":1.96,\"end\":2.42}]"} {"utterance_id": "3660-6517-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0007.flac", "task_type": "lexical_counting", "query": "space", "prompt": "Listen to the entire clip, then report the count sequence for the word \"space\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.81, 4.22, "one"], [6.36, 6.74, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.81,\"end\":4.22},{\"word\":\"two\",\"start\":6.36,\"end\":6.74}]"} {"utterance_id": "6267-65525-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0039.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.55, 1.05, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":0.55,\"end\":1.05}]"} {"utterance_id": "6467-97061-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.21, 6.51, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":6.21,\"end\":6.51}]"} {"utterance_id": "8254-115543-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0021.flac", "task_type": "lexical_counting", "query": "rajah", "prompt": "When the utterance is complete, report how the word \"rajah\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.85, "one"], [5.18, 5.67, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.38,\"end\":2.85},{\"word\":\"two\",\"start\":5.18,\"end\":5.67}]"} {"utterance_id": "4323-18416-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0029.flac", "task_type": "lexical_detection", "query": "threw", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"threw\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.93, 2.07, "threw"]], "response_mode": "offline", "answer": "[{\"word\":\"threw\",\"start\":1.93,\"end\":2.07}]"} {"utterance_id": "5849-50962-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.65, 2.02, "hal"]], "response_mode": "offline", "answer": "[{\"word\":\"hal\",\"start\":1.65,\"end\":2.02}]"} {"utterance_id": "5849-50873-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0035.flac", "task_type": "lexical_detection", "query": "connecticut", "prompt": "Once the audio has finished, identify every instance of the word \"connecticut\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.1, 2.69, "connecticut"]], "response_mode": "offline", "answer": "[{\"word\":\"connecticut\",\"start\":2.1,\"end\":2.69}]"} {"utterance_id": "6841-88291-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0030.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.27, 0.57, "calf"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":0.27,\"end\":0.57}]"} {"utterance_id": "8173-294714-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0006.flac", "task_type": "lexical_detection", "query": "mouth", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mouth\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.25, 1.64, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":1.25,\"end\":1.64}]"} {"utterance_id": "5849-50964-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.3, 0.63, "dorothy"], [2.76, 3.38, "cinderella's"]], "response_mode": "offline", "answer": "[{\"word\":\"dorothy\",\"start\":0.3,\"end\":0.63},{\"word\":\"cinderella's\",\"start\":2.76,\"end\":3.38}]"} {"utterance_id": "6467-56885-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.05, 2.31, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.05,\"end\":2.31}]"} {"utterance_id": "8254-115543-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0015.flac", "task_type": "lexical_detection", "query": "eaten", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"eaten\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.88, 1.17, "eaten"]], "response_mode": "offline", "answer": "[{\"word\":\"eaten\",\"start\":0.88,\"end\":1.17}]"} {"utterance_id": "8288-274162-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.61, 1.99, "louise"]], "response_mode": "offline", "answer": "[{\"word\":\"louise\",\"start\":1.61,\"end\":1.99}]"} {"utterance_id": "8254-84205-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0064.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 1.79, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.53,\"end\":1.79}]"} {"utterance_id": "3663-172528-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.14, "maid"]], "response_mode": "offline", "answer": "[{\"word\":\"maid\",\"start\":0.76,\"end\":1.14}]"} {"utterance_id": "7601-175351-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.16, 2.95, "hatteras"]], "response_mode": "offline", "answer": "[{\"word\":\"hatteras\",\"start\":2.16,\"end\":2.95}]"} {"utterance_id": "1701-141760-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0012.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.47, 4.78, "first"], [12.48, 12.78, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":4.47,\"end\":4.78},{\"word\":\"first\",\"start\":12.48,\"end\":12.78}]"} {"utterance_id": "7641-96684-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0033.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.99, 7.27, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.99,\"end\":7.27}]"} {"utterance_id": "3660-172182-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0023.flac", "task_type": "lexical_counting", "query": "heaven", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"heaven\" (word, start, end). If none, reply [].", "labels": [[0.5, 0.78, "one"], [3.13, 3.4, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":0.78},{\"word\":\"two\",\"start\":3.13,\"end\":3.4}]"} {"utterance_id": "1255-90413-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0022.flac", "task_type": "lexical_detection", "query": "dining", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"dining\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.77, 4.07, "dining"]], "response_mode": "offline", "answer": "[{\"word\":\"dining\",\"start\":3.77,\"end\":4.07}]"} {"utterance_id": "4515-11057-0097", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0097.flac", "task_type": "lexical_detection", "query": "happens", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"happens\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.99, "happens"]], "response_mode": "offline", "answer": "[{\"word\":\"happens\",\"start\":1.48,\"end\":1.99}]"} {"utterance_id": "6841-88291-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0010.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.41, 0.7, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.41,\"end\":0.7}]"} {"utterance_id": "2506-11278-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0010.flac", "task_type": "lexical_detection", "query": "lady", "prompt": "Once the audio has finished, identify every instance of the word \"lady\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.27, 4.86, "lady"]], "response_mode": "offline", "answer": "[{\"word\":\"lady\",\"start\":4.27,\"end\":4.86}]"} {"utterance_id": "6841-88294-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0016.flac", "task_type": "lexical_counting", "query": "didn't", "prompt": "After listening to the whole recording, output timed count words for the word \"didn't\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.66, 0.91, "one"], [2.19, 2.4, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.66,\"end\":0.91},{\"word\":\"two\",\"start\":2.19,\"end\":2.4}]"} {"utterance_id": "4570-14911-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0010.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.97, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":0.53,\"end\":0.97}]"} {"utterance_id": "1686-142278-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0014.flac", "task_type": "lexical_detection", "query": "bear", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"bear\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.47, "bear"]], "response_mode": "offline", "answer": "[{\"word\":\"bear\",\"start\":1.3,\"end\":1.47}]"} {"utterance_id": "6123-59186-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0001.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "When the utterance is complete, report all instances of a weather term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.48, 5.94, "storm"]], "response_mode": "offline", "answer": "[{\"word\":\"storm\",\"start\":5.48,\"end\":5.94}]"} {"utterance_id": "8254-115543-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0007.flac", "task_type": "lexical_detection", "query": "shouted", "prompt": "Process the full audio offline. Find every instance of the word \"shouted\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.75, "shouted"]], "response_mode": "offline", "answer": "[{\"word\":\"shouted\",\"start\":1.39,\"end\":1.75}]"} {"utterance_id": "4515-11057-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0061.flac", "task_type": "lexical_detection", "query": "refuses", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"refuses\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 2.62, "refuses"]], "response_mode": "offline", "answer": "[{\"word\":\"refuses\",\"start\":1.71,\"end\":2.62}]"} {"utterance_id": "6123-59150-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.35, 5.26, "anna karenina"]], "response_mode": "offline", "answer": "[{\"word\":\"anna karenina\",\"start\":4.35,\"end\":5.26}]"} {"utterance_id": "1650-173551-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0005.flac", "task_type": "lexical_counting", "query": "head", "prompt": "Once the audio has finished, list each occurrence of the word \"head\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[5.45, 5.75, "one"], [8.77, 9.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.45,\"end\":5.75},{\"word\":\"two\",\"start\":8.77,\"end\":9.03}]"} {"utterance_id": "7641-96670-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0017.flac", "task_type": "lexical_detection", "query": "fleece", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fleece\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.62, 3.0, "fleece"]], "response_mode": "offline", "answer": "[{\"word\":\"fleece\",\"start\":2.62,\"end\":3.0}]"} {"utterance_id": "700-122867-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0031.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.28, 2.62, "nose"]], "response_mode": "offline", "answer": "[{\"word\":\"nose\",\"start\":2.28,\"end\":2.62}]"} {"utterance_id": "8173-294714-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0017.flac", "task_type": "lexical_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of the word \"food\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.03, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":1.79,\"end\":2.03}]"} {"utterance_id": "5849-50873-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.42, 2.72, "feet"], [4.57, 4.92, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":2.42,\"end\":2.72},{\"word\":\"hand\",\"start\":4.57,\"end\":4.92}]"} {"utterance_id": "2506-11278-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0035.flac", "task_type": "lexical_detection", "query": "sweet", "prompt": "After listening to the whole recording, output every instance of the word \"sweet\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 1.98, "sweet"]], "response_mode": "offline", "answer": "[{\"word\":\"sweet\",\"start\":1.71,\"end\":1.98}]"} {"utterance_id": "6267-65525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0007.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[4.38, 4.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":4.38,\"end\":4.63}]"} {"utterance_id": "3663-172528-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.64, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":4.23,\"end\":4.64}]"} {"utterance_id": "116-288048-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0031.flac", "task_type": "lexical_detection", "query": "mean", "prompt": "Once the audio has finished, identify every instance of the word \"mean\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.81, 1.18, "mean"]], "response_mode": "offline", "answer": "[{\"word\":\"mean\",\"start\":0.81,\"end\":1.18}]"} {"utterance_id": "1650-157641-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.7, 1.65, "missus quilp"]], "response_mode": "offline", "answer": "[{\"word\":\"missus quilp\",\"start\":0.7,\"end\":1.65}]"} {"utterance_id": "6455-66379-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0019.flac", "task_type": "lexical_detection", "query": "stop", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"stop\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.05, "stop"]], "response_mode": "offline", "answer": "[{\"word\":\"stop\",\"start\":1.68,\"end\":2.05}]"} {"utterance_id": "3915-57461-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.33, 2.79, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":2.33,\"end\":2.79}]"} {"utterance_id": "1255-90407-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0028.flac", "task_type": "lexical_detection", "query": "conscious", "prompt": "When the utterance is complete, report all instances of the word \"conscious\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.8, "conscious"]], "response_mode": "offline", "answer": "[{\"word\":\"conscious\",\"start\":1.36,\"end\":1.8}]"} {"utterance_id": "4153-186223-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.07, "duncan"]], "response_mode": "offline", "answer": "[{\"word\":\"duncan\",\"start\":0.65,\"end\":1.07}]"} {"utterance_id": "8254-115543-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0038.flac", "task_type": "lexical_detection", "query": "beggars", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"beggars\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.97, "beggars"]], "response_mode": "offline", "answer": "[{\"word\":\"beggars\",\"start\":1.25,\"end\":1.97}]"} {"utterance_id": "8173-294714-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0030.flac", "task_type": "lexical_detection", "query": "something", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"something\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.91, "something"]], "response_mode": "offline", "answer": "[{\"word\":\"something\",\"start\":0.48,\"end\":0.91}]"} {"utterance_id": "3660-6517-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.78, 3.02, "thirty"]], "response_mode": "offline", "answer": "[{\"word\":\"thirty\",\"start\":2.78,\"end\":3.02}]"} {"utterance_id": "1686-142278-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.69, 4.02, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":3.69,\"end\":4.02}]"} {"utterance_id": "4153-186222-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0003.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.3, 2.77, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":2.3,\"end\":2.77}]"} {"utterance_id": "1686-142278-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0070.flac", "task_type": "lexical_detection", "query": "influenced", "prompt": "When the utterance is complete, report all instances of the word \"influenced\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.78, 3.28, "influenced"]], "response_mode": "offline", "answer": "[{\"word\":\"influenced\",\"start\":2.78,\"end\":3.28}]"} {"utterance_id": "8254-115543-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0025.flac", "task_type": "lexical_detection", "query": "stick", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"stick\" with fields word, start, end (seconds). If none, reply [].", "labels": [[7.02, 7.45, "stick"]], "response_mode": "offline", "answer": "[{\"word\":\"stick\",\"start\":7.02,\"end\":7.45}]"} {"utterance_id": "4515-11057-0106", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0106.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After listening to the whole recording, output every instance of a plant name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.35, 4.69, "tree"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":4.35,\"end\":4.69}]"} {"utterance_id": "4570-56594-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0012.flac", "task_type": "lexical_detection", "query": "opening", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"opening\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.54, 7.99, "opening"]], "response_mode": "offline", "answer": "[{\"word\":\"opening\",\"start\":7.54,\"end\":7.99}]"} {"utterance_id": "4323-55228-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0027.flac", "task_type": "lexical_detection", "query": "instead", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"instead\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.97, 3.28, "instead"]], "response_mode": "offline", "answer": "[{\"word\":\"instead\",\"start\":2.97,\"end\":3.28}]"} {"utterance_id": "5849-50873-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0031.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.6, "soldiers"]], "response_mode": "offline", "answer": "[{\"word\":\"soldiers\",\"start\":1.15,\"end\":1.6}]"} {"utterance_id": "3660-172182-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.72, 1.09, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":0.72,\"end\":1.09}]"} {"utterance_id": "8254-115543-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0023.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.58, 9.09, "rajah"]], "response_mode": "offline", "answer": "[{\"word\":\"rajah\",\"start\":8.58,\"end\":9.09}]"} {"utterance_id": "1686-142278-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0042.flac", "task_type": "lexical_detection", "query": "resign", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"resign\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.43, 14.81, "resign"]], "response_mode": "offline", "answer": "[{\"word\":\"resign\",\"start\":14.43,\"end\":14.81}]"} {"utterance_id": "8254-115543-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0021.flac", "task_type": "lexical_detection", "query": "rajah", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"rajah\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.85, "rajah"], [5.18, 5.67, "rajah"]], "response_mode": "offline", "answer": "[{\"word\":\"rajah\",\"start\":2.38,\"end\":2.85},{\"word\":\"rajah\",\"start\":5.18,\"end\":5.67}]"} {"utterance_id": "8254-84205-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0068.flac", "task_type": "lexical_detection", "query": "let", "prompt": "After listening to the whole recording, output every instance of the word \"let\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.16, "let"]], "response_mode": "offline", "answer": "[{\"word\":\"let\",\"start\":0.99,\"end\":1.16}]"} {"utterance_id": "4570-14911-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0014.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Once the audio has finished, list each occurrence of a body part as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.53, 1.06, "one"], [6.3, 6.88, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.53,\"end\":1.06},{\"word\":\"two\",\"start\":6.3,\"end\":6.88}]"} {"utterance_id": "1686-142278-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0022.flac", "task_type": "lexical_detection", "query": "answer", "prompt": "When the utterance is complete, report all instances of the word \"answer\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.74, "answer"]], "response_mode": "offline", "answer": "[{\"word\":\"answer\",\"start\":0.44,\"end\":0.74}]"} {"utterance_id": "6841-88291-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0005.flac", "task_type": "lexical_detection", "query": "corrals", "prompt": "After the audio ends, extract all occurrences of the word \"corrals\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.04, 3.53, "corrals"]], "response_mode": "offline", "answer": "[{\"word\":\"corrals\",\"start\":3.04,\"end\":3.53}]"} {"utterance_id": "1630-96099-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0011.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Process the full audio offline. Find every instance of a country name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.07, 10.66, "italian"]], "response_mode": "offline", "answer": "[{\"word\":\"italian\",\"start\":10.07,\"end\":10.66}]"} {"utterance_id": "4572-112381-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0009.flac", "task_type": "lexical_detection", "query": "task", "prompt": "Process the full audio offline. Find every instance of the word \"task\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.51, "task"]], "response_mode": "offline", "answer": "[{\"word\":\"task\",\"start\":1.06,\"end\":1.51}]"} {"utterance_id": "6455-67803-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0035.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.74, 0.97, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":0.74,\"end\":0.97}]"} {"utterance_id": "1651-136854-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0020.flac", "task_type": "lexical_detection", "query": "authority", "prompt": "After listening to the whole recording, output every instance of the word \"authority\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.99, "authority"]], "response_mode": "offline", "answer": "[{\"word\":\"authority\",\"start\":1.33,\"end\":1.99}]"} {"utterance_id": "6467-97061-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0015.flac", "task_type": "lexical_detection", "query": "metal", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"metal\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.25, 9.56, "metal"]], "response_mode": "offline", "answer": "[{\"word\":\"metal\",\"start\":9.25,\"end\":9.56}]"} {"utterance_id": "6267-65525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0007.flac", "task_type": "lexical_detection", "query": "kitchen", "prompt": "Process the full audio offline. Find every instance of the word \"kitchen\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 2.87, "kitchen"]], "response_mode": "offline", "answer": "[{\"word\":\"kitchen\",\"start\":2.43,\"end\":2.87}]"} {"utterance_id": "8254-115543-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0003.flac", "task_type": "lexical_detection", "query": "procession", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"procession\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.65, "procession"]], "response_mode": "offline", "answer": "[{\"word\":\"procession\",\"start\":1.09,\"end\":1.65}]"} {"utterance_id": "7697-105815-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0034.flac", "task_type": "lexical_detection", "query": "climb", "prompt": "When the utterance is complete, report all instances of the word \"climb\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.45, "climb"]], "response_mode": "offline", "answer": "[{\"word\":\"climb\",\"start\":2.08,\"end\":2.45}]"} {"utterance_id": "8254-84205-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0022.flac", "task_type": "lexical_detection", "query": "serve", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"serve\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.08, "serve"]], "response_mode": "offline", "answer": "[{\"word\":\"serve\",\"start\":1.74,\"end\":2.08}]"} {"utterance_id": "1255-74899-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0012.flac", "task_type": "lexical_detection", "query": "foremost", "prompt": "Process the full audio offline. Find every instance of the word \"foremost\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.66, "foremost"]], "response_mode": "offline", "answer": "[{\"word\":\"foremost\",\"start\":1.18,\"end\":1.66}]"} {"utterance_id": "116-288045-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.38, 7.91, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":7.38,\"end\":7.91}]"} {"utterance_id": "1686-142278-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0058.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.54, 3.03, "dobson"]], "response_mode": "offline", "answer": "[{\"word\":\"dobson\",\"start\":2.54,\"end\":3.03}]"} {"utterance_id": "4570-102353-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0004.flac", "task_type": "lexical_counting", "query": "partly", "prompt": "Listen to the full audio. Count every occurrence of the word \"partly\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.41, 3.79, "one"], [8.44, 8.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.41,\"end\":3.79},{\"word\":\"two\",\"start\":8.44,\"end\":8.78}]"} {"utterance_id": "6123-59150-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.84, 3.47, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":2.84,\"end\":3.47}]"} {"utterance_id": "1585-131718-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0010.flac", "task_type": "lexical_detection", "query": "hence", "prompt": "After the audio ends, extract all occurrences of the word \"hence\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.75, "hence"]], "response_mode": "offline", "answer": "[{\"word\":\"hence\",\"start\":0.31,\"end\":0.75}]"} {"utterance_id": "5849-50962-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0003.flac", "task_type": "lexical_detection", "query": "cable", "prompt": "Process the full audio offline. Find every instance of the word \"cable\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.12, 4.61, "cable"]], "response_mode": "offline", "answer": "[{\"word\":\"cable\",\"start\":4.12,\"end\":4.61}]"} {"utterance_id": "4570-14911-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 2.75, "foot"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":2.43,\"end\":2.75}]"} {"utterance_id": "3915-98647-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0014.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.59, 3.02, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":2.59,\"end\":3.02}]"} {"utterance_id": "4570-56594-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0009.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.73, 6.12, "preacher"]], "response_mode": "offline", "answer": "[{\"word\":\"preacher\",\"start\":5.73,\"end\":6.12}]"} {"utterance_id": "6267-53049-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.05, 2.88, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":2.05,\"end\":2.88}]"} {"utterance_id": "6841-88291-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0039.flac", "task_type": "lexical_detection", "query": "branding", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"branding\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.74, 1.13, "branding"]], "response_mode": "offline", "answer": "[{\"word\":\"branding\",\"start\":0.74,\"end\":1.13}]"} {"utterance_id": "1651-136854-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0025.flac", "task_type": "lexical_detection", "query": "greatest", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"greatest\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.9, "greatest"]], "response_mode": "offline", "answer": "[{\"word\":\"greatest\",\"start\":0.44,\"end\":0.9}]"} {"utterance_id": "4323-55228-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0012.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Process the full audio offline. For every occurrence of a body part, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 2.49, "one"], [4.78, 5.0, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.3,\"end\":2.49},{\"word\":\"two\",\"start\":4.78,\"end\":5.0}]"} {"utterance_id": "3663-172005-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0000.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Do not answer until the audio ends. Then list each instance of a time of day as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.71, 2.16, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.71,\"end\":2.16}]"} {"utterance_id": "7641-96252-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0011.flac", "task_type": "lexical_detection", "query": "gives", "prompt": "Once the audio has finished, identify every instance of the word \"gives\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.6, 9.86, "gives"]], "response_mode": "offline", "answer": "[{\"word\":\"gives\",\"start\":9.6,\"end\":9.86}]"} {"utterance_id": "1255-90407-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0016.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.42, 2.84, "budmouth"]], "response_mode": "offline", "answer": "[{\"word\":\"budmouth\",\"start\":2.42,\"end\":2.84}]"} {"utterance_id": "3663-172005-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0005.flac", "task_type": "lexical_detection", "query": "rapidly", "prompt": "When the utterance is complete, report all instances of the word \"rapidly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.24, 4.89, "rapidly"]], "response_mode": "offline", "answer": "[{\"word\":\"rapidly\",\"start\":4.24,\"end\":4.89}]"} {"utterance_id": "6841-88291-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0011.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.57, "cattle"]], "response_mode": "offline", "answer": "[{\"word\":\"cattle\",\"start\":3.1,\"end\":3.57}]"} {"utterance_id": "1686-142278-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0055.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Once the audio has finished, identify every instance of a disease name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.78, 8.39, "depression"]], "response_mode": "offline", "answer": "[{\"word\":\"depression\",\"start\":7.78,\"end\":8.39}]"} {"utterance_id": "6267-53049-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0015.flac", "task_type": "lexical_detection", "query": "new", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"new\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.53, 2.82, "new"]], "response_mode": "offline", "answer": "[{\"word\":\"new\",\"start\":2.53,\"end\":2.82}]"} {"utterance_id": "1630-73710-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0021.flac", "task_type": "lexical_detection", "query": "meet", "prompt": "After listening to the whole recording, output every instance of the word \"meet\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.11, "meet"]], "response_mode": "offline", "answer": "[{\"word\":\"meet\",\"start\":0.85,\"end\":1.11}]"} {"utterance_id": "4323-55228-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0035.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Listen to the full audio. After it finishes, find every instance of a date. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.91, 2.34, "morrow"]], "response_mode": "offline", "answer": "[{\"word\":\"morrow\",\"start\":1.91,\"end\":2.34}]"} {"utterance_id": "8173-294714-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0047.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Do not answer until the audio ends. Then list each instance of a money or currency term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.23, 2.53, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":2.23,\"end\":2.53}]"} {"utterance_id": "1650-167613-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.9, 2.37, "pierre"], [4.27, 5.04, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":1.9,\"end\":2.37},{\"word\":\"prince andrew\",\"start\":4.27,\"end\":5.04}]"} {"utterance_id": "4153-185072-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0006.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.24, 0.81, "one"], [3.9, 4.39, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.24,\"end\":0.81},{\"word\":\"two\",\"start\":3.9,\"end\":4.39}]"} {"utterance_id": "6455-66379-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.07, "ten"]], "response_mode": "offline", "answer": "[{\"word\":\"ten\",\"start\":0.86,\"end\":1.07}]"} {"utterance_id": "1686-142278-0090", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0090.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.75, "maria"]], "response_mode": "offline", "answer": "[{\"word\":\"maria\",\"start\":1.2,\"end\":1.75}]"} {"utterance_id": "116-288046-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.05, 8.52, "brow"]], "response_mode": "offline", "answer": "[{\"word\":\"brow\",\"start\":8.05,\"end\":8.52}]"} {"utterance_id": "4572-112381-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0011.flac", "task_type": "lexical_detection", "query": "hopes", "prompt": "After listening to the whole recording, output every instance of the word \"hopes\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.16, 4.71, "hopes"]], "response_mode": "offline", "answer": "[{\"word\":\"hopes\",\"start\":4.16,\"end\":4.71}]"} {"utterance_id": "1701-141760-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0011.flac", "task_type": "lexical_detection", "query": "notwithstanding", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"notwithstanding\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 1.23, "notwithstanding"]], "response_mode": "offline", "answer": "[{\"word\":\"notwithstanding\",\"start\":0.51,\"end\":1.23}]"} {"utterance_id": "8254-84205-0076", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0076.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.62, 4.92, "scout"]], "response_mode": "offline", "answer": "[{\"word\":\"scout\",\"start\":4.62,\"end\":4.92}]"} {"utterance_id": "7697-105815-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0053.flac", "task_type": "lexical_detection", "query": "told", "prompt": "Once the audio has finished, identify every instance of the word \"told\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.87, 1.12, "told"]], "response_mode": "offline", "answer": "[{\"word\":\"told\",\"start\":0.87,\"end\":1.12}]"} {"utterance_id": "7641-96670-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0005.flac", "task_type": "lexical_detection", "query": "sons", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sons\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.1, 4.57, "sons"]], "response_mode": "offline", "answer": "[{\"word\":\"sons\",\"start\":4.1,\"end\":4.57}]"} {"utterance_id": "6267-65525-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0032.flac", "task_type": "lexical_detection", "query": "woman", "prompt": "Once the audio has finished, identify every instance of the word \"woman\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.55, 0.89, "woman"]], "response_mode": "offline", "answer": "[{\"word\":\"woman\",\"start\":0.55,\"end\":0.89}]"} {"utterance_id": "6599-38591-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.84, "seven"], [3.79, 3.96, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"seven\",\"start\":0.46,\"end\":0.84},{\"word\":\"two\",\"start\":3.79,\"end\":3.96}]"} {"utterance_id": "4323-55228-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.61, 1.01, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":0.61,\"end\":1.01}]"} {"utterance_id": "1255-74899-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0008.flac", "task_type": "lexical_detection", "query": "tailor", "prompt": "When the utterance is complete, report all instances of the word \"tailor\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.01, "tailor"]], "response_mode": "offline", "answer": "[{\"word\":\"tailor\",\"start\":0.7,\"end\":1.01}]"} {"utterance_id": "4515-11057-0074", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0074.flac", "task_type": "lexical_detection", "query": "yourself", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"yourself\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.92, 3.44, "yourself"]], "response_mode": "offline", "answer": "[{\"word\":\"yourself\",\"start\":2.92,\"end\":3.44}]"} {"utterance_id": "6467-62797-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.86, 5.29, "insect"]], "response_mode": "offline", "answer": "[{\"word\":\"insect\",\"start\":4.86,\"end\":5.29}]"} {"utterance_id": "7641-96684-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.55, "alcestis"]], "response_mode": "offline", "answer": "[{\"word\":\"alcestis\",\"start\":0.91,\"end\":1.55}]"} {"utterance_id": "6123-59150-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0010.flac", "task_type": "lexical_detection", "query": "talk", "prompt": "After the audio ends, extract all occurrences of the word \"talk\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.3, "talk"]], "response_mode": "offline", "answer": "[{\"word\":\"talk\",\"start\":0.88,\"end\":1.3}]"} {"utterance_id": "8254-84205-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0064.flac", "task_type": "lexical_detection", "query": "sharply", "prompt": "Process the full audio offline. Find every instance of the word \"sharply\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.25, "sharply"]], "response_mode": "offline", "answer": "[{\"word\":\"sharply\",\"start\":1.79,\"end\":2.25}]"} {"utterance_id": "7697-105815-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0045.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.12, 2.52, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":2.12,\"end\":2.52}]"} {"utterance_id": "5849-50873-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0018.flac", "task_type": "lexical_detection", "query": "setting", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"setting\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.07, 1.51, "setting"]], "response_mode": "offline", "answer": "[{\"word\":\"setting\",\"start\":1.07,\"end\":1.51}]"} {"utterance_id": "4323-55228-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0033.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.73, 2.07, "one"], [3.11, 3.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.73,\"end\":2.07},{\"word\":\"two\",\"start\":3.11,\"end\":3.46}]"} {"utterance_id": "700-122868-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0013.flac", "task_type": "lexical_detection", "query": "occasion", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"occasion\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.75, "occasion"]], "response_mode": "offline", "answer": "[{\"word\":\"occasion\",\"start\":1.2,\"end\":1.75}]"} {"utterance_id": "7697-105815-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0043.flac", "task_type": "lexical_detection", "query": "seen", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"seen\" with fields word, start, end (seconds). If none, reply [].", "labels": [[8.97, 9.23, "seen"]], "response_mode": "offline", "answer": "[{\"word\":\"seen\",\"start\":8.97,\"end\":9.23}]"} {"utterance_id": "1686-142278-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0098.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.18, 3.7, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"daughter\",\"start\":3.18,\"end\":3.7}]"} {"utterance_id": "3915-57461-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0019.flac", "task_type": "lexical_detection", "query": "happily", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"happily\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.46, 1.14, "happily"]], "response_mode": "offline", "answer": "[{\"word\":\"happily\",\"start\":0.46,\"end\":1.14}]"} {"utterance_id": "1686-142278-0080", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0080.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[1.12, 1.55, "children"]], "response_mode": "offline", "answer": "[{\"word\":\"children\",\"start\":1.12,\"end\":1.55}]"} {"utterance_id": "5849-50873-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0036.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the full audio. After it finishes, find every instance of a plant name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.74, 6.05, "tree"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":5.74,\"end\":6.05}]"} {"utterance_id": "7697-105815-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0027.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[1.94, 2.25, "knees"]], "response_mode": "offline", "answer": "[{\"word\":\"knees\",\"start\":1.94,\"end\":2.25}]"} {"utterance_id": "7601-175351-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0004.flac", "task_type": "lexical_detection", "query": "warren", "prompt": "Process the full audio offline. Find every instance of the word \"warren\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 4.09, "warren"]], "response_mode": "offline", "answer": "[{\"word\":\"warren\",\"start\":3.56,\"end\":4.09}]"} {"utterance_id": "3660-172182-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0028.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.6, 5.04, "vassal"]], "response_mode": "offline", "answer": "[{\"word\":\"vassal\",\"start\":4.6,\"end\":5.04}]"} {"utterance_id": "6841-88291-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0052.flac", "task_type": "lexical_detection", "query": "squatted", "prompt": "Process the full audio offline. Find every instance of the word \"squatted\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.23, "squatted"]], "response_mode": "offline", "answer": "[{\"word\":\"squatted\",\"start\":0.79,\"end\":1.23}]"} {"utterance_id": "4323-18416-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0018.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.31, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.02,\"end\":2.31}]"} {"utterance_id": "1686-142278-0073", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0073.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.5, 1.02, "mister hale"], [8.26, 8.93, "mister bell"], [10.47, 11.49, "mister bell margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale\",\"start\":0.5,\"end\":1.02},{\"word\":\"mister bell\",\"start\":8.26,\"end\":8.93},{\"word\":\"mister bell margaret\",\"start\":10.47,\"end\":11.49}]"} {"utterance_id": "6841-88291-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0012.flac", "task_type": "lexical_detection", "query": "loop", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"loop\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.76, "loop"]], "response_mode": "offline", "answer": "[{\"word\":\"loop\",\"start\":0.53,\"end\":0.76}]"} {"utterance_id": "8254-84205-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.46, 1.81, "leg"]], "response_mode": "offline", "answer": "[{\"word\":\"leg\",\"start\":1.46,\"end\":1.81}]"} {"utterance_id": "4831-25894-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0022.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.58, 1.21, "fourth"]], "response_mode": "offline", "answer": "[{\"word\":\"fourth\",\"start\":0.58,\"end\":1.21}]"} {"utterance_id": "1651-136854-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0017.flac", "task_type": "lexical_detection", "query": "insolence", "prompt": "Process the full audio offline. Find every instance of the word \"insolence\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.88, "insolence"]], "response_mode": "offline", "answer": "[{\"word\":\"insolence\",\"start\":2.09,\"end\":2.88}]"} {"utterance_id": "3660-172183-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0008.flac", "task_type": "lexical_detection", "query": "heed", "prompt": "When the utterance is complete, report all instances of the word \"heed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.02, "heed"]], "response_mode": "offline", "answer": "[{\"word\":\"heed\",\"start\":0.76,\"end\":1.02}]"} {"utterance_id": "6123-59150-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0020.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 2.16, "cats"]], "response_mode": "offline", "answer": "[{\"word\":\"cats\",\"start\":1.59,\"end\":2.16}]"} {"utterance_id": "3660-172182-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.99, 5.28, "knight"]], "response_mode": "offline", "answer": "[{\"word\":\"knight\",\"start\":4.99,\"end\":5.28}]"} {"utterance_id": "1686-142278-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0012.flac", "task_type": "lexical_detection", "query": "lately", "prompt": "Once the audio has finished, identify every instance of the word \"lately\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.66, 4.13, "lately"]], "response_mode": "offline", "answer": "[{\"word\":\"lately\",\"start\":3.66,\"end\":4.13}]"} {"utterance_id": "7641-96684-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0009.flac", "task_type": "lexical_detection", "query": "sky", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"sky\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.15, 6.59, "sky"]], "response_mode": "offline", "answer": "[{\"word\":\"sky\",\"start\":6.15,\"end\":6.59}]"} {"utterance_id": "2506-11267-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0015.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.81, "dinners"]], "response_mode": "offline", "answer": "[{\"word\":\"dinners\",\"start\":3.33,\"end\":3.81}]"} {"utterance_id": "1686-142278-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0065.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.48, "frederick"]], "response_mode": "offline", "answer": "[{\"word\":\"frederick\",\"start\":1.95,\"end\":2.48}]"} {"utterance_id": "5543-27761-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0079.flac", "task_type": "lexical_detection", "query": "occasions", "prompt": "After the audio ends, extract all occurrences of the word \"occasions\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.11, 4.89, "occasions"]], "response_mode": "offline", "answer": "[{\"word\":\"occasions\",\"start\":4.11,\"end\":4.89}]"} {"utterance_id": "5849-50873-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0041.flac", "task_type": "semantic_counting", "query": "city", "prompt": "Listen to the entire clip, then report the count sequence for a city name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.18, 10.92, "one"], [17.21, 17.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":10.18,\"end\":10.92},{\"word\":\"two\",\"start\":17.21,\"end\":17.72}]"} {"utterance_id": "4831-25894-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0018.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.81, "faces"]], "response_mode": "offline", "answer": "[{\"word\":\"faces\",\"start\":2.31,\"end\":2.81}]"} {"utterance_id": "8254-84205-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.13, "chris"], [3.25, 3.57, "ned"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":1.75,\"end\":2.13},{\"word\":\"ned\",\"start\":3.25,\"end\":3.57}]"} {"utterance_id": "3660-6517-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0032.flac", "task_type": "lexical_detection", "query": "pump", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pump\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.76, 1.02, "pump"]], "response_mode": "offline", "answer": "[{\"word\":\"pump\",\"start\":0.76,\"end\":1.02}]"} {"utterance_id": "8254-84205-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0032.flac", "task_type": "lexical_detection", "query": "trees", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"trees\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.05, 3.46, "trees"]], "response_mode": "offline", "answer": "[{\"word\":\"trees\",\"start\":3.05,\"end\":3.46}]"} {"utterance_id": "1701-141759-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0029.flac", "task_type": "lexical_detection", "query": "blushed", "prompt": "After the audio ends, extract all occurrences of the word \"blushed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.7, 2.32, "blushed"]], "response_mode": "offline", "answer": "[{\"word\":\"blushed\",\"start\":1.7,\"end\":2.32}]"} {"utterance_id": "1701-141760-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.76, 4.42, "bagration"], [6.22, 7.12, "anna mikhaylovna"]], "response_mode": "offline", "answer": "[{\"word\":\"bagration\",\"start\":3.76,\"end\":4.42},{\"word\":\"anna mikhaylovna\",\"start\":6.22,\"end\":7.12}]"} {"utterance_id": "4323-55228-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0012.flac", "task_type": "lexical_detection", "query": "brushed", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"brushed\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.89, 1.3, "brushed"]], "response_mode": "offline", "answer": "[{\"word\":\"brushed\",\"start\":0.89,\"end\":1.3}]"} {"utterance_id": "8173-294714-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0032.flac", "task_type": "lexical_detection", "query": "complete", "prompt": "After listening to the whole recording, output every instance of the word \"complete\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.89, "complete"]], "response_mode": "offline", "answer": "[{\"word\":\"complete\",\"start\":0.39,\"end\":0.89}]"} {"utterance_id": "8173-294714-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0020.flac", "task_type": "lexical_detection", "query": "consent", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"consent\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.01, 4.46, "consent"]], "response_mode": "offline", "answer": "[{\"word\":\"consent\",\"start\":4.01,\"end\":4.46}]"} {"utterance_id": "5849-50873-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0039.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen carefully to the entire clip, then report every instance of an organization name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.92, 5.33, "assembly"]], "response_mode": "offline", "answer": "[{\"word\":\"assembly\",\"start\":4.92,\"end\":5.33}]"} {"utterance_id": "116-288048-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0019.flac", "task_type": "lexical_detection", "query": "against", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"against\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.86, 2.34, "against"]], "response_mode": "offline", "answer": "[{\"word\":\"against\",\"start\":1.86,\"end\":2.34}]"} {"utterance_id": "700-122866-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0017.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Process the full audio offline. Find every instance of a color name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.08, "white"]], "response_mode": "offline", "answer": "[{\"word\":\"white\",\"start\":0.7,\"end\":1.08}]"} {"utterance_id": "5849-50873-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.45, 1.03, "charles"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":0.45,\"end\":1.03}]"} {"utterance_id": "4572-64670-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.27, 4.5, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.27,\"end\":4.5}]"} {"utterance_id": "1255-74899-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0022.flac", "task_type": "lexical_detection", "query": "sit", "prompt": "After the audio ends, extract all occurrences of the word \"sit\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.69, "sit"]], "response_mode": "offline", "answer": "[{\"word\":\"sit\",\"start\":0.48,\"end\":0.69}]"} {"utterance_id": "6123-59186-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.4,\"end\":0.63}]"} {"utterance_id": "6841-88294-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0010.flac", "task_type": "lexical_detection", "query": "solemn", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"solemn\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.51, 3.99, "solemn"]], "response_mode": "offline", "answer": "[{\"word\":\"solemn\",\"start\":3.51,\"end\":3.99}]"} {"utterance_id": "116-288048-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0000.flac", "task_type": "lexical_detection", "query": "dates", "prompt": "When the utterance is complete, report all instances of the word \"dates\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.5, "dates"]], "response_mode": "offline", "answer": "[{\"word\":\"dates\",\"start\":1.21,\"end\":1.5}]"} {"utterance_id": "1255-138279-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0002.flac", "task_type": "lexical_counting", "query": "wall", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"wall\" with word/start/end. If none, reply [].", "labels": [[3.02, 3.48, "one"], [5.7, 6.2, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.02,\"end\":3.48},{\"word\":\"two\",\"start\":5.7,\"end\":6.2}]"} {"utterance_id": "4515-11057-0071", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0071.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After the audio ends, extract all occurrences of a weather term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.54, "rain"]], "response_mode": "offline", "answer": "[{\"word\":\"rain\",\"start\":2.16,\"end\":2.54}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Once the audio has finished, list each occurrence of a family relation as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.81, 1.11, "one"], [7.95, 8.31, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.81,\"end\":1.11},{\"word\":\"two\",\"start\":7.95,\"end\":8.31}]"} {"utterance_id": "6123-59186-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0005.flac", "task_type": "lexical_detection", "query": "image", "prompt": "When the utterance is complete, report all instances of the word \"image\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 4.81, "image"]], "response_mode": "offline", "answer": "[{\"word\":\"image\",\"start\":4.52,\"end\":4.81}]"} {"utterance_id": "1651-136854-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.24, 13.94, "socrates"]], "response_mode": "offline", "answer": "[{\"word\":\"socrates\",\"start\":13.24,\"end\":13.94}]"} {"utterance_id": "6841-88294-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0017.flac", "task_type": "lexical_detection", "query": "blazing", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"blazing\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.1, 3.48, "blazing"]], "response_mode": "offline", "answer": "[{\"word\":\"blazing\",\"start\":3.1,\"end\":3.48}]"} {"utterance_id": "7641-96684-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0002.flac", "task_type": "lexical_detection", "query": "companions", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"companions\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.73, 4.34, "companions"]], "response_mode": "offline", "answer": "[{\"word\":\"companions\",\"start\":3.73,\"end\":4.34}]"} {"utterance_id": "1255-138279-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0009.flac", "task_type": "lexical_detection", "query": "five", "prompt": "After the audio ends, extract all occurrences of the word \"five\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 2.21, "five"]], "response_mode": "offline", "answer": "[{\"word\":\"five\",\"start\":1.49,\"end\":2.21}]"} {"utterance_id": "1686-142278-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.52, 0.97, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.52,\"end\":0.97}]"} {"utterance_id": "6467-94831-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.61, 7.33, "cedarville"], [9.39, 10.11, "ithaca"]], "response_mode": "offline", "answer": "[{\"word\":\"cedarville\",\"start\":6.61,\"end\":7.33},{\"word\":\"ithaca\",\"start\":9.39,\"end\":10.11}]"} {"utterance_id": "3660-6517-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0031.flac", "task_type": "lexical_detection", "query": "draw", "prompt": "After listening to the whole recording, output every instance of the word \"draw\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.71, 2.95, "draw"]], "response_mode": "offline", "answer": "[{\"word\":\"draw\",\"start\":2.71,\"end\":2.95}]"} {"utterance_id": "3663-172528-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0030.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.29, 0.82, "one"], [6.43, 7.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.29,\"end\":0.82},{\"word\":\"two\",\"start\":6.43,\"end\":7.86}]"} {"utterance_id": "6841-88291-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0024.flac", "task_type": "semantic_detection", "query": "color", "prompt": "When the utterance is complete, report all instances of a color name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.38, "cherry"]], "response_mode": "offline", "answer": "[{\"word\":\"cherry\",\"start\":0.91,\"end\":1.38}]"} {"utterance_id": "7697-105815-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0044.flac", "task_type": "lexical_detection", "query": "glad", "prompt": "When the utterance is complete, report all instances of the word \"glad\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.74, 4.09, "glad"]], "response_mode": "offline", "answer": "[{\"word\":\"glad\",\"start\":3.74,\"end\":4.09}]"} {"utterance_id": "1686-142278-0082", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0082.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.85, 12.06, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":11.85,\"end\":12.06}]"} {"utterance_id": "1650-157641-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0005.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen carefully to the entire clip, then report every instance of a plant name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.55, "tobacco"]], "response_mode": "offline", "answer": "[{\"word\":\"tobacco\",\"start\":3.82,\"end\":4.55}]"} {"utterance_id": "1255-138279-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0020.flac", "task_type": "lexical_detection", "query": "money", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"money\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.77, 1.2, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":0.77,\"end\":1.2}]"} {"utterance_id": "7697-245712-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0016.flac", "task_type": "lexical_detection", "query": "inequality", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"inequality\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.3, 0.89, "inequality"]], "response_mode": "offline", "answer": "[{\"word\":\"inequality\",\"start\":0.3,\"end\":0.89}]"} {"utterance_id": "3660-172183-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0014.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an ordinal number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 3.83, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":3.36,\"end\":3.83}]"} {"utterance_id": "7601-101622-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0000.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[27.92, 28.32, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":27.92,\"end\":28.32}]"} {"utterance_id": "4515-11057-0080", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0080.flac", "task_type": "lexical_detection", "query": "world", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"world\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 2.86, "world"]], "response_mode": "offline", "answer": "[{\"word\":\"world\",\"start\":2.48,\"end\":2.86}]"} {"utterance_id": "4323-18416-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0020.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[1.67, 1.98, "aunt"], [7.81, 8.36, "grandpapa"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt\",\"start\":1.67,\"end\":1.98},{\"word\":\"grandpapa\",\"start\":7.81,\"end\":8.36}]"} {"utterance_id": "1255-90407-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0014.flac", "task_type": "lexical_detection", "query": "nets", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"nets\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.01, 1.47, "nets"]], "response_mode": "offline", "answer": "[{\"word\":\"nets\",\"start\":1.01,\"end\":1.47}]"} {"utterance_id": "2506-13150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0008.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.99, 5.72, "abdera"]], "response_mode": "offline", "answer": "[{\"word\":\"abdera\",\"start\":4.99,\"end\":5.72}]"} {"utterance_id": "7601-175351-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0020.flac", "task_type": "lexical_detection", "query": "snow", "prompt": "After listening to the whole recording, output every instance of the word \"snow\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.07, 13.52, "snow"]], "response_mode": "offline", "answer": "[{\"word\":\"snow\",\"start\":13.07,\"end\":13.52}]"} {"utterance_id": "3660-172182-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0031.flac", "task_type": "lexical_detection", "query": "between", "prompt": "When the utterance is complete, report all instances of the word \"between\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.78, 3.22, "between"]], "response_mode": "offline", "answer": "[{\"word\":\"between\",\"start\":2.78,\"end\":3.22}]"} {"utterance_id": "7697-105815-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0032.flac", "task_type": "lexical_detection", "query": "weak", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"weak\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.66, "weak"]], "response_mode": "offline", "answer": "[{\"word\":\"weak\",\"start\":1.36,\"end\":1.66}]"} {"utterance_id": "3663-172005-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0000.flac", "task_type": "semantic_detection", "query": "food", "prompt": "You will hear a complete utterance. When it ends, list every instance of a food name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.61, "meal"]], "response_mode": "offline", "answer": "[{\"word\":\"meal\",\"start\":2.16,\"end\":2.61}]"} {"utterance_id": "2506-13150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0001.flac", "task_type": "lexical_detection", "query": "upside", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"upside\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.84, 12.29, "upside"]], "response_mode": "offline", "answer": "[{\"word\":\"upside\",\"start\":11.84,\"end\":12.29}]"} {"utterance_id": "1630-73710-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[10.05, 10.64, "poet"]], "response_mode": "offline", "answer": "[{\"word\":\"poet\",\"start\":10.05,\"end\":10.64}]"} {"utterance_id": "1630-73710-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.56, "henrietta"]], "response_mode": "offline", "answer": "[{\"word\":\"henrietta\",\"start\":1.98,\"end\":2.56}]"} {"utterance_id": "4831-18525-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0025.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[1.38, 1.74, "one"], [2.87, 3.25, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.38,\"end\":1.74},{\"word\":\"two\",\"start\":2.87,\"end\":3.25}]"} {"utterance_id": "7697-105815-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0038.flac", "task_type": "lexical_detection", "query": "midst", "prompt": "When the utterance is complete, report all instances of the word \"midst\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.38, "midst"]], "response_mode": "offline", "answer": "[{\"word\":\"midst\",\"start\":1.04,\"end\":1.38}]"} {"utterance_id": "4323-55228-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0036.flac", "task_type": "lexical_detection", "query": "breakfast", "prompt": "Once the audio has finished, identify every instance of the word \"breakfast\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.17, 6.64, "breakfast"]], "response_mode": "offline", "answer": "[{\"word\":\"breakfast\",\"start\":6.17,\"end\":6.64}]"} {"utterance_id": "3915-57461-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[1.89, 2.8, "two hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"two hundred\",\"start\":1.89,\"end\":2.8}]"} {"utterance_id": "4831-25894-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0013.flac", "task_type": "lexical_detection", "query": "fat", "prompt": "Once the audio has finished, identify every instance of the word \"fat\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.62, 0.85, "fat"]], "response_mode": "offline", "answer": "[{\"word\":\"fat\",\"start\":0.62,\"end\":0.85}]"} {"utterance_id": "7601-101622-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0006.flac", "task_type": "lexical_detection", "query": "cheerfully", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"cheerfully\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.59, 5.29, "cheerfully"]], "response_mode": "offline", "answer": "[{\"word\":\"cheerfully\",\"start\":4.59,\"end\":5.29}]"} {"utterance_id": "4323-55228-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0036.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.17, 6.64, "breakfast"]], "response_mode": "offline", "answer": "[{\"word\":\"breakfast\",\"start\":6.17,\"end\":6.64}]"} {"utterance_id": "6467-97061-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0008.flac", "task_type": "lexical_detection", "query": "criminal", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"criminal\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.79, 6.19, "criminal"]], "response_mode": "offline", "answer": "[{\"word\":\"criminal\",\"start\":5.79,\"end\":6.19}]"} {"utterance_id": "3915-98647-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0020.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the full audio. Count every occurrence of a body part. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 1.04, "one"], [5.22, 5.65, "two"], [6.66, 7.15, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.56,\"end\":1.04},{\"word\":\"two\",\"start\":5.22,\"end\":5.65},{\"word\":\"three\",\"start\":6.66,\"end\":7.15}]"} {"utterance_id": "7697-105815-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0022.flac", "task_type": "lexical_detection", "query": "cave", "prompt": "After listening to the whole recording, output every instance of the word \"cave\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.06, "cave"]], "response_mode": "offline", "answer": "[{\"word\":\"cave\",\"start\":1.68,\"end\":2.06}]"} {"utterance_id": "4515-11057-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0039.flac", "task_type": "lexical_detection", "query": "delicate", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"delicate\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.7, 2.38, "delicate"]], "response_mode": "offline", "answer": "[{\"word\":\"delicate\",\"start\":1.7,\"end\":2.38}]"} {"utterance_id": "4323-55228-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0009.flac", "task_type": "lexical_counting", "query": "billy", "prompt": "Process the full audio offline. For every occurrence of the word \"billy\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.5, "one"], [6.69, 7.01, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.23,\"end\":0.5},{\"word\":\"two\",\"start\":6.69,\"end\":7.01}]"} {"utterance_id": "5849-50963-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0010.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen carefully to the entire clip, then report every instance of a money or currency term as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 3.97, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":3.54,\"end\":3.97}]"} {"utterance_id": "3663-172528-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0011.flac", "task_type": "lexical_detection", "query": "requires", "prompt": "When the utterance is complete, report all instances of the word \"requires\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.99, "requires"]], "response_mode": "offline", "answer": "[{\"word\":\"requires\",\"start\":1.16,\"end\":1.99}]"} {"utterance_id": "4323-18416-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 1.78, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.43,\"end\":1.78}]"} {"utterance_id": "1650-157641-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0015.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the full audio. After it finishes, find every instance of a plant name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.84, 5.34, "tobacco"]], "response_mode": "offline", "answer": "[{\"word\":\"tobacco\",\"start\":4.84,\"end\":5.34}]"} {"utterance_id": "4323-18416-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.45, 0.93, "charlotte"]], "response_mode": "offline", "answer": "[{\"word\":\"charlotte\",\"start\":0.45,\"end\":0.93}]"} {"utterance_id": "6123-59150-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0029.flac", "task_type": "lexical_detection", "query": "boldly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"boldly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.4, "boldly"]], "response_mode": "offline", "answer": "[{\"word\":\"boldly\",\"start\":1.02,\"end\":1.4}]"} {"utterance_id": "6841-88294-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0033.flac", "task_type": "lexical_detection", "query": "didn't", "prompt": "After the audio ends, extract all occurrences of the word \"didn't\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 0.76, "didn't"]], "response_mode": "offline", "answer": "[{\"word\":\"didn't\",\"start\":0.56,\"end\":0.76}]"} {"utterance_id": "1701-141760-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0012.flac", "task_type": "semantic_counting", "query": "ordinal", "prompt": "After the audio ends, tally every instance of an ordinal number. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[4.47, 4.78, "one"], [12.48, 12.78, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.47,\"end\":4.78},{\"word\":\"two\",\"start\":12.48,\"end\":12.78}]"} {"utterance_id": "6455-67804-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0013.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.94, "breakfast"]], "response_mode": "offline", "answer": "[{\"word\":\"breakfast\",\"start\":0.49,\"end\":0.94}]"} {"utterance_id": "6267-53049-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.35, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":2.61,\"end\":3.35}]"} {"utterance_id": "700-122867-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.53, 0.8, "anne's"], [7.1, 7.77, "josie pye"], [10.45, 10.76, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne's\",\"start\":0.53,\"end\":0.8},{\"word\":\"josie pye\",\"start\":7.1,\"end\":7.77},{\"word\":\"anne\",\"start\":10.45,\"end\":10.76}]"} {"utterance_id": "700-122868-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.19, 4.76, "gilbert blythe"]], "response_mode": "offline", "answer": "[{\"word\":\"gilbert blythe\",\"start\":4.19,\"end\":4.76}]"} {"utterance_id": "2506-11278-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0016.flac", "task_type": "lexical_detection", "query": "instances", "prompt": "When the utterance is complete, report all instances of the word \"instances\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.73, 2.37, "instances"]], "response_mode": "offline", "answer": "[{\"word\":\"instances\",\"start\":1.73,\"end\":2.37}]"} {"utterance_id": "3660-6517-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0007.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a food name with fields word, start, end (seconds). If none, reply [].", "labels": [[8.1, 8.55, "fodder"]], "response_mode": "offline", "answer": "[{\"word\":\"fodder\",\"start\":8.1,\"end\":8.55}]"} {"utterance_id": "7641-96684-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0023.flac", "task_type": "lexical_detection", "query": "palace", "prompt": "When the utterance is complete, report all instances of the word \"palace\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.84, 3.38, "palace"]], "response_mode": "offline", "answer": "[{\"word\":\"palace\",\"start\":2.84,\"end\":3.38}]"} {"utterance_id": "1255-74899-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0014.flac", "task_type": "lexical_detection", "query": "boil", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"boil\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.61, 2.05, "boil"], [3.72, 4.08, "boil"]], "response_mode": "offline", "answer": "[{\"word\":\"boil\",\"start\":1.61,\"end\":2.05},{\"word\":\"boil\",\"start\":3.72,\"end\":4.08}]"} {"utterance_id": "1650-167613-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0014.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the full audio. After it finishes, find every instance of a country name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.99, "russia"]], "response_mode": "offline", "answer": "[{\"word\":\"russia\",\"start\":0.63,\"end\":0.99}]"} {"utterance_id": "1686-142278-0074", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0074.flac", "task_type": "lexical_detection", "query": "never", "prompt": "After listening to the whole recording, output every instance of the word \"never\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 1.79, "never"]], "response_mode": "offline", "answer": "[{\"word\":\"never\",\"start\":1.51,\"end\":1.79}]"} {"utterance_id": "7641-96670-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0016.flac", "task_type": "lexical_detection", "query": "truly", "prompt": "Process the full audio offline. Find every instance of the word \"truly\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 1.88, "truly"]], "response_mode": "offline", "answer": "[{\"word\":\"truly\",\"start\":1.43,\"end\":1.88}]"} {"utterance_id": "7641-96670-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0020.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.69, 4.05, "bulls"]], "response_mode": "offline", "answer": "[{\"word\":\"bulls\",\"start\":3.69,\"end\":4.05}]"} {"utterance_id": "4515-11057-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0012.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 2.38, "whelpdale"], [3.52, 3.89, "biffen"]], "response_mode": "offline", "answer": "[{\"word\":\"whelpdale\",\"start\":1.4,\"end\":2.38},{\"word\":\"biffen\",\"start\":3.52,\"end\":3.89}]"} {"utterance_id": "7697-105815-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.96, "prudence"]], "response_mode": "offline", "answer": "[{\"word\":\"prudence\",\"start\":0.34,\"end\":0.96}]"} {"utterance_id": "7697-105815-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0026.flac", "task_type": "lexical_detection", "query": "felt", "prompt": "After the audio ends, extract all occurrences of the word \"felt\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.73, "felt"]], "response_mode": "offline", "answer": "[{\"word\":\"felt\",\"start\":0.45,\"end\":0.73}]"} {"utterance_id": "3660-172183-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.93, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.68,\"end\":0.93}]"} {"utterance_id": "4831-25894-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0033.flac", "task_type": "lexical_detection", "query": "delight", "prompt": "Once the audio has finished, identify every instance of the word \"delight\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.01, 2.43, "delight"]], "response_mode": "offline", "answer": "[{\"word\":\"delight\",\"start\":2.01,\"end\":2.43}]"} {"utterance_id": "1686-142278-0071", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0071.flac", "task_type": "lexical_detection", "query": "stand", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"stand\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.26, "stand"]], "response_mode": "offline", "answer": "[{\"word\":\"stand\",\"start\":0.96,\"end\":1.26}]"} {"utterance_id": "6599-38591-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0006.flac", "task_type": "lexical_detection", "query": "tapped", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"tapped\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[10.56, 10.85, "tapped"]], "response_mode": "offline", "answer": "[{\"word\":\"tapped\",\"start\":10.56,\"end\":10.85}]"} {"utterance_id": "8288-274162-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.1, "d'artagnan"]], "response_mode": "offline", "answer": "[{\"word\":\"d'artagnan\",\"start\":0.5,\"end\":1.1}]"} {"utterance_id": "1630-96099-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0000.flac", "task_type": "lexical_detection", "query": "undertake", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"undertake\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.88, 2.6, "undertake"]], "response_mode": "offline", "answer": "[{\"word\":\"undertake\",\"start\":1.88,\"end\":2.6}]"} {"utterance_id": "7641-96684-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0017.flac", "task_type": "lexical_counting", "query": "palace", "prompt": "When the utterance is complete, report how the word \"palace\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.04, 2.59, "one"], [14.59, 15.01, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.04,\"end\":2.59},{\"word\":\"two\",\"start\":14.59,\"end\":15.01}]"} {"utterance_id": "4323-55228-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0037.flac", "task_type": "lexical_detection", "query": "frown", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"frown\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.47, 3.83, "frown"]], "response_mode": "offline", "answer": "[{\"word\":\"frown\",\"start\":3.47,\"end\":3.83}]"} {"utterance_id": "700-122866-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0028.flac", "task_type": "lexical_detection", "query": "cultivate", "prompt": "After the audio ends, extract all occurrences of the word \"cultivate\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 1.07, "cultivate"]], "response_mode": "offline", "answer": "[{\"word\":\"cultivate\",\"start\":0.47,\"end\":1.07}]"} {"utterance_id": "700-122866-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0031.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.53, 7.81, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.53,\"end\":7.81}]"} {"utterance_id": "1701-141760-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0028.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.08, 3.54, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":3.08,\"end\":3.54}]"} {"utterance_id": "6455-67803-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0005.flac", "task_type": "lexical_counting", "query": "came", "prompt": "Listen to the entire clip, then report the count sequence for the word \"came\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.18, "one"], [3.41, 3.58, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.0,\"end\":1.18},{\"word\":\"two\",\"start\":3.41,\"end\":3.58}]"} {"utterance_id": "8288-274162-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.49, 1.06, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":0.49,\"end\":1.06}]"} {"utterance_id": "4323-55228-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0014.flac", "task_type": "lexical_detection", "query": "william's", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"william's\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.45, 2.84, "william's"]], "response_mode": "offline", "answer": "[{\"word\":\"william's\",\"start\":2.45,\"end\":2.84}]"} {"utterance_id": "6455-66379-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0016.flac", "task_type": "lexical_detection", "query": "changed", "prompt": "After listening to the whole recording, output every instance of the word \"changed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.28, "changed"]], "response_mode": "offline", "answer": "[{\"word\":\"changed\",\"start\":0.94,\"end\":1.28}]"} {"utterance_id": "7641-96684-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.13, 5.53, "admetus admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus admetus\",\"start\":4.13,\"end\":5.53}]"} {"utterance_id": "1255-90413-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0012.flac", "task_type": "lexical_detection", "query": "observed", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"observed\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.1, 3.51, "observed"]], "response_mode": "offline", "answer": "[{\"word\":\"observed\",\"start\":3.1,\"end\":3.51}]"} {"utterance_id": "6123-59186-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0028.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.98, 4.49, "doves"]], "response_mode": "offline", "answer": "[{\"word\":\"doves\",\"start\":3.98,\"end\":4.49}]"} {"utterance_id": "3660-6517-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0007.flac", "task_type": "lexical_detection", "query": "forecastle", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"forecastle\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.45, "forecastle"]], "response_mode": "offline", "answer": "[{\"word\":\"forecastle\",\"start\":0.76,\"end\":1.45}]"} {"utterance_id": "1650-167613-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 1.12, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":0.39,\"end\":1.12}]"} {"utterance_id": "700-122866-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0032.flac", "task_type": "lexical_detection", "query": "girl", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"girl\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.47, 0.68, "girl"]], "response_mode": "offline", "answer": "[{\"word\":\"girl\",\"start\":0.47,\"end\":0.68}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "lexical_counting", "query": "waist", "prompt": "After listening to the whole recording, output timed count words for the word \"waist\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[16.49, 16.84, "one"], [18.1, 18.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":16.49,\"end\":16.84},{\"word\":\"two\",\"start\":18.1,\"end\":18.53}]"} {"utterance_id": "6267-65525-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0036.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.59, "keeper"]], "response_mode": "offline", "answer": "[{\"word\":\"keeper\",\"start\":1.17,\"end\":1.59}]"} {"utterance_id": "1686-142278-0078", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0078.flac", "task_type": "lexical_detection", "query": "manufacturers", "prompt": "After listening to the whole recording, output every instance of the word \"manufacturers\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.78, 4.59, "manufacturers"]], "response_mode": "offline", "answer": "[{\"word\":\"manufacturers\",\"start\":3.78,\"end\":4.59}]"} {"utterance_id": "1255-138279-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0004.flac", "task_type": "lexical_detection", "query": "form", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"form\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.96, 2.45, "form"]], "response_mode": "offline", "answer": "[{\"word\":\"form\",\"start\":1.96,\"end\":2.45}]"} {"utterance_id": "6467-97061-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0019.flac", "task_type": "lexical_detection", "query": "mutual", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mutual\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.1, 1.46, "mutual"]], "response_mode": "offline", "answer": "[{\"word\":\"mutual\",\"start\":1.1,\"end\":1.46}]"} {"utterance_id": "700-122868-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0020.flac", "task_type": "lexical_detection", "query": "unfortunate", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"unfortunate\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.5, 4.02, "unfortunate"]], "response_mode": "offline", "answer": "[{\"word\":\"unfortunate\",\"start\":3.5,\"end\":4.02}]"} {"utterance_id": "7697-105815-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0050.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.4, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.93,\"end\":1.4}]"} {"utterance_id": "4831-25894-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0018.flac", "task_type": "lexical_detection", "query": "baskets", "prompt": "When the utterance is complete, report all instances of the word \"baskets\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.65, 5.25, "baskets"]], "response_mode": "offline", "answer": "[{\"word\":\"baskets\",\"start\":4.65,\"end\":5.25}]"} {"utterance_id": "7601-175351-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0023.flac", "task_type": "lexical_detection", "query": "succeed", "prompt": "After the audio ends, extract all occurrences of the word \"succeed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.76, "succeed"]], "response_mode": "offline", "answer": "[{\"word\":\"succeed\",\"start\":2.06,\"end\":2.76}]"} {"utterance_id": "3915-98647-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0005.flac", "task_type": "lexical_detection", "query": "owners", "prompt": "Once the audio has finished, identify every instance of the word \"owners\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.39, 7.84, "owners"]], "response_mode": "offline", "answer": "[{\"word\":\"owners\",\"start\":7.39,\"end\":7.84}]"} {"utterance_id": "6841-88291-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0029.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.21, "puncher"]], "response_mode": "offline", "answer": "[{\"word\":\"puncher\",\"start\":0.84,\"end\":1.21}]"} {"utterance_id": "700-122867-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0013.flac", "task_type": "lexical_detection", "query": "hair", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"hair\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.55, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":2.18,\"end\":2.55}]"} {"utterance_id": "5849-50962-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0021.flac", "task_type": "lexical_counting", "query": "man", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"man\" with word/start/end. If none, reply [].", "labels": [[1.28, 1.76, "one"], [4.2, 4.5, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.28,\"end\":1.76},{\"word\":\"two\",\"start\":4.2,\"end\":4.5}]"} {"utterance_id": "1650-173551-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.82, 1.27, "porter"]], "response_mode": "offline", "answer": "[{\"word\":\"porter\",\"start\":0.82,\"end\":1.27}]"} {"utterance_id": "1650-157641-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0010.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen to the full audio. After it finishes, find every instance of a money or currency term. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.72, "guinea"]], "response_mode": "offline", "answer": "[{\"word\":\"guinea\",\"start\":1.36,\"end\":1.72}]"} {"utterance_id": "700-122867-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0021.flac", "task_type": "lexical_detection", "query": "beautiful", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"beautiful\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.21, 2.81, "beautiful"]], "response_mode": "offline", "answer": "[{\"word\":\"beautiful\",\"start\":2.21,\"end\":2.81}]"} {"utterance_id": "116-288046-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.84, "bruno"]], "response_mode": "offline", "answer": "[{\"word\":\"bruno\",\"start\":0.49,\"end\":0.84}]"} {"utterance_id": "700-122867-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.32, 2.75, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":2.32,\"end\":2.75}]"} {"utterance_id": "1630-73710-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0018.flac", "task_type": "lexical_detection", "query": "confess", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"confess\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.45, 9.23, "confess"]], "response_mode": "offline", "answer": "[{\"word\":\"confess\",\"start\":8.45,\"end\":9.23}]"} {"utterance_id": "3915-98647-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0006.flac", "task_type": "lexical_counting", "query": "luck", "prompt": "Listen to the full audio. Count every occurrence of the word \"luck\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.15, "one"], [13.52, 13.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.91,\"end\":3.15},{\"word\":\"two\",\"start\":13.52,\"end\":13.76}]"} {"utterance_id": "6267-65525-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.78, 2.51, "john lincoln"]], "response_mode": "offline", "answer": "[{\"word\":\"john lincoln\",\"start\":1.78,\"end\":2.51}]"} {"utterance_id": "3660-6517-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0026.flac", "task_type": "lexical_detection", "query": "wave", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"wave\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.03, 4.28, "wave"]], "response_mode": "offline", "answer": "[{\"word\":\"wave\",\"start\":4.03,\"end\":4.28}]"} {"utterance_id": "6841-88294-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0053.flac", "task_type": "lexical_detection", "query": "book", "prompt": "Once the audio has finished, identify every instance of the word \"book\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.4, 3.69, "book"]], "response_mode": "offline", "answer": "[{\"word\":\"book\",\"start\":3.4,\"end\":3.69}]"} {"utterance_id": "8254-84205-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0025.flac", "task_type": "lexical_detection", "query": "chris", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"chris\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.13, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":1.75,\"end\":2.13}]"} {"utterance_id": "6267-65525-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0039.flac", "task_type": "lexical_detection", "query": "hands", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"hands\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.47, 3.0, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":2.47,\"end\":3.0}]"} {"utterance_id": "7697-245712-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0007.flac", "task_type": "lexical_counting", "query": "excepted", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"excepted\" (word, start, end). If none, reply [].", "labels": [[2.39, 2.89, "one"], [6.74, 7.17, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.39,\"end\":2.89},{\"word\":\"two\",\"start\":6.74,\"end\":7.17}]"} {"utterance_id": "7641-96670-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0022.flac", "task_type": "lexical_detection", "query": "submit", "prompt": "After listening to the whole recording, output every instance of the word \"submit\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.97, 5.36, "submit"]], "response_mode": "offline", "answer": "[{\"word\":\"submit\",\"start\":4.97,\"end\":5.36}]"} {"utterance_id": "6841-88294-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.25, 0.62, "mex"]], "response_mode": "offline", "answer": "[{\"word\":\"mex\",\"start\":0.25,\"end\":0.62}]"} {"utterance_id": "4831-25894-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0025.flac", "task_type": "lexical_detection", "query": "fairy", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"fairy\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.85, 3.29, "fairy"]], "response_mode": "offline", "answer": "[{\"word\":\"fairy\",\"start\":2.85,\"end\":3.29}]"} {"utterance_id": "7601-291468-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0004.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[23.25, 23.82, "general"]], "response_mode": "offline", "answer": "[{\"word\":\"general\",\"start\":23.25,\"end\":23.82}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 1.41, "prince albert"], [6.92, 7.98, "grinnell"]], "response_mode": "offline", "answer": "[{\"word\":\"prince albert\",\"start\":0.53,\"end\":1.41},{\"word\":\"grinnell\",\"start\":6.92,\"end\":7.98}]"} {"utterance_id": "4572-112375-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0015.flac", "task_type": "lexical_detection", "query": "lend", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"lend\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.8, "lend"]], "response_mode": "offline", "answer": "[{\"word\":\"lend\",\"start\":0.48,\"end\":0.8}]"} {"utterance_id": "4515-11057-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0050.flac", "task_type": "lexical_detection", "query": "mean", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mean\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.58, 1.94, "mean"]], "response_mode": "offline", "answer": "[{\"word\":\"mean\",\"start\":1.58,\"end\":1.94}]"} {"utterance_id": "4831-25894-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0018.flac", "task_type": "lexical_counting", "query": "how", "prompt": "Process the full audio offline. For every occurrence of the word \"how\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.18, 0.35, "one"], [1.61, 1.81, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.18,\"end\":0.35},{\"word\":\"two\",\"start\":1.61,\"end\":1.81}]"} {"utterance_id": "1701-141759-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0017.flac", "task_type": "lexical_detection", "query": "believed", "prompt": "After the audio ends, extract all occurrences of the word \"believed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.87, "believed"]], "response_mode": "offline", "answer": "[{\"word\":\"believed\",\"start\":0.44,\"end\":0.87}]"} {"utterance_id": "8254-115543-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.41, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":5.09,\"end\":5.41}]"} {"utterance_id": "4323-55228-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0010.flac", "task_type": "lexical_detection", "query": "receive", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"receive\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.41, 3.91, "receive"]], "response_mode": "offline", "answer": "[{\"word\":\"receive\",\"start\":3.41,\"end\":3.91}]"} {"utterance_id": "2506-11278-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 4.1, "right honourable mister vernon"]], "response_mode": "offline", "answer": "[{\"word\":\"right honourable mister vernon\",\"start\":2.39,\"end\":4.1}]"} {"utterance_id": "3915-98647-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0017.flac", "task_type": "lexical_detection", "query": "custom", "prompt": "Once the audio has finished, identify every instance of the word \"custom\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.4, 4.8, "custom"]], "response_mode": "offline", "answer": "[{\"word\":\"custom\",\"start\":4.4,\"end\":4.8}]"} {"utterance_id": "7641-96684-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0031.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.73, 5.25, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":4.73,\"end\":5.25}]"} {"utterance_id": "1630-73710-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0009.flac", "task_type": "lexical_detection", "query": "pony", "prompt": "When the utterance is complete, report all instances of the word \"pony\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.75, "pony"]], "response_mode": "offline", "answer": "[{\"word\":\"pony\",\"start\":1.18,\"end\":1.75}]"} {"utterance_id": "1701-141759-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0030.flac", "task_type": "lexical_detection", "query": "crying", "prompt": "When the utterance is complete, report all instances of the word \"crying\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.31, 1.8, "crying"]], "response_mode": "offline", "answer": "[{\"word\":\"crying\",\"start\":1.31,\"end\":1.8}]"} {"utterance_id": "4323-55228-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.43, 0.69, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":0.43,\"end\":0.69}]"} {"utterance_id": "3660-172182-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.38, "guenever"], [5.52, 5.93, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"guenever\",\"start\":3.94,\"end\":4.38},{\"word\":\"geraint\",\"start\":5.52,\"end\":5.93}]"} {"utterance_id": "2506-11267-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0006.flac", "task_type": "lexical_detection", "query": "ill", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"ill\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.19, 2.32, "ill"]], "response_mode": "offline", "answer": "[{\"word\":\"ill\",\"start\":2.19,\"end\":2.32}]"} {"utterance_id": "8288-274162-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0037.flac", "task_type": "lexical_detection", "query": "tone", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"tone\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.7, "tone"]], "response_mode": "offline", "answer": "[{\"word\":\"tone\",\"start\":1.4,\"end\":1.7}]"} {"utterance_id": "6467-94831-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0033.flac", "task_type": "lexical_detection", "query": "box", "prompt": "After the audio ends, extract all occurrences of the word \"box\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 3.03, "box"]], "response_mode": "offline", "answer": "[{\"word\":\"box\",\"start\":2.57,\"end\":3.03}]"} {"utterance_id": "3663-172528-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0049.flac", "task_type": "lexical_detection", "query": "nevertheless", "prompt": "Once the audio has finished, identify every instance of the word \"nevertheless\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.27, 1.15, "nevertheless"]], "response_mode": "offline", "answer": "[{\"word\":\"nevertheless\",\"start\":0.27,\"end\":1.15}]"} {"utterance_id": "700-122867-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0011.flac", "task_type": "lexical_detection", "query": "red", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"red\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.32, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":2.06,\"end\":2.32}]"} {"utterance_id": "1255-74899-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0006.flac", "task_type": "lexical_counting", "query": "pride", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"pride\" with word/start/end. If none, reply [].", "labels": [[0.5, 0.86, "one"], [1.98, 2.34, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":0.86},{\"word\":\"two\",\"start\":1.98,\"end\":2.34}]"} {"utterance_id": "1630-73710-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0018.flac", "task_type": "lexical_counting", "query": "know", "prompt": "Once the audio has finished, list each occurrence of the word \"know\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[4.03, 4.29, "one"], [5.51, 5.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.03,\"end\":4.29},{\"word\":\"two\",\"start\":5.51,\"end\":5.76}]"} {"utterance_id": "8288-274162-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0048.flac", "task_type": "lexical_detection", "query": "don't", "prompt": "After the audio ends, extract all occurrences of the word \"don't\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.25, "don't"]], "response_mode": "offline", "answer": "[{\"word\":\"don't\",\"start\":1.0,\"end\":1.25}]"} {"utterance_id": "3663-172005-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0001.flac", "task_type": "lexical_detection", "query": "caught", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"caught\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.48, 1.8, "caught"]], "response_mode": "offline", "answer": "[{\"word\":\"caught\",\"start\":1.48,\"end\":1.8}]"} {"utterance_id": "7601-175351-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0008.flac", "task_type": "lexical_detection", "query": "we've", "prompt": "Once the audio has finished, identify every instance of the word \"we've\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.01, 5.44, "we've"]], "response_mode": "offline", "answer": "[{\"word\":\"we've\",\"start\":5.01,\"end\":5.44}]"} {"utterance_id": "3915-57461-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0026.flac", "task_type": "lexical_detection", "query": "fit", "prompt": "Process the full audio offline. Find every instance of the word \"fit\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.67, "fit"]], "response_mode": "offline", "answer": "[{\"word\":\"fit\",\"start\":0.4,\"end\":0.67}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.13, 3.78, "wellington"]], "response_mode": "offline", "answer": "[{\"word\":\"wellington\",\"start\":3.13,\"end\":3.78}]"} {"utterance_id": "6267-53049-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0003.flac", "task_type": "lexical_detection", "query": "admitted", "prompt": "When the utterance is complete, report all instances of the word \"admitted\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.87, "admitted"]], "response_mode": "offline", "answer": "[{\"word\":\"admitted\",\"start\":5.09,\"end\":5.87}]"} {"utterance_id": "2506-11267-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0008.flac", "task_type": "lexical_detection", "query": "you'd", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"you'd\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.56, "you'd"]], "response_mode": "offline", "answer": "[{\"word\":\"you'd\",\"start\":1.41,\"end\":1.56}]"} {"utterance_id": "700-122866-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0026.flac", "task_type": "lexical_detection", "query": "pretended", "prompt": "After the audio ends, extract all occurrences of the word \"pretended\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.19, "pretended"]], "response_mode": "offline", "answer": "[{\"word\":\"pretended\",\"start\":0.65,\"end\":1.19}]"} {"utterance_id": "4831-25894-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0018.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Process the full audio offline. Find every instance of a date and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.85, 9.27, "christmas"]], "response_mode": "offline", "answer": "[{\"word\":\"christmas\",\"start\":8.85,\"end\":9.27}]"} {"utterance_id": "6123-59150-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0032.flac", "task_type": "lexical_detection", "query": "sought", "prompt": "Once the audio has finished, identify every instance of the word \"sought\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.59, 5.95, "sought"]], "response_mode": "offline", "answer": "[{\"word\":\"sought\",\"start\":5.59,\"end\":5.95}]"} {"utterance_id": "116-288045-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0024.flac", "task_type": "lexical_detection", "query": "person", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"person\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.55, "person"]], "response_mode": "offline", "answer": "[{\"word\":\"person\",\"start\":1.02,\"end\":1.55}]"} {"utterance_id": "5849-50873-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0038.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a date with fields word, start, end (seconds). If none, reply [].", "labels": [[5.29, 6.93, "august eighteen fifty six"]], "response_mode": "offline", "answer": "[{\"word\":\"august eighteen fifty six\",\"start\":5.29,\"end\":6.93}]"} {"utterance_id": "4572-112375-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0012.flac", "task_type": "lexical_detection", "query": "mild", "prompt": "After listening to the whole recording, output every instance of the word \"mild\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.92, "mild"]], "response_mode": "offline", "answer": "[{\"word\":\"mild\",\"start\":2.38,\"end\":2.92}]"} {"utterance_id": "5543-27761-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0018.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.92, 1.22, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":0.92,\"end\":1.22}]"} {"utterance_id": "4831-18525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.93, 2.32, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.93,\"end\":2.32}]"} {"utterance_id": "4153-185072-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.24, 0.81, "nettles"], [3.9, 4.39, "allen"]], "response_mode": "offline", "answer": "[{\"word\":\"nettles\",\"start\":0.24,\"end\":0.81},{\"word\":\"allen\",\"start\":3.9,\"end\":4.39}]"} {"utterance_id": "3663-172528-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0026.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 2.1, "physician"]], "response_mode": "offline", "answer": "[{\"word\":\"physician\",\"start\":1.46,\"end\":2.1}]"} {"utterance_id": "8173-294714-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0056.flac", "task_type": "lexical_detection", "query": "promise", "prompt": "After the audio ends, extract all occurrences of the word \"promise\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.64, "promise"]], "response_mode": "offline", "answer": "[{\"word\":\"promise\",\"start\":1.15,\"end\":1.64}]"} {"utterance_id": "4570-14911-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0012.flac", "task_type": "lexical_detection", "query": "throat", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"throat\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.95, 5.32, "throat"]], "response_mode": "offline", "answer": "[{\"word\":\"throat\",\"start\":4.95,\"end\":5.32}]"} {"utterance_id": "1630-141772-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.28, "soldiers"]], "response_mode": "offline", "answer": "[{\"word\":\"soldiers\",\"start\":0.59,\"end\":1.28}]"} {"utterance_id": "1686-142278-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0026.flac", "task_type": "lexical_detection", "query": "church", "prompt": "After the audio ends, extract all occurrences of the word \"church\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.15, "church"]], "response_mode": "offline", "answer": "[{\"word\":\"church\",\"start\":1.83,\"end\":2.15}]"} {"utterance_id": "7697-105815-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0000.flac", "task_type": "lexical_detection", "query": "times", "prompt": "When the utterance is complete, report all instances of the word \"times\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.79, 6.44, "times"]], "response_mode": "offline", "answer": "[{\"word\":\"times\",\"start\":5.79,\"end\":6.44}]"} {"utterance_id": "7697-245712-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0015.flac", "task_type": "lexical_detection", "query": "bodily", "prompt": "After listening to the whole recording, output every instance of the word \"bodily\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.07, "bodily"]], "response_mode": "offline", "answer": "[{\"word\":\"bodily\",\"start\":1.75,\"end\":2.07}]"} {"utterance_id": "6267-53049-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0028.flac", "task_type": "lexical_detection", "query": "broken", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"broken\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.89, 5.4, "broken"]], "response_mode": "offline", "answer": "[{\"word\":\"broken\",\"start\":4.89,\"end\":5.4}]"} {"utterance_id": "7641-96670-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0019.flac", "task_type": "semantic_counting", "query": "color", "prompt": "Process the full audio offline. For every occurrence of a color name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.16, "one"], [3.38, 3.64, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.92,\"end\":1.16},{\"word\":\"two\",\"start\":3.38,\"end\":3.64}]"} {"utterance_id": "3663-172528-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0039.flac", "task_type": "lexical_detection", "query": "injured", "prompt": "Process the full audio offline. Find every instance of the word \"injured\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.88, 5.47, "injured"]], "response_mode": "offline", "answer": "[{\"word\":\"injured\",\"start\":4.88,\"end\":5.47}]"} {"utterance_id": "3660-172183-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0021.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.11, 3.38, "meal"]], "response_mode": "offline", "answer": "[{\"word\":\"meal\",\"start\":3.11,\"end\":3.38}]"} {"utterance_id": "1701-141759-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0003.flac", "task_type": "lexical_detection", "query": "matters", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"matters\" with fields word, start, end (seconds). If none, reply [].", "labels": [[12.78, 13.37, "matters"]], "response_mode": "offline", "answer": "[{\"word\":\"matters\",\"start\":12.78,\"end\":13.37}]"} {"utterance_id": "1255-90413-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0027.flac", "task_type": "lexical_detection", "query": "companion", "prompt": "After the audio ends, extract all occurrences of the word \"companion\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.2, 6.7, "companion"]], "response_mode": "offline", "answer": "[{\"word\":\"companion\",\"start\":6.2,\"end\":6.7}]"} {"utterance_id": "7601-175351-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.48, 13.81, "dick"]], "response_mode": "offline", "answer": "[{\"word\":\"dick\",\"start\":13.48,\"end\":13.81}]"} {"utterance_id": "5849-50963-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0001.flac", "task_type": "lexical_detection", "query": "happy", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"happy\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.71, 2.2, "happy"]], "response_mode": "offline", "answer": "[{\"word\":\"happy\",\"start\":1.71,\"end\":2.2}]"} {"utterance_id": "1701-141760-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0012.flac", "task_type": "lexical_counting", "query": "steps", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"steps\" with word/start/end. If none, reply [].", "labels": [[4.78, 5.16, "one"], [12.78, 13.31, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.78,\"end\":5.16},{\"word\":\"two\",\"start\":12.78,\"end\":13.31}]"} {"utterance_id": "6455-67803-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0019.flac", "task_type": "lexical_detection", "query": "hint", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hint\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.41, 5.78, "hint"]], "response_mode": "offline", "answer": "[{\"word\":\"hint\",\"start\":5.41,\"end\":5.78}]"} {"utterance_id": "4153-186222-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0022.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 4.0, "cousin"]], "response_mode": "offline", "answer": "[{\"word\":\"cousin\",\"start\":3.54,\"end\":4.0}]"} {"utterance_id": "1686-142278-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0058.flac", "task_type": "lexical_detection", "query": "bid", "prompt": "Once the audio has finished, identify every instance of the word \"bid\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.03, 2.22, "bid"]], "response_mode": "offline", "answer": "[{\"word\":\"bid\",\"start\":2.03,\"end\":2.22}]"} {"utterance_id": "6599-38590-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0000.flac", "task_type": "lexical_detection", "query": "rome", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"rome\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.23, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":2.76,\"end\":3.23}]"} {"utterance_id": "8254-115543-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0022.flac", "task_type": "lexical_detection", "query": "decorated", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"decorated\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.67, 3.23, "decorated"]], "response_mode": "offline", "answer": "[{\"word\":\"decorated\",\"start\":2.67,\"end\":3.23}]"} {"utterance_id": "3915-98647-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0004.flac", "task_type": "lexical_detection", "query": "destiny", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"destiny\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.37, 7.86, "destiny"]], "response_mode": "offline", "answer": "[{\"word\":\"destiny\",\"start\":7.37,\"end\":7.86}]"} {"utterance_id": "6123-59150-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.56, "karamazovs"]], "response_mode": "offline", "answer": "[{\"word\":\"karamazovs\",\"start\":0.93,\"end\":1.56}]"} {"utterance_id": "700-122866-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 3.04, "miss stacy"]], "response_mode": "offline", "answer": "[{\"word\":\"miss stacy\",\"start\":2.52,\"end\":3.04}]"} {"utterance_id": "8254-84205-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0048.flac", "task_type": "lexical_detection", "query": "quarters", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"quarters\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.78, "quarters"]], "response_mode": "offline", "answer": "[{\"word\":\"quarters\",\"start\":1.33,\"end\":1.78}]"} {"utterance_id": "700-122866-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0008.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.97, "one"], [6.95, 7.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.68,\"end\":0.97},{\"word\":\"two\",\"start\":6.95,\"end\":7.49}]"} {"utterance_id": "5543-27761-0072", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0072.flac", "task_type": "lexical_detection", "query": "covered", "prompt": "When the utterance is complete, report all instances of the word \"covered\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.43, 3.78, "covered"]], "response_mode": "offline", "answer": "[{\"word\":\"covered\",\"start\":3.43,\"end\":3.78}]"} {"utterance_id": "7641-96684-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0021.flac", "task_type": "lexical_detection", "query": "words", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"words\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.93, "words"], [3.19, 3.67, "words"]], "response_mode": "offline", "answer": "[{\"word\":\"words\",\"start\":0.62,\"end\":0.93},{\"word\":\"words\",\"start\":3.19,\"end\":3.67}]"} {"utterance_id": "8288-274162-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0036.flac", "task_type": "lexical_detection", "query": "affair", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"affair\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.07, 1.4, "affair"]], "response_mode": "offline", "answer": "[{\"word\":\"affair\",\"start\":1.07,\"end\":1.4}]"} {"utterance_id": "4323-13259-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0001.flac", "task_type": "lexical_detection", "query": "members", "prompt": "When the utterance is complete, report all instances of the word \"members\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.06, 7.42, "members"]], "response_mode": "offline", "answer": "[{\"word\":\"members\",\"start\":7.06,\"end\":7.42}]"} {"utterance_id": "4323-13259-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0019.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.09, 3.72, "correspondent"]], "response_mode": "offline", "answer": "[{\"word\":\"correspondent\",\"start\":3.09,\"end\":3.72}]"} {"utterance_id": "6467-97061-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0022.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.37, 6.81, "soldiers"]], "response_mode": "offline", "answer": "[{\"word\":\"soldiers\",\"start\":6.37,\"end\":6.81}]"} {"utterance_id": "8288-274162-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.64, 4.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":4.64,\"end\":4.86}]"} {"utterance_id": "7601-101619-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0002.flac", "task_type": "lexical_counting", "query": "pleasure", "prompt": "Listen to the full audio. Count every occurrence of the word \"pleasure\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.94, 7.61, "one"], [19.65, 20.33, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.94,\"end\":7.61},{\"word\":\"two\",\"start\":19.65,\"end\":20.33}]"} {"utterance_id": "6123-59150-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0016.flac", "task_type": "lexical_detection", "query": "fears", "prompt": "When the utterance is complete, report all instances of the word \"fears\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.29, "fears"]], "response_mode": "offline", "answer": "[{\"word\":\"fears\",\"start\":1.8,\"end\":2.29}]"} {"utterance_id": "6455-66379-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0004.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.75, "pilot's"]], "response_mode": "offline", "answer": "[{\"word\":\"pilot's\",\"start\":0.39,\"end\":0.75}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.4, 0.95, "one"], [9.2, 9.82, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.4,\"end\":0.95},{\"word\":\"two\",\"start\":9.2,\"end\":9.82}]"} {"utterance_id": "5849-50964-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0003.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.75, 4.34, "sisters"]], "response_mode": "offline", "answer": "[{\"word\":\"sisters\",\"start\":3.75,\"end\":4.34}]"} {"utterance_id": "1585-131718-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0036.flac", "task_type": "lexical_detection", "query": "utter", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"utter\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.8, 3.97, "utter"]], "response_mode": "offline", "answer": "[{\"word\":\"utter\",\"start\":3.8,\"end\":3.97}]"} {"utterance_id": "4323-13259-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0021.flac", "task_type": "lexical_detection", "query": "concerts", "prompt": "When the utterance is complete, report all instances of the word \"concerts\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.97, 8.48, "concerts"]], "response_mode": "offline", "answer": "[{\"word\":\"concerts\",\"start\":7.97,\"end\":8.48}]"} {"utterance_id": "116-288048-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0025.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen to the full audio. After it finishes, find every instance of a disease name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.35, 13.14, "disease"], [15.2, 15.81, "insanity"]], "response_mode": "offline", "answer": "[{\"word\":\"disease\",\"start\":12.35,\"end\":13.14},{\"word\":\"insanity\",\"start\":15.2,\"end\":15.81}]"} {"utterance_id": "1585-131718-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0030.flac", "task_type": "lexical_counting", "query": "edison", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"edison\" (word, start, end). If none, reply [].", "labels": [[3.55, 4.1, "one"], [8.5, 8.93, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.55,\"end\":4.1},{\"word\":\"two\",\"start\":8.5,\"end\":8.93}]"} {"utterance_id": "1585-131718-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0037.flac", "task_type": "lexical_detection", "query": "cigar", "prompt": "When the utterance is complete, report all instances of the word \"cigar\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.38, "cigar"]], "response_mode": "offline", "answer": "[{\"word\":\"cigar\",\"start\":2.98,\"end\":3.38}]"} {"utterance_id": "6455-66379-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0018.flac", "task_type": "lexical_detection", "query": "view", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"view\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.66, 2.82, "view"]], "response_mode": "offline", "answer": "[{\"word\":\"view\",\"start\":2.66,\"end\":2.82}]"} {"utterance_id": "4153-185072-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0004.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.71, 1.12, "one"], [2.16, 2.54, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.71,\"end\":1.12},{\"word\":\"two\",\"start\":2.16,\"end\":2.54}]"} {"utterance_id": "8173-294714-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0055.flac", "task_type": "lexical_detection", "query": "borrow", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"borrow\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.39, "borrow"]], "response_mode": "offline", "answer": "[{\"word\":\"borrow\",\"start\":2.09,\"end\":2.39}]"} {"utterance_id": "1255-138279-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0013.flac", "task_type": "lexical_detection", "query": "irregularities", "prompt": "Process the full audio offline. Find every instance of the word \"irregularities\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.26, 10.07, "irregularities"]], "response_mode": "offline", "answer": "[{\"word\":\"irregularities\",\"start\":9.26,\"end\":10.07}]"} {"utterance_id": "8288-274150-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0007.flac", "task_type": "lexical_counting", "query": "partly", "prompt": "Once the audio has finished, list each occurrence of the word \"partly\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.83, 2.26, "one"], [3.75, 4.16, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.83,\"end\":2.26},{\"word\":\"two\",\"start\":3.75,\"end\":4.16}]"} {"utterance_id": "4323-18416-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0008.flac", "task_type": "lexical_detection", "query": "pleased", "prompt": "After listening to the whole recording, output every instance of the word \"pleased\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.4, "pleased"]], "response_mode": "offline", "answer": "[{\"word\":\"pleased\",\"start\":2.08,\"end\":2.4}]"} {"utterance_id": "4515-11057-0082", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0082.flac", "task_type": "lexical_detection", "query": "stand", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"stand\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.67, 5.04, "stand"]], "response_mode": "offline", "answer": "[{\"word\":\"stand\",\"start\":4.67,\"end\":5.04}]"} {"utterance_id": "6455-66379-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.15, 3.52, "thompson"]], "response_mode": "offline", "answer": "[{\"word\":\"thompson\",\"start\":3.15,\"end\":3.52}]"} {"utterance_id": "1255-138279-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0005.flac", "task_type": "lexical_detection", "query": "human", "prompt": "When the utterance is complete, report all instances of the word \"human\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.08, 4.71, "human"]], "response_mode": "offline", "answer": "[{\"word\":\"human\",\"start\":4.08,\"end\":4.71}]"} {"utterance_id": "4323-55228-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0043.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 1.86, "face"], [3.81, 3.99, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":1.64,\"end\":1.86},{\"word\":\"eyes\",\"start\":3.81,\"end\":3.99}]"} {"utterance_id": "5849-50963-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[12.34, 12.88, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":12.34,\"end\":12.88}]"} {"utterance_id": "8173-294714-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.83, 7.34, "van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"van brandt\",\"start\":6.83,\"end\":7.34}]"} {"utterance_id": "8173-294714-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0001.flac", "task_type": "lexical_detection", "query": "helpless", "prompt": "Once the audio has finished, identify every instance of the word \"helpless\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.04, 3.71, "helpless"]], "response_mode": "offline", "answer": "[{\"word\":\"helpless\",\"start\":3.04,\"end\":3.71}]"} {"utterance_id": "1686-142278-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0018.flac", "task_type": "lexical_detection", "query": "mysterious", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mysterious\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.1, 3.01, "mysterious"]], "response_mode": "offline", "answer": "[{\"word\":\"mysterious\",\"start\":2.1,\"end\":3.01}]"} {"utterance_id": "4572-64670-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0002.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.1, 3.34, "wood"]], "response_mode": "offline", "answer": "[{\"word\":\"wood\",\"start\":3.1,\"end\":3.34}]"} {"utterance_id": "700-122867-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0011.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Once the audio has finished, identify every instance of a color name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.06, 2.32, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":2.06,\"end\":2.32}]"} {"utterance_id": "8254-84205-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0006.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After listening to the whole recording, output every instance of a time of day with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.01, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.57,\"end\":2.01}]"} {"utterance_id": "3915-57461-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.23, 11.62, "buddha"]], "response_mode": "offline", "answer": "[{\"word\":\"buddha\",\"start\":11.23,\"end\":11.62}]"} {"utterance_id": "6841-88291-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.61, "tail"]], "response_mode": "offline", "answer": "[{\"word\":\"tail\",\"start\":1.3,\"end\":1.61}]"} {"utterance_id": "4572-112375-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0014.flac", "task_type": "lexical_detection", "query": "pure", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"pure\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.98, 4.45, "pure"]], "response_mode": "offline", "answer": "[{\"word\":\"pure\",\"start\":3.98,\"end\":4.45}]"} {"utterance_id": "116-288048-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0003.flac", "task_type": "semantic_detection", "query": "country", "prompt": "You will hear a complete utterance. When it ends, list every instance of a country name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.35, 7.91, "egypt"]], "response_mode": "offline", "answer": "[{\"word\":\"egypt\",\"start\":7.35,\"end\":7.91}]"} {"utterance_id": "4323-55228-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0042.flac", "task_type": "lexical_detection", "query": "besides", "prompt": "Once the audio has finished, identify every instance of the word \"besides\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.3, 1.05, "besides"]], "response_mode": "offline", "answer": "[{\"word\":\"besides\",\"start\":0.3,\"end\":1.05}]"} {"utterance_id": "1650-167613-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0010.flac", "task_type": "lexical_detection", "query": "forbidden", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"forbidden\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 2.04, "forbidden"]], "response_mode": "offline", "answer": "[{\"word\":\"forbidden\",\"start\":1.36,\"end\":2.04}]"} {"utterance_id": "1630-96099-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0018.flac", "task_type": "lexical_detection", "query": "word", "prompt": "After listening to the whole recording, output every instance of the word \"word\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.9, 3.22, "word"]], "response_mode": "offline", "answer": "[{\"word\":\"word\",\"start\":2.9,\"end\":3.22}]"} {"utterance_id": "3660-172182-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0030.flac", "task_type": "lexical_counting", "query": "geraint", "prompt": "Process the full audio offline. For every occurrence of the word \"geraint\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.26, "one"], [6.86, 7.33, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.96,\"end\":1.26},{\"word\":\"two\",\"start\":6.86,\"end\":7.33}]"} {"utterance_id": "4570-56594-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0015.flac", "task_type": "lexical_detection", "query": "thar", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"thar\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 3.93, "thar"]], "response_mode": "offline", "answer": "[{\"word\":\"thar\",\"start\":3.52,\"end\":3.93}]"} {"utterance_id": "3660-172182-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0034.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.6, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":1.05,\"end\":1.6}]"} {"utterance_id": "116-288046-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0006.flac", "task_type": "lexical_detection", "query": "building", "prompt": "After listening to the whole recording, output every instance of the word \"building\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.93, 5.25, "building"]], "response_mode": "offline", "answer": "[{\"word\":\"building\",\"start\":4.93,\"end\":5.25}]"} {"utterance_id": "1255-138279-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0006.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[7.56, 7.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":7.56,\"end\":7.76}]"} {"utterance_id": "4323-55228-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0021.flac", "task_type": "lexical_detection", "query": "sigh", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"sigh\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.37, 1.78, "sigh"]], "response_mode": "offline", "answer": "[{\"word\":\"sigh\",\"start\":1.37,\"end\":1.78}]"} {"utterance_id": "7697-105815-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0041.flac", "task_type": "lexical_detection", "query": "watchful", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"watchful\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.37, "watchful"]], "response_mode": "offline", "answer": "[{\"word\":\"watchful\",\"start\":0.83,\"end\":1.37}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an organization name with fields word, start, end (seconds). If none, reply [].", "labels": [[7.16, 7.48, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"army\",\"start\":7.16,\"end\":7.48}]"} {"utterance_id": "5849-50873-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0009.flac", "task_type": "lexical_detection", "query": "debate", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"debate\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.85, 3.35, "debate"]], "response_mode": "offline", "answer": "[{\"word\":\"debate\",\"start\":2.85,\"end\":3.35}]"} {"utterance_id": "3663-172528-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0041.flac", "task_type": "lexical_detection", "query": "stamping", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"stamping\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.32, 7.86, "stamping"]], "response_mode": "offline", "answer": "[{\"word\":\"stamping\",\"start\":7.32,\"end\":7.86}]"} {"utterance_id": "7641-96684-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0035.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.35, "features"]], "response_mode": "offline", "answer": "[{\"word\":\"features\",\"start\":2.77,\"end\":3.35}]"} {"utterance_id": "4153-186223-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0020.flac", "task_type": "lexical_detection", "query": "brother", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"brother\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.15, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":1.8,\"end\":2.15}]"} {"utterance_id": "6267-65525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0002.flac", "task_type": "lexical_detection", "query": "vulgar", "prompt": "After the audio ends, extract all occurrences of the word \"vulgar\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.25, 5.8, "vulgar"]], "response_mode": "offline", "answer": "[{\"word\":\"vulgar\",\"start\":5.25,\"end\":5.8}]"} {"utterance_id": "1650-173552-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0003.flac", "task_type": "lexical_detection", "query": "twinkling", "prompt": "After listening to the whole recording, output every instance of the word \"twinkling\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.52, 7.09, "twinkling"]], "response_mode": "offline", "answer": "[{\"word\":\"twinkling\",\"start\":6.52,\"end\":7.09}]"} {"utterance_id": "1650-167613-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.46, 6.26, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":5.46,\"end\":6.26}]"} {"utterance_id": "1630-102884-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.37, 1.84, "buddha"]], "response_mode": "offline", "answer": "[{\"word\":\"buddha\",\"start\":1.37,\"end\":1.84}]"} {"utterance_id": "4153-186223-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0002.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen carefully to the entire clip, then report every instance of a plant name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.71, "hedge"]], "response_mode": "offline", "answer": "[{\"word\":\"hedge\",\"start\":2.31,\"end\":2.71}]"} {"utterance_id": "1255-90413-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0028.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.08, 3.59, "workmen"]], "response_mode": "offline", "answer": "[{\"word\":\"workmen\",\"start\":3.08,\"end\":3.59}]"} {"utterance_id": "1701-141760-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.51, 0.89, "boris"], [3.97, 4.41, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":0.51,\"end\":0.89},{\"word\":\"berg\",\"start\":3.97,\"end\":4.41}]"} {"utterance_id": "6455-67803-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.89, "grandmamma octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"grandmamma octavia\",\"start\":1.75,\"end\":2.89}]"} {"utterance_id": "1630-96099-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0023.flac", "task_type": "lexical_detection", "query": "threshold", "prompt": "When the utterance is complete, report all instances of the word \"threshold\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 2.42, "threshold"]], "response_mode": "offline", "answer": "[{\"word\":\"threshold\",\"start\":1.64,\"end\":2.42}]"} {"utterance_id": "8288-274150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.47, 3.44, "saint aignan"]], "response_mode": "offline", "answer": "[{\"word\":\"saint aignan\",\"start\":2.47,\"end\":3.44}]"} {"utterance_id": "3660-172182-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0037.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "You will hear a complete utterance. When it ends, list every instance of a state or province name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.8, "cornwall"]], "response_mode": "offline", "answer": "[{\"word\":\"cornwall\",\"start\":2.35,\"end\":2.8}]"} {"utterance_id": "8288-274162-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.92, 5.47, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":4.92,\"end\":5.47}]"} {"utterance_id": "4831-25894-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.52, "hand"], [4.94, 5.23, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":1.25,\"end\":1.52},{\"word\":\"face\",\"start\":4.94,\"end\":5.23}]"} {"utterance_id": "700-122868-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0003.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, list every instance of a plant name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.32, 1.59, "lily"]], "response_mode": "offline", "answer": "[{\"word\":\"lily\",\"start\":1.32,\"end\":1.59}]"} {"utterance_id": "4515-11057-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0021.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen to the full audio. After it finishes, find every instance of a disease name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.72, 3.43, "blindness"]], "response_mode": "offline", "answer": "[{\"word\":\"blindness\",\"start\":2.72,\"end\":3.43}]"} {"utterance_id": "4153-186222-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0011.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Process the full audio offline. For every occurrence of a family relation, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.93, "one"], [7.69, 8.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.53,\"end\":0.93},{\"word\":\"two\",\"start\":7.69,\"end\":8.44}]"} {"utterance_id": "5543-27761-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0043.flac", "task_type": "lexical_detection", "query": "proud", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"proud\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.66, 5.97, "proud"]], "response_mode": "offline", "answer": "[{\"word\":\"proud\",\"start\":5.66,\"end\":5.97}]"} {"utterance_id": "1686-142278-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0011.flac", "task_type": "lexical_detection", "query": "refusal", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"refusal\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.2, 5.81, "refusal"]], "response_mode": "offline", "answer": "[{\"word\":\"refusal\",\"start\":5.2,\"end\":5.81}]"} {"utterance_id": "4323-13259-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 1.11, "mister maxwell"]], "response_mode": "offline", "answer": "[{\"word\":\"mister maxwell\",\"start\":0.39,\"end\":1.11}]"} {"utterance_id": "700-122866-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0005.flac", "task_type": "lexical_detection", "query": "interests", "prompt": "When the utterance is complete, report all instances of the word \"interests\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.92, 6.65, "interests"]], "response_mode": "offline", "answer": "[{\"word\":\"interests\",\"start\":5.92,\"end\":6.65}]"} {"utterance_id": "700-122867-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.4, "matthew"]], "response_mode": "offline", "answer": "[{\"word\":\"matthew\",\"start\":1.97,\"end\":2.4}]"} {"utterance_id": "6267-53049-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0027.flac", "task_type": "lexical_detection", "query": "devoted", "prompt": "When the utterance is complete, report all instances of the word \"devoted\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.99, 5.73, "devoted"]], "response_mode": "offline", "answer": "[{\"word\":\"devoted\",\"start\":4.99,\"end\":5.73}]"} {"utterance_id": "4153-186222-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0012.flac", "task_type": "lexical_detection", "query": "earn", "prompt": "After the audio ends, extract all occurrences of the word \"earn\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.28, 5.45, "earn"]], "response_mode": "offline", "answer": "[{\"word\":\"earn\",\"start\":5.28,\"end\":5.45}]"} {"utterance_id": "1650-167613-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0051.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.53, 5.36, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":4.53,\"end\":5.36}]"} {"utterance_id": "7601-291468-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0003.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "When the utterance is complete, report how an animal name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.64, "one"], [8.74, 9.49, "two"], [10.49, 11.19, "three"], [14.22, 14.92, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.88,\"end\":2.64},{\"word\":\"two\",\"start\":8.74,\"end\":9.49},{\"word\":\"three\",\"start\":10.49,\"end\":11.19},{\"word\":\"four\",\"start\":14.22,\"end\":14.92}]"} {"utterance_id": "6267-65525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0001.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[4.13, 4.36, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.13,\"end\":4.36}]"} {"utterance_id": "116-288046-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0011.flac", "task_type": "lexical_detection", "query": "industry", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"industry\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.65, 2.18, "industry"]], "response_mode": "offline", "answer": "[{\"word\":\"industry\",\"start\":1.65,\"end\":2.18}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for an organization name with word/start/end. If none, reply [].", "labels": [[6.84, 7.14, "one"], [8.49, 8.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.84,\"end\":7.14},{\"word\":\"two\",\"start\":8.49,\"end\":8.86}]"} {"utterance_id": "6267-53049-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0026.flac", "task_type": "lexical_detection", "query": "recognize", "prompt": "After listening to the whole recording, output every instance of the word \"recognize\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.03, 5.68, "recognize"]], "response_mode": "offline", "answer": "[{\"word\":\"recognize\",\"start\":5.03,\"end\":5.68}]"} {"utterance_id": "6841-88294-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0011.flac", "task_type": "lexical_detection", "query": "sort", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"sort\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.22, 4.47, "sort"]], "response_mode": "offline", "answer": "[{\"word\":\"sort\",\"start\":4.22,\"end\":4.47}]"} {"utterance_id": "4831-25894-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0030.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.28, 2.71, "fourth"]], "response_mode": "offline", "answer": "[{\"word\":\"fourth\",\"start\":2.28,\"end\":2.71}]"} {"utterance_id": "8254-84205-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0011.flac", "task_type": "lexical_detection", "query": "plains", "prompt": "Process the full audio offline. Find every instance of the word \"plains\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.83, 3.38, "plains"]], "response_mode": "offline", "answer": "[{\"word\":\"plains\",\"start\":2.83,\"end\":3.38}]"} {"utterance_id": "4153-186222-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0021.flac", "task_type": "lexical_detection", "query": "louise", "prompt": "After the audio ends, extract all occurrences of the word \"louise\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.93, "louise"]], "response_mode": "offline", "answer": "[{\"word\":\"louise\",\"start\":0.53,\"end\":0.93}]"} {"utterance_id": "700-122868-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0017.flac", "task_type": "lexical_detection", "query": "flowery", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"flowery\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.38, 3.76, "flowery"]], "response_mode": "offline", "answer": "[{\"word\":\"flowery\",\"start\":3.38,\"end\":3.76}]"} {"utterance_id": "4831-18525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0015.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.86, 4.08, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":3.86,\"end\":4.08}]"} {"utterance_id": "4572-64670-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0009.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.0, 6.42, "one"], [8.86, 9.29, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.0,\"end\":6.42},{\"word\":\"two\",\"start\":8.86,\"end\":9.29}]"} {"utterance_id": "2506-11267-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0010.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.68, 4.15, "features"]], "response_mode": "offline", "answer": "[{\"word\":\"features\",\"start\":3.68,\"end\":4.15}]"} {"utterance_id": "7601-291468-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0002.flac", "task_type": "semantic_counting", "query": "country", "prompt": "After the audio ends, tally every instance of a country name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[3.32, 3.93, "one"], [12.94, 13.58, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.32,\"end\":3.93},{\"word\":\"two\",\"start\":12.94,\"end\":13.58}]"} {"utterance_id": "5849-50962-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0006.flac", "task_type": "lexical_detection", "query": "buoy", "prompt": "After listening to the whole recording, output every instance of the word \"buoy\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.4, 4.83, "buoy"]], "response_mode": "offline", "answer": "[{\"word\":\"buoy\",\"start\":4.4,\"end\":4.83}]"} {"utterance_id": "4323-18416-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0028.flac", "task_type": "lexical_detection", "query": "boy", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"boy\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.36, "boy"]], "response_mode": "offline", "answer": "[{\"word\":\"boy\",\"start\":1.18,\"end\":1.36}]"} {"utterance_id": "1255-74899-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0012.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[3.0, 3.53, "master"]], "response_mode": "offline", "answer": "[{\"word\":\"master\",\"start\":3.0,\"end\":3.53}]"} {"utterance_id": "1686-142278-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0032.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.13, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.9,\"end\":2.13}]"} {"utterance_id": "4153-186223-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0011.flac", "task_type": "lexical_detection", "query": "merrick", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"merrick\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.7, "merrick"]], "response_mode": "offline", "answer": "[{\"word\":\"merrick\",\"start\":2.25,\"end\":2.7}]"} {"utterance_id": "4323-55228-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0046.flac", "task_type": "lexical_detection", "query": "hannah's", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"hannah's\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.93, "hannah's"]], "response_mode": "offline", "answer": "[{\"word\":\"hannah's\",\"start\":0.62,\"end\":0.93}]"} {"utterance_id": "6267-65525-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.82, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":0.16,\"end\":0.82}]"} {"utterance_id": "3663-172528-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0006.flac", "task_type": "lexical_detection", "query": "ugly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ugly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.65, 2.14, "ugly"]], "response_mode": "offline", "answer": "[{\"word\":\"ugly\",\"start\":1.65,\"end\":2.14}]"} {"utterance_id": "7601-101619-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0003.flac", "task_type": "lexical_detection", "query": "observe", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"observe\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[16.68, 17.29, "observe"]], "response_mode": "offline", "answer": "[{\"word\":\"observe\",\"start\":16.68,\"end\":17.29}]"} {"utterance_id": "8173-294714-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0009.flac", "task_type": "lexical_detection", "query": "wrong", "prompt": "After the audio ends, extract all occurrences of the word \"wrong\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.96, "wrong"]], "response_mode": "offline", "answer": "[{\"word\":\"wrong\",\"start\":0.6,\"end\":0.96}]"} {"utterance_id": "1701-141760-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0020.flac", "task_type": "lexical_detection", "query": "getting", "prompt": "When the utterance is complete, report all instances of the word \"getting\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.18, "getting"]], "response_mode": "offline", "answer": "[{\"word\":\"getting\",\"start\":3.82,\"end\":4.18}]"} {"utterance_id": "5849-50873-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0013.flac", "task_type": "lexical_detection", "query": "charles", "prompt": "When the utterance is complete, report all instances of the word \"charles\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.6, 1.97, "charles"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":1.6,\"end\":1.97}]"} {"utterance_id": "6841-88291-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[1.87, 2.22, "cook's"]], "response_mode": "offline", "answer": "[{\"word\":\"cook's\",\"start\":1.87,\"end\":2.22}]"} {"utterance_id": "8254-84205-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0005.flac", "task_type": "lexical_detection", "query": "ned", "prompt": "After listening to the whole recording, output every instance of the word \"ned\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 1.86, "ned"]], "response_mode": "offline", "answer": "[{\"word\":\"ned\",\"start\":1.54,\"end\":1.86}]"} {"utterance_id": "7641-96684-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.99, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":0.47,\"end\":0.99}]"} {"utterance_id": "7601-175351-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.82, "warren"]], "response_mode": "offline", "answer": "[{\"word\":\"warren\",\"start\":3.21,\"end\":3.82}]"} {"utterance_id": "2506-13150-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0002.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Once the audio has finished, identify every instance of a disease name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.52, 2.19, "sick sick"]], "response_mode": "offline", "answer": "[{\"word\":\"sick sick\",\"start\":0.52,\"end\":2.19}]"} {"utterance_id": "1255-138279-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0015.flac", "task_type": "lexical_detection", "query": "noise", "prompt": "After listening to the whole recording, output every instance of the word \"noise\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.09, "noise"]], "response_mode": "offline", "answer": "[{\"word\":\"noise\",\"start\":0.76,\"end\":1.09}]"} {"utterance_id": "3663-172005-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0000.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.85, "fawn"]], "response_mode": "offline", "answer": "[{\"word\":\"fawn\",\"start\":0.47,\"end\":0.85}]"} {"utterance_id": "1701-141760-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0025.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a family relation with word/start/end. If none, reply [].", "labels": [[1.37, 1.78, "one"], [10.58, 10.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.37,\"end\":1.78},{\"word\":\"two\",\"start\":10.58,\"end\":10.91}]"} {"utterance_id": "1650-173552-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[12.03, 12.87, "alexander"]], "response_mode": "offline", "answer": "[{\"word\":\"alexander\",\"start\":12.03,\"end\":12.87}]"} {"utterance_id": "1686-142278-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0054.flac", "task_type": "lexical_detection", "query": "indifferent", "prompt": "Process the full audio offline. Find every instance of the word \"indifferent\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.29, 3.8, "indifferent"]], "response_mode": "offline", "answer": "[{\"word\":\"indifferent\",\"start\":3.29,\"end\":3.8}]"} {"utterance_id": "3915-98647-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0025.flac", "task_type": "lexical_detection", "query": "renewed", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"renewed\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.55, 6.95, "renewed"]], "response_mode": "offline", "answer": "[{\"word\":\"renewed\",\"start\":6.55,\"end\":6.95}]"} {"utterance_id": "6267-65525-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0016.flac", "task_type": "lexical_detection", "query": "enough", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"enough\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.04, 1.47, "enough"]], "response_mode": "offline", "answer": "[{\"word\":\"enough\",\"start\":1.04,\"end\":1.47}]"} {"utterance_id": "4153-186222-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0017.flac", "task_type": "lexical_counting", "query": "you're", "prompt": "After listening to the whole recording, output timed count words for the word \"you're\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.38, 0.53, "one"], [3.26, 3.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.38,\"end\":0.53},{\"word\":\"two\",\"start\":3.26,\"end\":3.36}]"} {"utterance_id": "5849-50963-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0009.flac", "task_type": "semantic_detection", "query": "money", "prompt": "When the utterance is complete, report all instances of a money or currency term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.65, 12.04, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":11.65,\"end\":12.04}]"} {"utterance_id": "6455-67804-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0034.flac", "task_type": "lexical_detection", "query": "ladyship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ladyship\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.88, 2.44, "ladyship"]], "response_mode": "offline", "answer": "[{\"word\":\"ladyship\",\"start\":1.88,\"end\":2.44}]"} {"utterance_id": "1630-96099-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0002.flac", "task_type": "lexical_detection", "query": "evident", "prompt": "Once the audio has finished, identify every instance of the word \"evident\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.19, 2.63, "evident"]], "response_mode": "offline", "answer": "[{\"word\":\"evident\",\"start\":2.19,\"end\":2.63}]"} {"utterance_id": "1650-157641-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0007.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.09, 4.73, "tobacco"]], "response_mode": "offline", "answer": "[{\"word\":\"tobacco\",\"start\":4.09,\"end\":4.73}]"} {"utterance_id": "4831-25894-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0027.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.64, 2.31, "one"], [3.87, 4.32, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.64,\"end\":2.31},{\"word\":\"two\",\"start\":3.87,\"end\":4.32}]"} {"utterance_id": "116-288047-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0003.flac", "task_type": "lexical_detection", "query": "mythus", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"mythus\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.54, "mythus"]], "response_mode": "offline", "answer": "[{\"word\":\"mythus\",\"start\":0.87,\"end\":1.54}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "When the utterance is complete, report all instances of an organization name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.84, 7.14, "louvre"], [8.49, 8.86, "louvre"]], "response_mode": "offline", "answer": "[{\"word\":\"louvre\",\"start\":6.84,\"end\":7.14},{\"word\":\"louvre\",\"start\":8.49,\"end\":8.86}]"} {"utterance_id": "6267-53049-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0018.flac", "task_type": "lexical_detection", "query": "heirloom", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"heirloom\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.35, 3.09, "heirloom"]], "response_mode": "offline", "answer": "[{\"word\":\"heirloom\",\"start\":2.35,\"end\":3.09}]"} {"utterance_id": "6841-88294-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0032.flac", "task_type": "lexical_detection", "query": "empty", "prompt": "Once the audio has finished, identify every instance of the word \"empty\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.92, 2.19, "empty"]], "response_mode": "offline", "answer": "[{\"word\":\"empty\",\"start\":1.92,\"end\":2.19}]"} {"utterance_id": "6267-65525-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0046.flac", "task_type": "lexical_detection", "query": "keep", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"keep\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.49, 2.79, "keep"]], "response_mode": "offline", "answer": "[{\"word\":\"keep\",\"start\":2.49,\"end\":2.79}]"} {"utterance_id": "4572-112375-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0008.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.5, "one"], [3.21, 3.65, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.96,\"end\":1.5},{\"word\":\"two\",\"start\":3.21,\"end\":3.65}]"} {"utterance_id": "6123-59186-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0003.flac", "task_type": "lexical_detection", "query": "express", "prompt": "Once the audio has finished, identify every instance of the word \"express\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.93, 1.5, "express"]], "response_mode": "offline", "answer": "[{\"word\":\"express\",\"start\":0.93,\"end\":1.5}]"} {"utterance_id": "4831-18525-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0016.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.25, "white"]], "response_mode": "offline", "answer": "[{\"word\":\"white\",\"start\":2.0,\"end\":2.25}]"} {"utterance_id": "4323-55228-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0010.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.49, 5.72, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"lips\",\"start\":5.49,\"end\":5.72}]"} {"utterance_id": "6467-97061-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0007.flac", "task_type": "lexical_detection", "query": "pity", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"pity\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.57, "pity"]], "response_mode": "offline", "answer": "[{\"word\":\"pity\",\"start\":1.15,\"end\":1.57}]"} {"utterance_id": "5849-50873-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0034.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Process the full audio offline. Find every instance of an organization name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.74, "assembly"]], "response_mode": "offline", "answer": "[{\"word\":\"assembly\",\"start\":1.33,\"end\":1.74}]"} {"utterance_id": "1630-73710-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0004.flac", "task_type": "lexical_detection", "query": "mine", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"mine\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.55, 1.02, "mine"]], "response_mode": "offline", "answer": "[{\"word\":\"mine\",\"start\":0.55,\"end\":1.02}]"} {"utterance_id": "4153-186223-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.12, 7.37, "lap"]], "response_mode": "offline", "answer": "[{\"word\":\"lap\",\"start\":7.12,\"end\":7.37}]"} {"utterance_id": "1651-136854-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0021.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Listen to the full audio. Count every occurrence of a family relation. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.47, "one"], [4.28, 4.79, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.04,\"end\":1.47},{\"word\":\"two\",\"start\":4.28,\"end\":4.79}]"} {"utterance_id": "5543-27761-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0005.flac", "task_type": "lexical_detection", "query": "cold", "prompt": "Process the full audio offline. Find every instance of the word \"cold\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.2, "cold"]], "response_mode": "offline", "answer": "[{\"word\":\"cold\",\"start\":0.78,\"end\":1.2}]"} {"utterance_id": "1650-167613-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0041.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "After listening to the whole recording, output every instance of a disease name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.07, 12.47, "cramp"]], "response_mode": "offline", "answer": "[{\"word\":\"cramp\",\"start\":12.07,\"end\":12.47}]"} {"utterance_id": "6455-67803-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0026.flac", "task_type": "lexical_detection", "query": "calmly", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"calmly\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 1.93, "calmly"]], "response_mode": "offline", "answer": "[{\"word\":\"calmly\",\"start\":1.43,\"end\":1.93}]"} {"utterance_id": "4572-64670-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0003.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a body part with word/start/end. If none, reply [].", "labels": [[1.37, 1.71, "one"], [4.17, 4.51, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.37,\"end\":1.71},{\"word\":\"two\",\"start\":4.17,\"end\":4.51}]"} {"utterance_id": "8254-115543-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0013.flac", "task_type": "lexical_detection", "query": "beggars", "prompt": "Process the full audio offline. Find every instance of the word \"beggars\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.86, 3.34, "beggars"]], "response_mode": "offline", "answer": "[{\"word\":\"beggars\",\"start\":2.86,\"end\":3.34}]"} {"utterance_id": "116-288045-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0031.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[4.86, 5.32, "one"], [6.35, 7.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.86,\"end\":5.32},{\"word\":\"two\",\"start\":6.35,\"end\":7.02}]"} {"utterance_id": "1630-141772-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0010.flac", "task_type": "lexical_detection", "query": "napoleon", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"napoleon\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.67, 3.26, "napoleon"]], "response_mode": "offline", "answer": "[{\"word\":\"napoleon\",\"start\":2.67,\"end\":3.26}]"} {"utterance_id": "1585-157660-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.54, "william"]], "response_mode": "offline", "answer": "[{\"word\":\"william\",\"start\":3.14,\"end\":3.54}]"} {"utterance_id": "6123-59186-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0015.flac", "task_type": "lexical_detection", "query": "sense", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"sense\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.89, 1.14, "sense"]], "response_mode": "offline", "answer": "[{\"word\":\"sense\",\"start\":0.89,\"end\":1.14}]"} {"utterance_id": "4831-29134-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0017.flac", "task_type": "lexical_detection", "query": "honest", "prompt": "After listening to the whole recording, output every instance of the word \"honest\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.76, 2.36, "honest"]], "response_mode": "offline", "answer": "[{\"word\":\"honest\",\"start\":1.76,\"end\":2.36}]"} {"utterance_id": "6841-88294-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0021.flac", "task_type": "lexical_detection", "query": "minute", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"minute\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.82, "minute"]], "response_mode": "offline", "answer": "[{\"word\":\"minute\",\"start\":0.52,\"end\":0.82}]"} {"utterance_id": "8254-115543-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0037.flac", "task_type": "lexical_detection", "query": "cried", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"cried\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.26, 1.66, "cried"]], "response_mode": "offline", "answer": "[{\"word\":\"cried\",\"start\":1.26,\"end\":1.66}]"} {"utterance_id": "8254-84205-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0012.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After listening to the whole recording, output every instance of a time of day with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.56, "daybreak"]], "response_mode": "offline", "answer": "[{\"word\":\"daybreak\",\"start\":2.03,\"end\":2.56}]"} {"utterance_id": "4153-186223-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0007.flac", "task_type": "lexical_counting", "query": "aunt", "prompt": "Once the audio has finished, list each occurrence of the word \"aunt\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.4, 0.67, "one"], [2.4, 2.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.4,\"end\":0.67},{\"word\":\"two\",\"start\":2.4,\"end\":2.63}]"} {"utterance_id": "3663-172528-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0017.flac", "task_type": "lexical_detection", "query": "cure", "prompt": "After listening to the whole recording, output every instance of the word \"cure\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.69, "cure"]], "response_mode": "offline", "answer": "[{\"word\":\"cure\",\"start\":2.25,\"end\":2.69}]"} {"utterance_id": "4153-185072-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0008.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.43, 7.22, "mount morris"]], "response_mode": "offline", "answer": "[{\"word\":\"mount morris\",\"start\":6.43,\"end\":7.22}]"} {"utterance_id": "700-122868-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0015.flac", "task_type": "lexical_detection", "query": "scream", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"scream\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.9, 2.51, "scream"]], "response_mode": "offline", "answer": "[{\"word\":\"scream\",\"start\":1.9,\"end\":2.51}]"} {"utterance_id": "5849-50873-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0029.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.85, "soldiers"], [8.81, 9.52, "soldiers"]], "response_mode": "offline", "answer": "[{\"word\":\"soldiers\",\"start\":0.33,\"end\":0.85},{\"word\":\"soldiers\",\"start\":8.81,\"end\":9.52}]"} {"utterance_id": "116-288045-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0030.flac", "task_type": "lexical_detection", "query": "quivering", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"quivering\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.2, 3.54, "quivering"]], "response_mode": "offline", "answer": "[{\"word\":\"quivering\",\"start\":3.2,\"end\":3.54}]"} {"utterance_id": "5849-50962-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0025.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen carefully to the entire clip, then report every instance of a time of day as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.17, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":0.69,\"end\":1.17}]"} {"utterance_id": "6841-88291-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0011.flac", "task_type": "lexical_detection", "query": "wooden", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"wooden\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.09, "wooden"]], "response_mode": "offline", "answer": "[{\"word\":\"wooden\",\"start\":0.65,\"end\":1.09}]"} {"utterance_id": "5849-50963-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0012.flac", "task_type": "lexical_detection", "query": "remove", "prompt": "After listening to the whole recording, output every instance of the word \"remove\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.37, "remove"]], "response_mode": "offline", "answer": "[{\"word\":\"remove\",\"start\":1.11,\"end\":1.37}]"} {"utterance_id": "8288-274150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.64, "malicorne"]], "response_mode": "offline", "answer": "[{\"word\":\"malicorne\",\"start\":0.91,\"end\":1.64}]"} {"utterance_id": "5849-50963-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 4.81, "eyes"], [7.01, 7.37, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":4.52,\"end\":4.81},{\"word\":\"face\",\"start\":7.01,\"end\":7.37}]"} {"utterance_id": "8254-115543-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0029.flac", "task_type": "lexical_detection", "query": "plates", "prompt": "After listening to the whole recording, output every instance of the word \"plates\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 1.88, "plates"]], "response_mode": "offline", "answer": "[{\"word\":\"plates\",\"start\":1.42,\"end\":1.88}]"} {"utterance_id": "700-122868-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0031.flac", "task_type": "lexical_detection", "query": "laughable", "prompt": "Process the full audio offline. Find every instance of the word \"laughable\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.11, 3.6, "laughable"]], "response_mode": "offline", "answer": "[{\"word\":\"laughable\",\"start\":3.11,\"end\":3.6}]"} {"utterance_id": "7641-96670-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0013.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.02, "one"], [3.77, 4.07, "two"], [5.12, 5.59, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.77,\"end\":1.02},{\"word\":\"two\",\"start\":3.77,\"end\":4.07},{\"word\":\"three\",\"start\":5.12,\"end\":5.59}]"} {"utterance_id": "1255-90413-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.04, 1.42, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":1.04,\"end\":1.42}]"} {"utterance_id": "5849-50873-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0023.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.17, "four"], [5.78, 6.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":2.98,\"end\":3.17},{\"word\":\"two\",\"start\":5.78,\"end\":6.02}]"} {"utterance_id": "8254-115543-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0045.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.0, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.68,\"end\":2.0}]"} {"utterance_id": "1650-167613-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0039.flac", "task_type": "lexical_detection", "query": "people's", "prompt": "When the utterance is complete, report all instances of the word \"people's\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.35, 1.88, "people's"]], "response_mode": "offline", "answer": "[{\"word\":\"people's\",\"start\":1.35,\"end\":1.88}]"} {"utterance_id": "3660-172182-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0000.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.88, 2.39, "stag"]], "response_mode": "offline", "answer": "[{\"word\":\"stag\",\"start\":1.88,\"end\":2.39}]"} {"utterance_id": "700-122866-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0019.flac", "task_type": "lexical_detection", "query": "rival", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"rival\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.85, "rival"]], "response_mode": "offline", "answer": "[{\"word\":\"rival\",\"start\":1.38,\"end\":1.85}]"} {"utterance_id": "3915-98647-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0026.flac", "task_type": "lexical_detection", "query": "defend", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"defend\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.41, 2.86, "defend"]], "response_mode": "offline", "answer": "[{\"word\":\"defend\",\"start\":2.41,\"end\":2.86}]"} {"utterance_id": "6455-67803-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.77, 2.28, "lucia's"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia's\",\"start\":1.77,\"end\":2.28}]"} {"utterance_id": "4570-56594-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.0, 6.51, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":6.0,\"end\":6.51}]"} {"utterance_id": "6841-88294-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.03, 5.44, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":5.03,\"end\":5.44}]"} {"utterance_id": "8254-84205-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0045.flac", "task_type": "lexical_detection", "query": "shooting", "prompt": "After the audio ends, extract all occurrences of the word \"shooting\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.29, 5.75, "shooting"]], "response_mode": "offline", "answer": "[{\"word\":\"shooting\",\"start\":5.29,\"end\":5.75}]"} {"utterance_id": "1255-90413-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0023.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.05, 1.43, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":1.05,\"end\":1.43}]"} {"utterance_id": "1255-138279-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0001.flac", "task_type": "semantic_counting", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a plant name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[13.3, 13.81, "one"], [14.83, 15.27, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":13.3,\"end\":13.81},{\"word\":\"two\",\"start\":14.83,\"end\":15.27}]"} {"utterance_id": "8173-294714-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0024.flac", "task_type": "lexical_counting", "query": "how", "prompt": "Once the audio has finished, list each occurrence of the word \"how\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.19, 0.44, "one"], [1.45, 1.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.19,\"end\":0.44},{\"word\":\"two\",\"start\":1.45,\"end\":1.7}]"} {"utterance_id": "6267-65525-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0005.flac", "task_type": "lexical_detection", "query": "alike", "prompt": "Once the audio has finished, identify every instance of the word \"alike\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.66, 2.05, "alike"]], "response_mode": "offline", "answer": "[{\"word\":\"alike\",\"start\":1.66,\"end\":2.05}]"} {"utterance_id": "1650-167613-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0017.flac", "task_type": "lexical_detection", "query": "commander", "prompt": "After the audio ends, extract all occurrences of the word \"commander\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.77, "commander"]], "response_mode": "offline", "answer": "[{\"word\":\"commander\",\"start\":1.09,\"end\":1.77}]"} {"utterance_id": "700-122867-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0014.flac", "task_type": "lexical_detection", "query": "shirley", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"shirley\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.68, 1.26, "shirley"]], "response_mode": "offline", "answer": "[{\"word\":\"shirley\",\"start\":0.68,\"end\":1.26}]"} {"utterance_id": "5849-50963-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0000.flac", "task_type": "lexical_detection", "query": "dreadful", "prompt": "When the utterance is complete, report all instances of the word \"dreadful\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.78, "dreadful"]], "response_mode": "offline", "answer": "[{\"word\":\"dreadful\",\"start\":2.35,\"end\":2.78}]"} {"utterance_id": "700-122867-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0038.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a body part: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.47, 1.67, "one"], [4.78, 5.06, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.47,\"end\":1.67},{\"word\":\"two\",\"start\":4.78,\"end\":5.06}]"} {"utterance_id": "6455-66379-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0002.flac", "task_type": "lexical_detection", "query": "deeply", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"deeply\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.34, 6.65, "deeply"]], "response_mode": "offline", "answer": "[{\"word\":\"deeply\",\"start\":6.34,\"end\":6.65}]"} {"utterance_id": "1255-90413-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0020.flac", "task_type": "lexical_detection", "query": "avon", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"avon\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.17, "avon"]], "response_mode": "offline", "answer": "[{\"word\":\"avon\",\"start\":0.98,\"end\":1.17}]"} {"utterance_id": "700-122868-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0025.flac", "task_type": "lexical_detection", "query": "difficult", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"difficult\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.88, 2.35, "difficult"]], "response_mode": "offline", "answer": "[{\"word\":\"difficult\",\"start\":1.88,\"end\":2.35}]"} {"utterance_id": "3660-172182-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0027.flac", "task_type": "lexical_detection", "query": "kingdom", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"kingdom\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.93, 6.29, "kingdom"]], "response_mode": "offline", "answer": "[{\"word\":\"kingdom\",\"start\":5.93,\"end\":6.29}]"} {"utterance_id": "1686-142278-0073", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0073.flac", "task_type": "lexical_detection", "query": "bell", "prompt": "After the audio ends, extract all occurrences of the word \"bell\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.55, 8.93, "bell"], [10.78, 11.02, "bell"]], "response_mode": "offline", "answer": "[{\"word\":\"bell\",\"start\":8.55,\"end\":8.93},{\"word\":\"bell\",\"start\":10.78,\"end\":11.02}]"} {"utterance_id": "4153-186222-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0005.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 0.91, "two"], [3.77, 4.0, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":0.71,\"end\":0.91},{\"word\":\"one\",\"start\":3.77,\"end\":4.0}]"} {"utterance_id": "5849-50964-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.26, 1.63, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":1.26,\"end\":1.63}]"} {"utterance_id": "7697-105815-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0042.flac", "task_type": "lexical_detection", "query": "bent", "prompt": "Process the full audio offline. Find every instance of the word \"bent\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.35, "bent"]], "response_mode": "offline", "answer": "[{\"word\":\"bent\",\"start\":1.14,\"end\":1.35}]"} {"utterance_id": "5849-50962-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0009.flac", "task_type": "lexical_detection", "query": "minturn", "prompt": "After listening to the whole recording, output every instance of the word \"minturn\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.87, 2.48, "minturn"]], "response_mode": "offline", "answer": "[{\"word\":\"minturn\",\"start\":1.87,\"end\":2.48}]"} {"utterance_id": "6841-88294-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0055.flac", "task_type": "lexical_detection", "query": "treasure", "prompt": "After the audio ends, extract all occurrences of the word \"treasure\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.03, 1.37, "treasure"]], "response_mode": "offline", "answer": "[{\"word\":\"treasure\",\"start\":1.03,\"end\":1.37}]"} {"utterance_id": "7601-101622-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0000.flac", "task_type": "lexical_detection", "query": "lands", "prompt": "When the utterance is complete, report all instances of the word \"lands\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[15.19, 15.65, "lands"]], "response_mode": "offline", "answer": "[{\"word\":\"lands\",\"start\":15.19,\"end\":15.65}]"} {"utterance_id": "8254-84205-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.45, "ned"]], "response_mode": "offline", "answer": "[{\"word\":\"ned\",\"start\":2.26,\"end\":2.45}]"} {"utterance_id": "4831-25894-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 2.31, "fairy prince"], [3.87, 4.32, "tessa's"]], "response_mode": "offline", "answer": "[{\"word\":\"fairy prince\",\"start\":1.64,\"end\":2.31},{\"word\":\"tessa's\",\"start\":3.87,\"end\":4.32}]"} {"utterance_id": "8288-274150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0008.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[3.17, 3.56, "knees"]], "response_mode": "offline", "answer": "[{\"word\":\"knees\",\"start\":3.17,\"end\":3.56}]"} {"utterance_id": "8254-84205-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0010.flac", "task_type": "lexical_detection", "query": "i'm", "prompt": "Process the full audio offline. Find every instance of the word \"i'm\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 0.87, "i'm"]], "response_mode": "offline", "answer": "[{\"word\":\"i'm\",\"start\":0.69,\"end\":0.87}]"} {"utterance_id": "1255-138279-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0007.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 4.94, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.52,\"end\":4.94}]"} {"utterance_id": "6599-38591-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.38, 1.1, "montmorency"]], "response_mode": "offline", "answer": "[{\"word\":\"montmorency\",\"start\":0.38,\"end\":1.1}]"} {"utterance_id": "6455-67804-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0000.flac", "task_type": "lexical_detection", "query": "ladyship", "prompt": "After the audio ends, extract all occurrences of the word \"ladyship\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.68, 4.05, "ladyship"]], "response_mode": "offline", "answer": "[{\"word\":\"ladyship\",\"start\":3.68,\"end\":4.05}]"} {"utterance_id": "5849-50962-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0018.flac", "task_type": "lexical_detection", "query": "george", "prompt": "Process the full audio offline. Find every instance of the word \"george\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.06, "george"]], "response_mode": "offline", "answer": "[{\"word\":\"george\",\"start\":0.77,\"end\":1.06}]"} {"utterance_id": "1686-142278-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0064.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.87, "mamma"]], "response_mode": "offline", "answer": "[{\"word\":\"mamma\",\"start\":0.53,\"end\":0.87}]"} {"utterance_id": "7697-245712-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0004.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.18, 0.81, "thirdly"]], "response_mode": "offline", "answer": "[{\"word\":\"thirdly\",\"start\":0.18,\"end\":0.81}]"} {"utterance_id": "5849-50964-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.21, 3.72, "dorothy"]], "response_mode": "offline", "answer": "[{\"word\":\"dorothy\",\"start\":3.21,\"end\":3.72}]"} {"utterance_id": "4570-24733-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0008.flac", "task_type": "lexical_counting", "query": "attributes", "prompt": "When the utterance is complete, report how the word \"attributes\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.4, 7.21, "one"], [10.0, 10.82, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.4,\"end\":7.21},{\"word\":\"two\",\"start\":10.0,\"end\":10.82}]"} {"utterance_id": "1701-141760-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.19, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"rostov\",\"start\":1.77,\"end\":2.19}]"} {"utterance_id": "8254-84205-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0039.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[1.67, 2.08, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.67,\"end\":2.08}]"} {"utterance_id": "1255-90413-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0011.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[8.33, 8.68, "one"], [13.56, 14.06, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.33,\"end\":8.68},{\"word\":\"two\",\"start\":13.56,\"end\":14.06}]"} {"utterance_id": "3915-98647-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.47, "mother"], [5.58, 6.0, "child"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":2.12,\"end\":2.47},{\"word\":\"child\",\"start\":5.58,\"end\":6.0}]"} {"utterance_id": "4515-11057-0100", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0100.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Do not answer until the audio ends. Then list each instance of a weather term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.67, 0.99, "rain"], [5.41, 5.93, "mist"]], "response_mode": "offline", "answer": "[{\"word\":\"rain\",\"start\":0.67,\"end\":0.99},{\"word\":\"mist\",\"start\":5.41,\"end\":5.93}]"} {"utterance_id": "5849-50873-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0002.flac", "task_type": "lexical_detection", "query": "treat", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"treat\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.85, 5.22, "treat"]], "response_mode": "offline", "answer": "[{\"word\":\"treat\",\"start\":4.85,\"end\":5.22}]"} {"utterance_id": "5849-50964-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0004.flac", "task_type": "lexical_detection", "query": "hand", "prompt": "After the audio ends, extract all occurrences of the word \"hand\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.63, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":1.26,\"end\":1.63}]"} {"utterance_id": "4323-18416-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0021.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[3.7, 4.26, "one"], [6.02, 6.47, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.7,\"end\":4.26},{\"word\":\"two\",\"start\":6.02,\"end\":6.47}]"} {"utterance_id": "4323-18416-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0026.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.28, 3.52, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"lips\",\"start\":3.28,\"end\":3.52}]"} {"utterance_id": "6455-67803-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0035.flac", "task_type": "lexical_detection", "query": "raised", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"raised\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.65, "raised"]], "response_mode": "offline", "answer": "[{\"word\":\"raised\",\"start\":0.45,\"end\":0.65}]"} {"utterance_id": "1630-96099-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0011.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.03, 11.44, "note"]], "response_mode": "offline", "answer": "[{\"word\":\"note\",\"start\":11.03,\"end\":11.44}]"} {"utterance_id": "1686-142278-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0019.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[1.36, 2.0, "clergyman"]], "response_mode": "offline", "answer": "[{\"word\":\"clergyman\",\"start\":1.36,\"end\":2.0}]"} {"utterance_id": "6123-59150-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0031.flac", "task_type": "lexical_detection", "query": "tavern", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"tavern\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.81, 9.32, "tavern"]], "response_mode": "offline", "answer": "[{\"word\":\"tavern\",\"start\":8.81,\"end\":9.32}]"} {"utterance_id": "7697-105815-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0051.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen carefully to the entire clip, then report every instance of a city name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.72, 9.1, "gath"]], "response_mode": "offline", "answer": "[{\"word\":\"gath\",\"start\":8.72,\"end\":9.1}]"} {"utterance_id": "4323-18416-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0004.flac", "task_type": "lexical_detection", "query": "dinner", "prompt": "After the audio ends, extract all occurrences of the word \"dinner\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.52, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":1.21,\"end\":1.52}]"} {"utterance_id": "116-288048-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0023.flac", "task_type": "lexical_detection", "query": "impossible", "prompt": "After listening to the whole recording, output every instance of the word \"impossible\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.44, "impossible"]], "response_mode": "offline", "answer": "[{\"word\":\"impossible\",\"start\":0.8,\"end\":1.44}]"} {"utterance_id": "1585-131718-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0012.flac", "task_type": "lexical_detection", "query": "association", "prompt": "After listening to the whole recording, output every instance of the word \"association\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 5.38, "association"]], "response_mode": "offline", "answer": "[{\"word\":\"association\",\"start\":4.52,\"end\":5.38}]"} {"utterance_id": "6467-97061-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0010.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 2.86, "elephant"], [13.65, 14.08, "elephant"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant\",\"start\":2.4,\"end\":2.86},{\"word\":\"elephant\",\"start\":13.65,\"end\":14.08}]"} {"utterance_id": "6455-67804-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 2.36, "martin bassett"]], "response_mode": "offline", "answer": "[{\"word\":\"martin bassett\",\"start\":1.54,\"end\":2.36}]"} {"utterance_id": "1650-167613-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0037.flac", "task_type": "lexical_detection", "query": "muttered", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"muttered\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.38, 1.9, "muttered"]], "response_mode": "offline", "answer": "[{\"word\":\"muttered\",\"start\":1.38,\"end\":1.9}]"} {"utterance_id": "7697-105815-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0045.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.42, 6.26, "mount zion"]], "response_mode": "offline", "answer": "[{\"word\":\"mount zion\",\"start\":5.42,\"end\":6.26}]"} {"utterance_id": "4570-56594-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0018.flac", "task_type": "lexical_detection", "query": "open", "prompt": "Once the audio has finished, identify every instance of the word \"open\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.32, 1.67, "open"]], "response_mode": "offline", "answer": "[{\"word\":\"open\",\"start\":1.32,\"end\":1.67}]"} {"utterance_id": "4572-112375-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.96, 1.5, "sugriva"], [3.21, 3.65, "rama"]], "response_mode": "offline", "answer": "[{\"word\":\"sugriva\",\"start\":0.96,\"end\":1.5},{\"word\":\"rama\",\"start\":3.21,\"end\":3.65}]"} {"utterance_id": "4572-112383-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0005.flac", "task_type": "lexical_detection", "query": "spirits", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"spirits\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.39, "spirits"]], "response_mode": "offline", "answer": "[{\"word\":\"spirits\",\"start\":0.82,\"end\":1.39}]"} {"utterance_id": "4323-55228-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0052.flac", "task_type": "lexical_detection", "query": "account", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"account\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.29, "account"]], "response_mode": "offline", "answer": "[{\"word\":\"account\",\"start\":2.03,\"end\":2.29}]"} {"utterance_id": "1701-141759-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.52, 1.07, "natasha"]], "response_mode": "offline", "answer": "[{\"word\":\"natasha\",\"start\":0.52,\"end\":1.07}]"} {"utterance_id": "4153-186222-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0028.flac", "task_type": "lexical_detection", "query": "invite", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"invite\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.56, "invite"]], "response_mode": "offline", "answer": "[{\"word\":\"invite\",\"start\":1.14,\"end\":1.56}]"} {"utterance_id": "7641-96252-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0014.flac", "task_type": "lexical_detection", "query": "laws", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"laws\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.02, 6.42, "laws"]], "response_mode": "offline", "answer": "[{\"word\":\"laws\",\"start\":6.02,\"end\":6.42}]"} {"utterance_id": "8254-84205-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0021.flac", "task_type": "lexical_detection", "query": "quarrel", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"quarrel\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.67, "quarrel"]], "response_mode": "offline", "answer": "[{\"word\":\"quarrel\",\"start\":2.22,\"end\":2.67}]"} {"utterance_id": "7601-101622-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0004.flac", "task_type": "lexical_counting", "query": "thy", "prompt": "Listen to the full audio. Count every occurrence of the word \"thy\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.09, 3.32, "one"], [5.59, 5.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.09,\"end\":3.32},{\"word\":\"two\",\"start\":5.59,\"end\":5.86}]"} {"utterance_id": "1650-173552-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0004.flac", "task_type": "lexical_detection", "query": "spare", "prompt": "Process the full audio offline. Find every instance of the word \"spare\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.03, 1.5, "spare"]], "response_mode": "offline", "answer": "[{\"word\":\"spare\",\"start\":1.03,\"end\":1.5}]"} {"utterance_id": "6123-59150-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0009.flac", "task_type": "lexical_detection", "query": "action", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"action\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.59, 3.08, "action"]], "response_mode": "offline", "answer": "[{\"word\":\"action\",\"start\":2.59,\"end\":3.08}]"} {"utterance_id": "7697-105815-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.43, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.94,\"end\":1.43}]"} {"utterance_id": "1585-131718-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0023.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a money or currency term with fields word, start, end (seconds). If none, reply [].", "labels": [[1.22, 1.59, "cents"]], "response_mode": "offline", "answer": "[{\"word\":\"cents\",\"start\":1.22,\"end\":1.59}]"} {"utterance_id": "3660-172182-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0034.flac", "task_type": "lexical_detection", "query": "feeble", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"feeble\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.86, 3.28, "feeble"]], "response_mode": "offline", "answer": "[{\"word\":\"feeble\",\"start\":2.86,\"end\":3.28}]"} {"utterance_id": "7601-175351-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0018.flac", "task_type": "lexical_detection", "query": "sun", "prompt": "When the utterance is complete, report all instances of the word \"sun\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.73, "sun"]], "response_mode": "offline", "answer": "[{\"word\":\"sun\",\"start\":1.37,\"end\":1.73}]"} {"utterance_id": "2506-11267-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.2, 9.29, "historiographer"]], "response_mode": "offline", "answer": "[{\"word\":\"historiographer\",\"start\":8.2,\"end\":9.29}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[23.06, 23.79, "morsel"]], "response_mode": "offline", "answer": "[{\"word\":\"morsel\",\"start\":23.06,\"end\":23.79}]"} {"utterance_id": "4153-186223-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0017.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the full audio. Count every occurrence of a number. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.75, 6.92, "one"], [7.94, 8.39, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.75,\"end\":6.92},{\"word\":\"two\",\"start\":7.94,\"end\":8.39}]"} {"utterance_id": "700-122868-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.97, "maid"]], "response_mode": "offline", "answer": "[{\"word\":\"maid\",\"start\":1.59,\"end\":1.97}]"} {"utterance_id": "1585-157660-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0007.flac", "task_type": "lexical_counting", "query": "expectations", "prompt": "Listen to the entire clip, then report the count sequence for the word \"expectations\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.01, 4.03, "one"], [8.12, 9.1, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.01,\"end\":4.03},{\"word\":\"two\",\"start\":8.12,\"end\":9.1}]"} {"utterance_id": "1630-96099-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0009.flac", "task_type": "lexical_detection", "query": "restaurant", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"restaurant\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 2.04, "restaurant"]], "response_mode": "offline", "answer": "[{\"word\":\"restaurant\",\"start\":1.38,\"end\":2.04}]"} {"utterance_id": "6467-94831-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0024.flac", "task_type": "lexical_detection", "query": "joke", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"joke\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.47, "joke"]], "response_mode": "offline", "answer": "[{\"word\":\"joke\",\"start\":1.02,\"end\":1.47}]"} {"utterance_id": "7641-96670-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0000.flac", "task_type": "lexical_detection", "query": "sky", "prompt": "After the audio ends, extract all occurrences of the word \"sky\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.89, 4.4, "sky"]], "response_mode": "offline", "answer": "[{\"word\":\"sky\",\"start\":3.89,\"end\":4.4}]"} {"utterance_id": "4572-112375-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.06, 4.62, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.06,\"end\":4.62}]"} {"utterance_id": "5849-50963-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.39, 4.1, "mc laughlin"]], "response_mode": "offline", "answer": "[{\"word\":\"mc laughlin\",\"start\":3.39,\"end\":4.1}]"} {"utterance_id": "6467-94831-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 0.81, "sam"]], "response_mode": "offline", "answer": "[{\"word\":\"sam\",\"start\":0.59,\"end\":0.81}]"} {"utterance_id": "7601-175351-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0012.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.54, 1.37, "engineer"]], "response_mode": "offline", "answer": "[{\"word\":\"engineer\",\"start\":0.54,\"end\":1.37}]"} {"utterance_id": "6467-97061-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0022.flac", "task_type": "lexical_detection", "query": "front", "prompt": "When the utterance is complete, report all instances of the word \"front\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.95, 7.18, "front"]], "response_mode": "offline", "answer": "[{\"word\":\"front\",\"start\":6.95,\"end\":7.18}]"} {"utterance_id": "6599-38591-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0006.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.36, 4.79, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":4.36,\"end\":4.79}]"} {"utterance_id": "1686-142278-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0049.flac", "task_type": "lexical_detection", "query": "indeed", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"indeed\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.7, "indeed"]], "response_mode": "offline", "answer": "[{\"word\":\"indeed\",\"start\":1.27,\"end\":1.7}]"} {"utterance_id": "6467-62797-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.41, "breast"]], "response_mode": "offline", "answer": "[{\"word\":\"breast\",\"start\":1.95,\"end\":2.41}]"} {"utterance_id": "1686-142278-0078", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0078.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 1.09, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.66,\"end\":1.09}]"} {"utterance_id": "3663-172528-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0004.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "When the utterance is complete, report how a number accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.87, 2.38, "one"], [9.36, 9.52, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.87,\"end\":2.38},{\"word\":\"two\",\"start\":9.36,\"end\":9.52}]"} {"utterance_id": "1585-131718-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0012.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for an organization name with word/start/end. If none, reply [].", "labels": [[3.2, 5.38, "one"], [10.63, 12.07, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.2,\"end\":5.38},{\"word\":\"two\",\"start\":10.63,\"end\":12.07}]"} {"utterance_id": "3660-6517-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0012.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.11, "forage"]], "response_mode": "offline", "answer": "[{\"word\":\"forage\",\"start\":0.52,\"end\":1.11}]"} {"utterance_id": "7601-175351-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0005.flac", "task_type": "lexical_detection", "query": "hear", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"hear\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.01, 1.36, "hear"]], "response_mode": "offline", "answer": "[{\"word\":\"hear\",\"start\":1.01,\"end\":1.36}]"} {"utterance_id": "700-122867-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[2.18, 2.55, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":2.18,\"end\":2.55}]"} {"utterance_id": "1255-74899-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.72, 4.38, "aunt julia"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt julia\",\"start\":3.72,\"end\":4.38}]"} {"utterance_id": "2506-11278-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.03, 5.23, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.03,\"end\":5.23}]"} {"utterance_id": "7697-245712-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0017.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.18,\"end\":1.63}]"} {"utterance_id": "5849-50964-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0011.flac", "task_type": "lexical_detection", "query": "packing", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"packing\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.73, 8.22, "packing"]], "response_mode": "offline", "answer": "[{\"word\":\"packing\",\"start\":7.73,\"end\":8.22}]"} {"utterance_id": "1651-136854-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0026.flac", "task_type": "lexical_detection", "query": "perform", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"perform\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.54, "perform"]], "response_mode": "offline", "answer": "[{\"word\":\"perform\",\"start\":2.12,\"end\":2.54}]"} {"utterance_id": "1630-102884-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0008.flac", "task_type": "lexical_detection", "query": "buddha", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"buddha\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.84, "buddha"]], "response_mode": "offline", "answer": "[{\"word\":\"buddha\",\"start\":1.37,\"end\":1.84}]"} {"utterance_id": "4515-11057-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0015.flac", "task_type": "lexical_detection", "query": "utterly", "prompt": "When the utterance is complete, report all instances of the word \"utterly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.5, "utterly"]], "response_mode": "offline", "answer": "[{\"word\":\"utterly\",\"start\":2.98,\"end\":3.5}]"} {"utterance_id": "2506-11278-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.58, 6.3, "landlady"]], "response_mode": "offline", "answer": "[{\"word\":\"landlady\",\"start\":5.58,\"end\":6.3}]"} {"utterance_id": "3915-98647-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0033.flac", "task_type": "lexical_counting", "query": "sent", "prompt": "Listen to the entire clip, then report the count sequence for the word \"sent\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.65, "one"], [12.59, 13.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.38,\"end\":2.65},{\"word\":\"two\",\"start\":12.59,\"end\":13.02}]"} {"utterance_id": "116-288048-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0029.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.38, "preachers"]], "response_mode": "offline", "answer": "[{\"word\":\"preachers\",\"start\":0.98,\"end\":1.38}]"} {"utterance_id": "2506-11278-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0027.flac", "task_type": "lexical_detection", "query": "middle", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"middle\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.51, 2.79, "middle"]], "response_mode": "offline", "answer": "[{\"word\":\"middle\",\"start\":2.51,\"end\":2.79}]"} {"utterance_id": "116-288045-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.26, "heart"], [6.72, 7.18, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":1.8,\"end\":2.26},{\"word\":\"heart\",\"start\":6.72,\"end\":7.18}]"} {"utterance_id": "8254-84205-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0062.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 1.96, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":1.53,\"end\":1.96}]"} {"utterance_id": "3915-98647-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0036.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After the audio ends, extract all occurrences of a plant name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.93, 9.48, "flowers"]], "response_mode": "offline", "answer": "[{\"word\":\"flowers\",\"start\":8.93,\"end\":9.48}]"} {"utterance_id": "7641-96684-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0026.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.31, "one"], [4.26, 4.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.76,\"end\":1.31},{\"word\":\"two\",\"start\":4.26,\"end\":4.91}]"} {"utterance_id": "4153-185072-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0008.flac", "task_type": "lexical_detection", "query": "built", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"built\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.92, 3.14, "built"]], "response_mode": "offline", "answer": "[{\"word\":\"built\",\"start\":2.92,\"end\":3.14}]"} {"utterance_id": "1255-90413-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0019.flac", "task_type": "lexical_detection", "query": "marry", "prompt": "When the utterance is complete, report all instances of the word \"marry\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.86, "marry"]], "response_mode": "offline", "answer": "[{\"word\":\"marry\",\"start\":1.59,\"end\":1.86}]"} {"utterance_id": "116-288047-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0020.flac", "task_type": "lexical_detection", "query": "command", "prompt": "After the audio ends, extract all occurrences of the word \"command\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.29, 1.67, "command"]], "response_mode": "offline", "answer": "[{\"word\":\"command\",\"start\":1.29,\"end\":1.67}]"} {"utterance_id": "7601-175351-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0014.flac", "task_type": "lexical_detection", "query": "temperature", "prompt": "Process the full audio offline. Find every instance of the word \"temperature\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[19.54, 20.3, "temperature"]], "response_mode": "offline", "answer": "[{\"word\":\"temperature\",\"start\":19.54,\"end\":20.3}]"} {"utterance_id": "2506-11278-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.28, 0.6, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.28,\"end\":0.6}]"} {"utterance_id": "7697-245715-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0001.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.64, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":3.33,\"end\":3.64}]"} {"utterance_id": "3660-172183-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0023.flac", "task_type": "lexical_detection", "query": "washed", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"washed\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.58, 1.01, "washed"]], "response_mode": "offline", "answer": "[{\"word\":\"washed\",\"start\":0.58,\"end\":1.01}]"} {"utterance_id": "3663-172005-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0003.flac", "task_type": "lexical_detection", "query": "safe", "prompt": "When the utterance is complete, report all instances of the word \"safe\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.55, 5.01, "safe"]], "response_mode": "offline", "answer": "[{\"word\":\"safe\",\"start\":4.55,\"end\":5.01}]"} {"utterance_id": "4572-112375-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0014.flac", "task_type": "lexical_counting", "query": "thou", "prompt": "Listen to the entire clip, then report the count sequence for the word \"thou\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.31, 1.83, "one"], [8.27, 8.43, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.31,\"end\":1.83},{\"word\":\"two\",\"start\":8.27,\"end\":8.43}]"} {"utterance_id": "6267-65525-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.36, 2.81, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.36,\"end\":2.81}]"} {"utterance_id": "116-288047-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0014.flac", "task_type": "lexical_detection", "query": "create", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"create\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.42, "create"]], "response_mode": "offline", "answer": "[{\"word\":\"create\",\"start\":1.11,\"end\":1.42}]"} {"utterance_id": "2506-11278-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0025.flac", "task_type": "lexical_detection", "query": "creatures", "prompt": "Process the full audio offline. Find every instance of the word \"creatures\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.12, "creatures"]], "response_mode": "offline", "answer": "[{\"word\":\"creatures\",\"start\":0.59,\"end\":1.12}]"} {"utterance_id": "1585-131718-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0016.flac", "task_type": "lexical_detection", "query": "enemies", "prompt": "After the audio ends, extract all occurrences of the word \"enemies\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.76, 12.19, "enemies"]], "response_mode": "offline", "answer": "[{\"word\":\"enemies\",\"start\":11.76,\"end\":12.19}]"} {"utterance_id": "5543-27761-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0008.flac", "task_type": "lexical_detection", "query": "bride", "prompt": "Process the full audio offline. Find every instance of the word \"bride\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.88, "bride"]], "response_mode": "offline", "answer": "[{\"word\":\"bride\",\"start\":0.62,\"end\":0.88}]"} {"utterance_id": "1686-142278-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0060.flac", "task_type": "lexical_detection", "query": "shrink", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"shrink\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 2.84, "shrink"]], "response_mode": "offline", "answer": "[{\"word\":\"shrink\",\"start\":2.58,\"end\":2.84}]"} {"utterance_id": "4831-18525-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.87, 6.14, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":5.87,\"end\":6.14}]"} {"utterance_id": "700-122868-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.49, 0.79, "jane"]], "response_mode": "offline", "answer": "[{\"word\":\"jane\",\"start\":0.49,\"end\":0.79}]"} {"utterance_id": "4515-11057-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0063.flac", "task_type": "lexical_detection", "query": "accept", "prompt": "Once the audio has finished, identify every instance of the word \"accept\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.36, 4.9, "accept"]], "response_mode": "offline", "answer": "[{\"word\":\"accept\",\"start\":4.36,\"end\":4.9}]"} {"utterance_id": "4515-11057-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.23, 0.61, "amy"]], "response_mode": "offline", "answer": "[{\"word\":\"amy\",\"start\":0.23,\"end\":0.61}]"} {"utterance_id": "8288-274162-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0055.flac", "task_type": "lexical_detection", "query": "look", "prompt": "After the audio ends, extract all occurrences of the word \"look\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 0.76, "look"]], "response_mode": "offline", "answer": "[{\"word\":\"look\",\"start\":0.61,\"end\":0.76}]"} {"utterance_id": "116-288045-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0011.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.87, 1.27, "athens"]], "response_mode": "offline", "answer": "[{\"word\":\"athens\",\"start\":0.87,\"end\":1.27}]"} {"utterance_id": "6455-67803-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0022.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.18, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":1.82,\"end\":2.18}]"} {"utterance_id": "6467-62797-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0005.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.61, "milk"]], "response_mode": "offline", "answer": "[{\"word\":\"milk\",\"start\":1.28,\"end\":1.61}]"} {"utterance_id": "7641-96670-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0020.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.77,\"end\":3.03}]"} {"utterance_id": "7641-96252-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0022.flac", "task_type": "lexical_counting", "query": "weeps", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"weeps\" with word/start/end. If none, reply [].", "labels": [[1.43, 1.82, "one"], [3.64, 3.95, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.43,\"end\":1.82},{\"word\":\"two\",\"start\":3.64,\"end\":3.95}]"} {"utterance_id": "700-122866-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0033.flac", "task_type": "lexical_detection", "query": "montmorency", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"montmorency\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 3.27, "montmorency"]], "response_mode": "offline", "answer": "[{\"word\":\"montmorency\",\"start\":2.52,\"end\":3.27}]"} {"utterance_id": "4323-13259-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0023.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.34, 0.69, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":0.34,\"end\":0.69}]"} {"utterance_id": "1650-157641-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0011.flac", "task_type": "semantic_counting", "query": "country", "prompt": "Once the audio has finished, list each occurrence of a country name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.41, 0.97, "one"], [3.65, 4.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.41,\"end\":0.97},{\"word\":\"two\",\"start\":3.65,\"end\":4.02}]"} {"utterance_id": "6467-94831-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.24, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":1.97,\"end\":2.24}]"} {"utterance_id": "7697-105815-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0013.flac", "task_type": "lexical_detection", "query": "whence", "prompt": "When the utterance is complete, report all instances of the word \"whence\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.39, "whence"]], "response_mode": "offline", "answer": "[{\"word\":\"whence\",\"start\":2.09,\"end\":2.39}]"} {"utterance_id": "1651-136854-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0002.flac", "task_type": "lexical_detection", "query": "loss", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"loss\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 4.0, "loss"]], "response_mode": "offline", "answer": "[{\"word\":\"loss\",\"start\":3.6,\"end\":4.0}]"} {"utterance_id": "1585-131718-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.08, "manager"]], "response_mode": "offline", "answer": "[{\"word\":\"manager\",\"start\":0.69,\"end\":1.08}]"} {"utterance_id": "3660-172182-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0015.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.88, "one"], [3.13, 3.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.47,\"end\":1.88},{\"word\":\"two\",\"start\":3.13,\"end\":3.6}]"} {"utterance_id": "6455-67803-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0022.flac", "task_type": "lexical_detection", "query": "wondered", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wondered\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.13, "wondered"]], "response_mode": "offline", "answer": "[{\"word\":\"wondered\",\"start\":0.77,\"end\":1.13}]"} {"utterance_id": "6267-65525-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0033.flac", "task_type": "lexical_detection", "query": "found", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"found\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.66, 3.97, "found"]], "response_mode": "offline", "answer": "[{\"word\":\"found\",\"start\":3.66,\"end\":3.97}]"} {"utterance_id": "6267-53049-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0010.flac", "task_type": "lexical_detection", "query": "faded", "prompt": "When the utterance is complete, report all instances of the word \"faded\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.88, 4.39, "faded"]], "response_mode": "offline", "answer": "[{\"word\":\"faded\",\"start\":3.88,\"end\":4.39}]"} {"utterance_id": "6841-88291-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0000.flac", "task_type": "lexical_detection", "query": "reported", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"reported\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.37, 3.83, "reported"]], "response_mode": "offline", "answer": "[{\"word\":\"reported\",\"start\":3.37,\"end\":3.83}]"} {"utterance_id": "6467-62797-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0003.flac", "task_type": "lexical_detection", "query": "call", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"call\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.23, "call"]], "response_mode": "offline", "answer": "[{\"word\":\"call\",\"start\":0.72,\"end\":1.23}]"} {"utterance_id": "4153-186223-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0005.flac", "task_type": "lexical_detection", "query": "jane", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"jane\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.4, "jane"]], "response_mode": "offline", "answer": "[{\"word\":\"jane\",\"start\":1.19,\"end\":1.4}]"} {"utterance_id": "6123-59186-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0016.flac", "task_type": "lexical_detection", "query": "poverty", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"poverty\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.62, 3.25, "poverty"]], "response_mode": "offline", "answer": "[{\"word\":\"poverty\",\"start\":2.62,\"end\":3.25}]"} {"utterance_id": "8173-294714-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0038.flac", "task_type": "lexical_detection", "query": "relieve", "prompt": "After the audio ends, extract all occurrences of the word \"relieve\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.64, "relieve"]], "response_mode": "offline", "answer": "[{\"word\":\"relieve\",\"start\":2.27,\"end\":2.64}]"} {"utterance_id": "8254-115543-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0000.flac", "task_type": "lexical_detection", "query": "preparations", "prompt": "After listening to the whole recording, output every instance of the word \"preparations\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.84, 3.54, "preparations"]], "response_mode": "offline", "answer": "[{\"word\":\"preparations\",\"start\":2.84,\"end\":3.54}]"} {"utterance_id": "7641-96684-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.15, 5.75, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":5.15,\"end\":5.75}]"} {"utterance_id": "4153-61735-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[19.0, 19.3, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":19.0,\"end\":19.3}]"} {"utterance_id": "6467-94831-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.85, "missus green"]], "response_mode": "offline", "answer": "[{\"word\":\"missus green\",\"start\":0.26,\"end\":0.85}]"} {"utterance_id": "3915-57461-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0030.flac", "task_type": "lexical_detection", "query": "beak", "prompt": "Process the full audio offline. Find every instance of the word \"beak\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.65, 7.04, "beak"]], "response_mode": "offline", "answer": "[{\"word\":\"beak\",\"start\":6.65,\"end\":7.04}]"} {"utterance_id": "4153-186222-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.1, 3.66, "aunt jane's"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt jane's\",\"start\":3.1,\"end\":3.66}]"} {"utterance_id": "1650-157641-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.3, 0.85, "dwarf's"]], "response_mode": "offline", "answer": "[{\"word\":\"dwarf's\",\"start\":0.3,\"end\":0.85}]"} {"utterance_id": "700-122868-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0040.flac", "task_type": "lexical_detection", "query": "helped", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"helped\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.5, 5.66, "helped"]], "response_mode": "offline", "answer": "[{\"word\":\"helped\",\"start\":5.5,\"end\":5.66}]"} {"utterance_id": "4153-186222-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0014.flac", "task_type": "lexical_detection", "query": "kissed", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"kissed\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.94, "kissed"]], "response_mode": "offline", "answer": "[{\"word\":\"kissed\",\"start\":0.64,\"end\":0.94}]"} {"utterance_id": "6841-88291-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0033.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.74, "bull calf"]], "response_mode": "offline", "answer": "[{\"word\":\"bull calf\",\"start\":1.07,\"end\":1.74}]"} {"utterance_id": "6455-67803-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0003.flac", "task_type": "lexical_detection", "query": "explained", "prompt": "When the utterance is complete, report all instances of the word \"explained\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.52, "explained"]], "response_mode": "offline", "answer": "[{\"word\":\"explained\",\"start\":2.08,\"end\":2.52}]"} {"utterance_id": "1686-142278-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0029.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[3.67, 4.13, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":3.67,\"end\":4.13}]"} {"utterance_id": "6123-59150-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0020.flac", "task_type": "lexical_detection", "query": "fond", "prompt": "Once the audio has finished, identify every instance of the word \"fond\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.71, 1.05, "fond"]], "response_mode": "offline", "answer": "[{\"word\":\"fond\",\"start\":0.71,\"end\":1.05}]"} {"utterance_id": "1630-73710-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0013.flac", "task_type": "lexical_detection", "query": "cost", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"cost\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.75, 2.16, "cost"]], "response_mode": "offline", "answer": "[{\"word\":\"cost\",\"start\":1.75,\"end\":2.16}]"} {"utterance_id": "5849-50873-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0001.flac", "task_type": "lexical_counting", "query": "dangerous", "prompt": "After the audio ends, tally every instance of the word \"dangerous\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[15.58, 16.24, "one"], [20.65, 21.19, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":15.58,\"end\":16.24},{\"word\":\"two\",\"start\":20.65,\"end\":21.19}]"} {"utterance_id": "6467-94831-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 1.07, "sam rover"]], "response_mode": "offline", "answer": "[{\"word\":\"sam rover\",\"start\":0.21,\"end\":1.07}]"} {"utterance_id": "3915-98647-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 0.91, "animal"]], "response_mode": "offline", "answer": "[{\"word\":\"animal\",\"start\":0.59,\"end\":0.91}]"} {"utterance_id": "6123-59150-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.23, 3.39, "vicomte de voguee"]], "response_mode": "offline", "answer": "[{\"word\":\"vicomte de voguee\",\"start\":2.23,\"end\":3.39}]"} {"utterance_id": "1630-73710-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0019.flac", "task_type": "lexical_detection", "query": "post", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"post\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.11, 5.51, "post"]], "response_mode": "offline", "answer": "[{\"word\":\"post\",\"start\":5.11,\"end\":5.51}]"} {"utterance_id": "7697-105815-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0010.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Process the full audio offline. For every occurrence of a number, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.82, 3.06, "one"], [4.83, 5.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.82,\"end\":3.06},{\"word\":\"two\",\"start\":4.83,\"end\":5.03}]"} {"utterance_id": "2506-11267-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0009.flac", "task_type": "lexical_detection", "query": "houndsditch", "prompt": "Once the audio has finished, identify every instance of the word \"houndsditch\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.61, 2.28, "houndsditch"]], "response_mode": "offline", "answer": "[{\"word\":\"houndsditch\",\"start\":1.61,\"end\":2.28}]"} {"utterance_id": "6599-38590-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0009.flac", "task_type": "lexical_detection", "query": "actors", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"actors\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 1.87, "actors"]], "response_mode": "offline", "answer": "[{\"word\":\"actors\",\"start\":1.49,\"end\":1.87}]"} {"utterance_id": "8288-274162-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0050.flac", "task_type": "lexical_detection", "query": "affair", "prompt": "Process the full audio offline. Find every instance of the word \"affair\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.48, 3.92, "affair"]], "response_mode": "offline", "answer": "[{\"word\":\"affair\",\"start\":3.48,\"end\":3.92}]"} {"utterance_id": "2506-11267-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0005.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 1.44, "one"], [8.67, 9.24, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.53,\"end\":1.44},{\"word\":\"two\",\"start\":8.67,\"end\":9.24}]"} {"utterance_id": "4323-18416-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0010.flac", "task_type": "lexical_detection", "query": "highly", "prompt": "When the utterance is complete, report all instances of the word \"highly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.31, "highly"]], "response_mode": "offline", "answer": "[{\"word\":\"highly\",\"start\":2.0,\"end\":2.31}]"} {"utterance_id": "1686-142278-0096", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0096.flac", "task_type": "lexical_detection", "query": "hurt", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"hurt\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.56, 7.8, "hurt"]], "response_mode": "offline", "answer": "[{\"word\":\"hurt\",\"start\":7.56,\"end\":7.8}]"} {"utterance_id": "6267-53049-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 2.84, "missus fairweather's"]], "response_mode": "offline", "answer": "[{\"word\":\"missus fairweather's\",\"start\":1.34,\"end\":2.84}]"} {"utterance_id": "1701-141760-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0003.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Process the full audio offline. Find every instance of an organization name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.9, 6.44, "company"]], "response_mode": "offline", "answer": "[{\"word\":\"company\",\"start\":5.9,\"end\":6.44}]"} {"utterance_id": "3915-57461-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0017.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.01, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":0.81,\"end\":1.01}]"} {"utterance_id": "3660-172182-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0037.flac", "task_type": "lexical_detection", "query": "ambassadors", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"ambassadors\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.86, "ambassadors"]], "response_mode": "offline", "answer": "[{\"word\":\"ambassadors\",\"start\":1.27,\"end\":1.86}]"} {"utterance_id": "1701-141760-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0001.flac", "task_type": "lexical_detection", "query": "march", "prompt": "Once the audio has finished, identify every instance of the word \"march\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.72, 2.24, "march"]], "response_mode": "offline", "answer": "[{\"word\":\"march\",\"start\":1.72,\"end\":2.24}]"} {"utterance_id": "4831-18525-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0016.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[3.87, 4.24, "one"], [8.5, 8.81, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.87,\"end\":4.24},{\"word\":\"two\",\"start\":8.5,\"end\":8.81}]"} {"utterance_id": "1630-102884-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0004.flac", "task_type": "lexical_detection", "query": "tied", "prompt": "Once the audio has finished, identify every instance of the word \"tied\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.12, 7.57, "tied"]], "response_mode": "offline", "answer": "[{\"word\":\"tied\",\"start\":7.12,\"end\":7.57}]"} {"utterance_id": "116-288048-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0029.flac", "task_type": "lexical_detection", "query": "today", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"today\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.88, "today"]], "response_mode": "offline", "answer": "[{\"word\":\"today\",\"start\":1.48,\"end\":1.88}]"} {"utterance_id": "8254-115543-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.3, "porter"], [3.57, 4.06, "servants"]], "response_mode": "offline", "answer": "[{\"word\":\"porter\",\"start\":0.92,\"end\":1.3},{\"word\":\"servants\",\"start\":3.57,\"end\":4.06}]"} {"utterance_id": "700-122867-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.43, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":3.15,\"end\":3.43}]"} {"utterance_id": "700-122866-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 1.92, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":1.54,\"end\":1.92}]"} {"utterance_id": "8254-84205-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.69, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":3.39,\"end\":3.69}]"} {"utterance_id": "6841-88291-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0048.flac", "task_type": "lexical_detection", "query": "toward", "prompt": "After listening to the whole recording, output every instance of the word \"toward\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.38, 0.7, "toward"]], "response_mode": "offline", "answer": "[{\"word\":\"toward\",\"start\":0.38,\"end\":0.7}]"} {"utterance_id": "4323-18416-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.14, 1.51, "pickering"]], "response_mode": "offline", "answer": "[{\"word\":\"pickering\",\"start\":1.14,\"end\":1.51}]"} {"utterance_id": "6841-88291-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0031.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.37, 7.66, "teeth"]], "response_mode": "offline", "answer": "[{\"word\":\"teeth\",\"start\":7.37,\"end\":7.66}]"} {"utterance_id": "8173-294714-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0017.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a food name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.79, 2.03, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":1.79,\"end\":2.03}]"} {"utterance_id": "7641-96670-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0014.flac", "task_type": "lexical_detection", "query": "quiet", "prompt": "After listening to the whole recording, output every instance of the word \"quiet\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.36, "quiet"]], "response_mode": "offline", "answer": "[{\"word\":\"quiet\",\"start\":2.03,\"end\":2.36}]"} {"utterance_id": "4153-186222-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0011.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[0.53, 0.93, "mother"], [7.69, 8.44, "father's sister"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.53,\"end\":0.93},{\"word\":\"father's sister\",\"start\":7.69,\"end\":8.44}]"} {"utterance_id": "6841-88291-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0025.flac", "task_type": "lexical_detection", "query": "tail", "prompt": "After listening to the whole recording, output every instance of the word \"tail\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.61, "tail"]], "response_mode": "offline", "answer": "[{\"word\":\"tail\",\"start\":1.3,\"end\":1.61}]"} {"utterance_id": "4153-61735-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0001.flac", "task_type": "lexical_detection", "query": "earth", "prompt": "Process the full audio offline. Find every instance of the word \"earth\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.36, "earth"]], "response_mode": "offline", "answer": "[{\"word\":\"earth\",\"start\":2.02,\"end\":2.36}]"} {"utterance_id": "6841-88294-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0042.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.22, 1.54, "one"], [2.75, 3.24, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.22,\"end\":1.54},{\"word\":\"two\",\"start\":2.75,\"end\":3.24}]"} {"utterance_id": "6267-65525-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0034.flac", "task_type": "lexical_detection", "query": "road", "prompt": "Process the full audio offline. Find every instance of the word \"road\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.17, "road"]], "response_mode": "offline", "answer": "[{\"word\":\"road\",\"start\":0.8,\"end\":1.17}]"} {"utterance_id": "6267-53049-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0032.flac", "task_type": "lexical_detection", "query": "fairweather", "prompt": "When the utterance is complete, report all instances of the word \"fairweather\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.94, "fairweather"]], "response_mode": "offline", "answer": "[{\"word\":\"fairweather\",\"start\":1.17,\"end\":1.94}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "lexical_detection", "query": "garrison", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"garrison\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.54, 3.23, "garrison"]], "response_mode": "offline", "answer": "[{\"word\":\"garrison\",\"start\":2.54,\"end\":3.23}]"} {"utterance_id": "6267-65525-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0020.flac", "task_type": "lexical_detection", "query": "itself", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"itself\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.26, 1.74, "itself"]], "response_mode": "offline", "answer": "[{\"word\":\"itself\",\"start\":1.26,\"end\":1.74}]"} {"utterance_id": "6267-65525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.25, 0.81, "aunt jane"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt jane\",\"start\":0.25,\"end\":0.81}]"} {"utterance_id": "4323-55228-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.98, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.49,\"end\":0.98}]"} {"utterance_id": "700-122868-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0014.flac", "task_type": "lexical_detection", "query": "landing", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"landing\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.12, 1.67, "landing"]], "response_mode": "offline", "answer": "[{\"word\":\"landing\",\"start\":1.12,\"end\":1.67}]"} {"utterance_id": "8254-84205-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0041.flac", "task_type": "lexical_detection", "query": "redskins", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"redskins\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.49, 2.18, "redskins"]], "response_mode": "offline", "answer": "[{\"word\":\"redskins\",\"start\":1.49,\"end\":2.18}]"} {"utterance_id": "3663-172005-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0004.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.56, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.26,\"end\":0.56}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.08, 1.73, "prince andrew's"], [5.73, 6.38, "rostov"], [14.3, 14.76, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew's\",\"start\":1.08,\"end\":1.73},{\"word\":\"rostov\",\"start\":5.73,\"end\":6.38},{\"word\":\"rostov\",\"start\":14.3,\"end\":14.76}]"} {"utterance_id": "8173-294714-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0025.flac", "task_type": "lexical_detection", "query": "yielded", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"yielded\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.96, "yielded"]], "response_mode": "offline", "answer": "[{\"word\":\"yielded\",\"start\":0.45,\"end\":0.96}]"} {"utterance_id": "4323-13259-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0014.flac", "task_type": "lexical_detection", "query": "gray", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"gray\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.38, 0.58, "gray"]], "response_mode": "offline", "answer": "[{\"word\":\"gray\",\"start\":0.38,\"end\":0.58}]"} {"utterance_id": "700-122866-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0023.flac", "task_type": "lexical_detection", "query": "twelve", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"twelve\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.48, 2.93, "twelve"]], "response_mode": "offline", "answer": "[{\"word\":\"twelve\",\"start\":2.48,\"end\":2.93}]"} {"utterance_id": "116-288045-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0014.flac", "task_type": "lexical_detection", "query": "produce", "prompt": "Once the audio has finished, identify every instance of the word \"produce\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.51, 0.97, "produce"]], "response_mode": "offline", "answer": "[{\"word\":\"produce\",\"start\":0.51,\"end\":0.97}]"} {"utterance_id": "6599-38591-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0003.flac", "task_type": "lexical_detection", "query": "retreating", "prompt": "After the audio ends, extract all occurrences of the word \"retreating\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 2.21, "retreating"]], "response_mode": "offline", "answer": "[{\"word\":\"retreating\",\"start\":1.64,\"end\":2.21}]"} {"utterance_id": "4515-11057-0090", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0090.flac", "task_type": "lexical_detection", "query": "power", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"power\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.59, "power"]], "response_mode": "offline", "answer": "[{\"word\":\"power\",\"start\":1.21,\"end\":1.59}]"} {"utterance_id": "3660-172182-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0026.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After listening to the whole recording, output every instance of a country name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.01, 9.37, "britain"]], "response_mode": "offline", "answer": "[{\"word\":\"britain\",\"start\":9.01,\"end\":9.37}]"} {"utterance_id": "4153-186222-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0016.flac", "task_type": "lexical_detection", "query": "invalids", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"invalids\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 2.02, "invalids"]], "response_mode": "offline", "answer": "[{\"word\":\"invalids\",\"start\":1.43,\"end\":2.02}]"} {"utterance_id": "4831-25894-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0019.flac", "task_type": "lexical_detection", "query": "wistfully", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"wistfully\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.52, 5.03, "wistfully"]], "response_mode": "offline", "answer": "[{\"word\":\"wistfully\",\"start\":4.52,\"end\":5.03}]"} {"utterance_id": "6455-67803-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.51, "lady theobald"]], "response_mode": "offline", "answer": "[{\"word\":\"lady theobald\",\"start\":1.74,\"end\":2.51}]"} {"utterance_id": "4570-56594-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.56, "mouth"], [7.11, 7.59, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":1.06,\"end\":1.56},{\"word\":\"face\",\"start\":7.11,\"end\":7.59}]"} {"utterance_id": "1255-74899-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[13.49, 13.75, "lovel"]], "response_mode": "offline", "answer": "[{\"word\":\"lovel\",\"start\":13.49,\"end\":13.75}]"} {"utterance_id": "4323-55228-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0041.flac", "task_type": "lexical_detection", "query": "miles", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"miles\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 1.78, "miles"]], "response_mode": "offline", "answer": "[{\"word\":\"miles\",\"start\":1.49,\"end\":1.78}]"} {"utterance_id": "1585-131718-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0020.flac", "task_type": "lexical_detection", "query": "inertia", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"inertia\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.34, 2.9, "inertia"]], "response_mode": "offline", "answer": "[{\"word\":\"inertia\",\"start\":2.34,\"end\":2.9}]"} {"utterance_id": "7601-175351-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0019.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a time of day with fields word, start, end (seconds). If none, reply [].", "labels": [[15.81, 16.26, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":15.81,\"end\":16.26}]"} {"utterance_id": "1686-142278-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0079.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.52, 9.94, "oxford"]], "response_mode": "offline", "answer": "[{\"word\":\"oxford\",\"start\":9.52,\"end\":9.94}]"} {"utterance_id": "4515-11057-0087", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0087.flac", "task_type": "lexical_detection", "query": "couldn't", "prompt": "When the utterance is complete, report all instances of the word \"couldn't\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.98, "couldn't"]], "response_mode": "offline", "answer": "[{\"word\":\"couldn't\",\"start\":0.53,\"end\":0.98}]"} {"utterance_id": "3915-57461-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0001.flac", "task_type": "lexical_detection", "query": "boasting", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"boasting\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.37, 5.84, "boasting"]], "response_mode": "offline", "answer": "[{\"word\":\"boasting\",\"start\":5.37,\"end\":5.84}]"} {"utterance_id": "3660-6517-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0016.flac", "task_type": "lexical_detection", "query": "telegram", "prompt": "Process the full audio offline. Find every instance of the word \"telegram\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.8, "telegram"]], "response_mode": "offline", "answer": "[{\"word\":\"telegram\",\"start\":0.31,\"end\":0.8}]"} {"utterance_id": "1651-136854-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0029.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.37, "thirty two"]], "response_mode": "offline", "answer": "[{\"word\":\"thirty two\",\"start\":0.6,\"end\":1.37}]"} {"utterance_id": "6123-59186-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0023.flac", "task_type": "lexical_detection", "query": "mood", "prompt": "After the audio ends, extract all occurrences of the word \"mood\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.86, "mood"], [6.23, 6.62, "mood"]], "response_mode": "offline", "answer": "[{\"word\":\"mood\",\"start\":0.64,\"end\":0.86},{\"word\":\"mood\",\"start\":6.23,\"end\":6.62}]"} {"utterance_id": "2506-11278-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0011.flac", "task_type": "lexical_detection", "query": "proposed", "prompt": "After listening to the whole recording, output every instance of the word \"proposed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.69, 6.15, "proposed"]], "response_mode": "offline", "answer": "[{\"word\":\"proposed\",\"start\":5.69,\"end\":6.15}]"} {"utterance_id": "7697-245712-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0020.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[4.63, 4.85, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":4.63,\"end\":4.85}]"} {"utterance_id": "4153-186222-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.85, "father"], [2.33, 2.61, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":0.51,\"end\":0.85},{\"word\":\"mother\",\"start\":2.33,\"end\":2.61}]"} {"utterance_id": "5849-50873-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0041.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.18, 10.92, "rhode island"], [17.21, 17.72, "new york"]], "response_mode": "offline", "answer": "[{\"word\":\"rhode island\",\"start\":10.18,\"end\":10.92},{\"word\":\"new york\",\"start\":17.21,\"end\":17.72}]"} {"utterance_id": "1255-90413-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.77, 7.19, "jocelyn"]], "response_mode": "offline", "answer": "[{\"word\":\"jocelyn\",\"start\":6.77,\"end\":7.19}]"} {"utterance_id": "6267-53049-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.18, 0.81, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.18,\"end\":0.81}]"} {"utterance_id": "8288-274162-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0035.flac", "task_type": "lexical_counting", "query": "never", "prompt": "Once the audio has finished, list each occurrence of the word \"never\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.59, 0.91, "one"], [4.18, 4.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.59,\"end\":0.91},{\"word\":\"two\",\"start\":4.18,\"end\":4.44}]"} {"utterance_id": "6123-59186-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0021.flac", "task_type": "lexical_detection", "query": "triumphant", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"triumphant\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.68, "triumphant"]], "response_mode": "offline", "answer": "[{\"word\":\"triumphant\",\"start\":1.92,\"end\":2.68}]"} {"utterance_id": "1686-142278-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0064.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.5, 5.91, "aunt shaw"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt shaw\",\"start\":5.5,\"end\":5.91}]"} {"utterance_id": "7697-105817-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0008.flac", "task_type": "lexical_detection", "query": "try", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"try\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.09, 3.33, "try"]], "response_mode": "offline", "answer": "[{\"word\":\"try\",\"start\":3.09,\"end\":3.33}]"} {"utterance_id": "6123-59150-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0026.flac", "task_type": "lexical_detection", "query": "show", "prompt": "When the utterance is complete, report all instances of the word \"show\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.06, "show"]], "response_mode": "offline", "answer": "[{\"word\":\"show\",\"start\":1.82,\"end\":2.06}]"} {"utterance_id": "4570-14911-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0014.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 4.03, "dog"]], "response_mode": "offline", "answer": "[{\"word\":\"dog\",\"start\":3.52,\"end\":4.03}]"} {"utterance_id": "700-122866-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0011.flac", "task_type": "lexical_detection", "query": "gillis", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"gillis\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.11, "gillis"]], "response_mode": "offline", "answer": "[{\"word\":\"gillis\",\"start\":0.78,\"end\":1.11}]"} {"utterance_id": "5849-50962-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0017.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.21, 2.61, "eyes"], [6.94, 7.28, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":2.21,\"end\":2.61},{\"word\":\"heart\",\"start\":6.94,\"end\":7.28}]"} {"utterance_id": "6841-88294-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0008.flac", "task_type": "semantic_counting", "query": "food", "prompt": "Listen to the entire clip, then report the count sequence for a food name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.94, "one"], [4.05, 4.57, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.68,\"end\":0.94},{\"word\":\"two\",\"start\":4.05,\"end\":4.57}]"} {"utterance_id": "3663-172528-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0025.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Once the audio has finished, identify every instance of a disease name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[11.65, 12.17, "scab"]], "response_mode": "offline", "answer": "[{\"word\":\"scab\",\"start\":11.65,\"end\":12.17}]"} {"utterance_id": "7641-96684-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0029.flac", "task_type": "lexical_counting", "query": "underworld", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"underworld\" with word/start/end. If none, reply [].", "labels": [[1.32, 1.93, "one"], [6.9, 7.4, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.32,\"end\":1.93},{\"word\":\"two\",\"start\":6.9,\"end\":7.4}]"} {"utterance_id": "6467-56885-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0010.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After the audio ends, extract all occurrences of a plant name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.43, "sandal"]], "response_mode": "offline", "answer": "[{\"word\":\"sandal\",\"start\":5.09,\"end\":5.43}]"} {"utterance_id": "1650-167613-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0043.flac", "task_type": "lexical_detection", "query": "war", "prompt": "Once the audio has finished, identify every instance of the word \"war\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.25, 1.6, "war"], [3.4, 3.95, "war"]], "response_mode": "offline", "answer": "[{\"word\":\"war\",\"start\":1.25,\"end\":1.6},{\"word\":\"war\",\"start\":3.4,\"end\":3.95}]"} {"utterance_id": "7601-101622-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0005.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.66, 1.29, "beasts"]], "response_mode": "offline", "answer": "[{\"word\":\"beasts\",\"start\":0.66,\"end\":1.29}]"} {"utterance_id": "1686-142278-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0004.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.64, 0.93, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.64,\"end\":0.93}]"} {"utterance_id": "4515-11057-0096", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0096.flac", "task_type": "lexical_detection", "query": "truth", "prompt": "Process the full audio offline. Find every instance of the word \"truth\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.93, "truth"]], "response_mode": "offline", "answer": "[{\"word\":\"truth\",\"start\":1.44,\"end\":1.93}]"} {"utterance_id": "8173-294714-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0055.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Do not answer until the audio ends. Then list each instance of a money or currency term as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.39, 2.79, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":2.39,\"end\":2.79}]"} {"utterance_id": "5849-50962-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0004.flac", "task_type": "lexical_detection", "query": "directions", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"directions\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.34, "directions"]], "response_mode": "offline", "answer": "[{\"word\":\"directions\",\"start\":0.67,\"end\":1.34}]"} {"utterance_id": "6267-53049-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0022.flac", "task_type": "lexical_detection", "query": "stepped", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"stepped\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.41, 2.96, "stepped"]], "response_mode": "offline", "answer": "[{\"word\":\"stepped\",\"start\":2.41,\"end\":2.96}]"} {"utterance_id": "3660-6517-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0013.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Once the audio has finished, identify every instance of a plant name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.92, 8.42, "flowers"]], "response_mode": "offline", "answer": "[{\"word\":\"flowers\",\"start\":7.92,\"end\":8.42}]"} {"utterance_id": "4570-14911-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 1.11, "shoulders"]], "response_mode": "offline", "answer": "[{\"word\":\"shoulders\",\"start\":0.56,\"end\":1.11}]"} {"utterance_id": "700-122867-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0019.flac", "task_type": "lexical_detection", "query": "talking", "prompt": "After listening to the whole recording, output every instance of the word \"talking\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.32, "talking"]], "response_mode": "offline", "answer": "[{\"word\":\"talking\",\"start\":1.98,\"end\":2.32}]"} {"utterance_id": "1701-141759-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.88, 1.25, "god's"]], "response_mode": "offline", "answer": "[{\"word\":\"god's\",\"start\":0.88,\"end\":1.25}]"} {"utterance_id": "1650-167613-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0047.flac", "task_type": "lexical_detection", "query": "begun", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"begun\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 2.02, "begun"]], "response_mode": "offline", "answer": "[{\"word\":\"begun\",\"start\":1.63,\"end\":2.02}]"} {"utterance_id": "1651-136854-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0000.flac", "task_type": "lexical_detection", "query": "fifty", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"fifty\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.01, 1.42, "fifty"]], "response_mode": "offline", "answer": "[{\"word\":\"fifty\",\"start\":1.01,\"end\":1.42}]"} {"utterance_id": "7697-105817-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0010.flac", "task_type": "lexical_detection", "query": "fault", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"fault\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.17, 3.61, "fault"]], "response_mode": "offline", "answer": "[{\"word\":\"fault\",\"start\":3.17,\"end\":3.61}]"} {"utterance_id": "5849-50962-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0016.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[2.03, 2.62, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.03,\"end\":2.62}]"} {"utterance_id": "8288-274162-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0008.flac", "task_type": "lexical_detection", "query": "assume", "prompt": "Once the audio has finished, identify every instance of the word \"assume\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.25, 2.63, "assume"]], "response_mode": "offline", "answer": "[{\"word\":\"assume\",\"start\":2.25,\"end\":2.63}]"} {"utterance_id": "3915-98647-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0035.flac", "task_type": "lexical_detection", "query": "seems", "prompt": "After the audio ends, extract all occurrences of the word \"seems\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.77, "seems"]], "response_mode": "offline", "answer": "[{\"word\":\"seems\",\"start\":0.5,\"end\":0.77}]"} {"utterance_id": "7641-96252-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0001.flac", "task_type": "lexical_detection", "query": "distinguished", "prompt": "Process the full audio offline. Find every instance of the word \"distinguished\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.28, 6.92, "distinguished"]], "response_mode": "offline", "answer": "[{\"word\":\"distinguished\",\"start\":6.28,\"end\":6.92}]"} {"utterance_id": "5849-50963-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0003.flac", "task_type": "lexical_detection", "query": "crew", "prompt": "When the utterance is complete, report all instances of the word \"crew\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.94, 5.26, "crew"]], "response_mode": "offline", "answer": "[{\"word\":\"crew\",\"start\":4.94,\"end\":5.26}]"} {"utterance_id": "5849-50873-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0031.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Do not answer until the audio ends. Then list each instance of a state or province name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.94, 4.45, "connecticut"]], "response_mode": "offline", "answer": "[{\"word\":\"connecticut\",\"start\":3.94,\"end\":4.45}]"} {"utterance_id": "1585-157660-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.32, 3.77, "actors authors"]], "response_mode": "offline", "answer": "[{\"word\":\"actors authors\",\"start\":2.32,\"end\":3.77}]"} {"utterance_id": "6267-65525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0013.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Do not answer until the audio ends. Then list each instance of an organization name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.06, 9.03, "maple leaf"]], "response_mode": "offline", "answer": "[{\"word\":\"maple leaf\",\"start\":8.06,\"end\":9.03}]"} {"utterance_id": "6455-67804-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0039.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.93, 2.37, "wrist"]], "response_mode": "offline", "answer": "[{\"word\":\"wrist\",\"start\":1.93,\"end\":2.37}]"} {"utterance_id": "7641-96670-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0010.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the full audio. After it finishes, find every instance of a country name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.68, 3.19, "greece"]], "response_mode": "offline", "answer": "[{\"word\":\"greece\",\"start\":2.68,\"end\":3.19}]"} {"utterance_id": "6467-94831-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.02, 11.25, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":11.02,\"end\":11.25}]"} {"utterance_id": "6467-94831-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0034.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.73, 3.34, "snuggers"]], "response_mode": "offline", "answer": "[{\"word\":\"snuggers\",\"start\":2.73,\"end\":3.34}]"} {"utterance_id": "7601-101622-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0002.flac", "task_type": "semantic_detection", "query": "money", "prompt": "When the utterance is complete, report all instances of a money or currency term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.95, 14.45, "pay"]], "response_mode": "offline", "answer": "[{\"word\":\"pay\",\"start\":13.95,\"end\":14.45}]"} {"utterance_id": "5849-50873-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0005.flac", "task_type": "lexical_detection", "query": "bounded", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"bounded\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.02, "bounded"]], "response_mode": "offline", "answer": "[{\"word\":\"bounded\",\"start\":1.57,\"end\":2.02}]"} {"utterance_id": "3915-57461-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0015.flac", "task_type": "lexical_detection", "query": "undone", "prompt": "After listening to the whole recording, output every instance of the word \"undone\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.24, 1.95, "undone"]], "response_mode": "offline", "answer": "[{\"word\":\"undone\",\"start\":1.24,\"end\":1.95}]"} {"utterance_id": "7697-105815-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0050.flac", "task_type": "lexical_counting", "query": "till", "prompt": "After listening to the whole recording, output timed count words for the word \"till\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.29, 2.46, "one"], [8.93, 9.21, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.29,\"end\":2.46},{\"word\":\"two\",\"start\":8.93,\"end\":9.21}]"} {"utterance_id": "7601-175351-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0011.flac", "task_type": "lexical_detection", "query": "murmur", "prompt": "Once the audio has finished, identify every instance of the word \"murmur\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.99, "murmur"]], "response_mode": "offline", "answer": "[{\"word\":\"murmur\",\"start\":0.5,\"end\":0.99}]"} {"utterance_id": "1255-138279-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0012.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "When the utterance is complete, report all instances of an ordinal number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.72, 4.07, "fifth"]], "response_mode": "offline", "answer": "[{\"word\":\"fifth\",\"start\":3.72,\"end\":4.07}]"} {"utterance_id": "8254-115543-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0018.flac", "task_type": "lexical_detection", "query": "farmer", "prompt": "Process the full audio offline. Find every instance of the word \"farmer\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.66, 2.08, "farmer"]], "response_mode": "offline", "answer": "[{\"word\":\"farmer\",\"start\":1.66,\"end\":2.08}]"} {"utterance_id": "1686-142278-0092", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0092.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.29, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":1.84,\"end\":2.29}]"} {"utterance_id": "1650-173552-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0001.flac", "task_type": "lexical_detection", "query": "morning", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"morning\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.67, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":1.22,\"end\":1.67}]"} {"utterance_id": "4831-18525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0011.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 1.81, "one"], [5.21, 5.65, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.53,\"end\":1.81},{\"word\":\"two\",\"start\":5.21,\"end\":5.65}]"} {"utterance_id": "1585-131718-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0007.flac", "task_type": "lexical_detection", "query": "harder", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"harder\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 4.14, "harder"]], "response_mode": "offline", "answer": "[{\"word\":\"harder\",\"start\":3.6,\"end\":4.14}]"} {"utterance_id": "700-122868-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0009.flac", "task_type": "lexical_detection", "query": "shawl", "prompt": "Once the audio has finished, identify every instance of the word \"shawl\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.88, 1.11, "shawl"]], "response_mode": "offline", "answer": "[{\"word\":\"shawl\",\"start\":0.88,\"end\":1.11}]"} {"utterance_id": "6467-97061-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0005.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.31, 4.73, "champa"]], "response_mode": "offline", "answer": "[{\"word\":\"champa\",\"start\":4.31,\"end\":4.73}]"} {"utterance_id": "6455-67803-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0034.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the full audio. Count every occurrence of a body part. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.93, 2.43, "one"], [3.97, 4.41, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.93,\"end\":2.43},{\"word\":\"two\",\"start\":3.97,\"end\":4.41}]"} {"utterance_id": "116-288047-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.75, 1.07, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":0.75,\"end\":1.07}]"} {"utterance_id": "116-288046-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0011.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[9.08, 9.34, "bread"]], "response_mode": "offline", "answer": "[{\"word\":\"bread\",\"start\":9.08,\"end\":9.34}]"} {"utterance_id": "700-122866-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0034.flac", "task_type": "lexical_detection", "query": "effect", "prompt": "After listening to the whole recording, output every instance of the word \"effect\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.35, "effect"]], "response_mode": "offline", "answer": "[{\"word\":\"effect\",\"start\":1.82,\"end\":2.35}]"} {"utterance_id": "6267-65525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0002.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After the audio ends, extract all occurrences of a plant name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.63, 7.44, "sunflower"]], "response_mode": "offline", "answer": "[{\"word\":\"sunflower\",\"start\":6.63,\"end\":7.44}]"} {"utterance_id": "3660-6517-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0005.flac", "task_type": "lexical_detection", "query": "cook", "prompt": "Once the audio has finished, identify every instance of the word \"cook\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.94, 5.21, "cook"]], "response_mode": "offline", "answer": "[{\"word\":\"cook\",\"start\":4.94,\"end\":5.21}]"} {"utterance_id": "3663-172005-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0002.flac", "task_type": "lexical_detection", "query": "danger", "prompt": "When the utterance is complete, report all instances of the word \"danger\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.05, "danger"]], "response_mode": "offline", "answer": "[{\"word\":\"danger\",\"start\":0.59,\"end\":1.05}]"} {"utterance_id": "1650-167613-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0025.flac", "task_type": "lexical_detection", "query": "we've", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"we've\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.25, 0.56, "we've"]], "response_mode": "offline", "answer": "[{\"word\":\"we've\",\"start\":0.25,\"end\":0.56}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "lexical_counting", "query": "bull", "prompt": "After the audio ends, tally every instance of the word \"bull\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.83, 1.23, "one"], [10.14, 10.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.83,\"end\":1.23},{\"word\":\"two\",\"start\":10.14,\"end\":10.46}]"} {"utterance_id": "4831-18525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.68, "polly"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":2.31,\"end\":2.68}]"} {"utterance_id": "8288-274162-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.28, 0.86, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":0.28,\"end\":0.86}]"} {"utterance_id": "1650-173552-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0002.flac", "task_type": "lexical_detection", "query": "ready", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"ready\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.04, 2.34, "ready"]], "response_mode": "offline", "answer": "[{\"word\":\"ready\",\"start\":2.04,\"end\":2.34}]"} {"utterance_id": "5849-50873-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0029.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "After the audio ends, tally every instance of a profession. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.33, 0.85, "one"], [8.81, 9.52, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.33,\"end\":0.85},{\"word\":\"two\",\"start\":8.81,\"end\":9.52}]"} {"utterance_id": "7697-105815-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0028.flac", "task_type": "lexical_detection", "query": "ease", "prompt": "After listening to the whole recording, output every instance of the word \"ease\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.15, 7.53, "ease"]], "response_mode": "offline", "answer": "[{\"word\":\"ease\",\"start\":7.15,\"end\":7.53}]"} {"utterance_id": "6599-38591-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.72, 2.23, "garcilaso de la vega"], [4.95, 5.6, "petrarch"]], "response_mode": "offline", "answer": "[{\"word\":\"garcilaso de la vega\",\"start\":0.72,\"end\":2.23},{\"word\":\"petrarch\",\"start\":4.95,\"end\":5.6}]"} {"utterance_id": "6841-88291-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0044.flac", "task_type": "lexical_detection", "query": "calves", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"calves\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.14, 1.51, "calves"]], "response_mode": "offline", "answer": "[{\"word\":\"calves\",\"start\":1.14,\"end\":1.51}]"} {"utterance_id": "1686-142278-0088", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0088.flac", "task_type": "lexical_detection", "query": "recovered", "prompt": "Once the audio has finished, identify every instance of the word \"recovered\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.81, 1.26, "recovered"]], "response_mode": "offline", "answer": "[{\"word\":\"recovered\",\"start\":0.81,\"end\":1.26}]"} {"utterance_id": "7641-96670-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0013.flac", "task_type": "lexical_detection", "query": "filled", "prompt": "After listening to the whole recording, output every instance of the word \"filled\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.62, "filled"]], "response_mode": "offline", "answer": "[{\"word\":\"filled\",\"start\":1.22,\"end\":1.62}]"} {"utterance_id": "6455-66379-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0008.flac", "task_type": "lexical_detection", "query": "yellow", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"yellow\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.37, 2.57, "yellow"]], "response_mode": "offline", "answer": "[{\"word\":\"yellow\",\"start\":2.37,\"end\":2.57}]"} {"utterance_id": "700-122867-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0010.flac", "task_type": "lexical_detection", "query": "anne", "prompt": "After the audio ends, extract all occurrences of the word \"anne\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 2.65, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":2.34,\"end\":2.65}]"} {"utterance_id": "2506-11267-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0018.flac", "task_type": "lexical_detection", "query": "next", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"next\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.95, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":0.66,\"end\":0.95}]"} {"utterance_id": "7641-96670-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0002.flac", "task_type": "lexical_detection", "query": "taught", "prompt": "When the utterance is complete, report all instances of the word \"taught\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.45, "taught"]], "response_mode": "offline", "answer": "[{\"word\":\"taught\",\"start\":3.15,\"end\":3.45}]"} {"utterance_id": "4515-11057-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0079.flac", "task_type": "lexical_detection", "query": "parents", "prompt": "Process the full audio offline. Find every instance of the word \"parents\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.9, "parents"]], "response_mode": "offline", "answer": "[{\"word\":\"parents\",\"start\":4.23,\"end\":4.9}]"} {"utterance_id": "1630-102884-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0011.flac", "task_type": "lexical_detection", "query": "harmony", "prompt": "When the utterance is complete, report all instances of the word \"harmony\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.54, 7.14, "harmony"]], "response_mode": "offline", "answer": "[{\"word\":\"harmony\",\"start\":6.54,\"end\":7.14}]"} {"utterance_id": "6841-88294-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.16, 0.55, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.16,\"end\":0.55}]"} {"utterance_id": "1686-142278-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0065.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.44, 0.95, "seventy"]], "response_mode": "offline", "answer": "[{\"word\":\"seventy\",\"start\":0.44,\"end\":0.95}]"} {"utterance_id": "6599-38590-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0003.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.45, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":4.23,\"end\":4.45}]"} {"utterance_id": "1585-131718-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0046.flac", "task_type": "lexical_detection", "query": "york", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"york\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.67, 3.05, "york"]], "response_mode": "offline", "answer": "[{\"word\":\"york\",\"start\":2.67,\"end\":3.05}]"} {"utterance_id": "5849-50873-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.28, "mister prince"]], "response_mode": "offline", "answer": "[{\"word\":\"mister prince\",\"start\":0.5,\"end\":1.28}]"} {"utterance_id": "4515-11057-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0031.flac", "task_type": "lexical_detection", "query": "falling", "prompt": "When the utterance is complete, report all instances of the word \"falling\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 1.99, "falling"]], "response_mode": "offline", "answer": "[{\"word\":\"falling\",\"start\":1.52,\"end\":1.99}]"} {"utterance_id": "6841-88294-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0001.flac", "task_type": "lexical_detection", "query": "lookout", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"lookout\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.05, 4.42, "lookout"]], "response_mode": "offline", "answer": "[{\"word\":\"lookout\",\"start\":4.05,\"end\":4.42}]"} {"utterance_id": "7697-105817-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.86, 1.25, "judge"]], "response_mode": "offline", "answer": "[{\"word\":\"judge\",\"start\":0.86,\"end\":1.25}]"} {"utterance_id": "6455-67803-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0006.flac", "task_type": "lexical_detection", "query": "broke", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"broke\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.55, 0.76, "broke"]], "response_mode": "offline", "answer": "[{\"word\":\"broke\",\"start\":0.55,\"end\":0.76}]"} {"utterance_id": "5849-50873-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 2.96, "mister prince wadsworth"], [6.76, 7.61, "governor andros"], [9.01, 9.95, "mister prince"], [21.62, 22.37, "mister prince"]], "response_mode": "offline", "answer": "[{\"word\":\"mister prince wadsworth\",\"start\":1.08,\"end\":2.96},{\"word\":\"governor andros\",\"start\":6.76,\"end\":7.61},{\"word\":\"mister prince\",\"start\":9.01,\"end\":9.95},{\"word\":\"mister prince\",\"start\":21.62,\"end\":22.37}]"} {"utterance_id": "3663-172528-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0008.flac", "task_type": "lexical_detection", "query": "comfortable", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"comfortable\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.0, 4.64, "comfortable"]], "response_mode": "offline", "answer": "[{\"word\":\"comfortable\",\"start\":4.0,\"end\":4.64}]"} {"utterance_id": "7641-96670-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0008.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.1, "dove"], [3.53, 3.95, "hawk"], [5.21, 5.59, "hawk's"]], "response_mode": "offline", "answer": "[{\"word\":\"dove\",\"start\":0.81,\"end\":1.1},{\"word\":\"hawk\",\"start\":3.53,\"end\":3.95},{\"word\":\"hawk's\",\"start\":5.21,\"end\":5.59}]"} {"utterance_id": "1630-141772-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.57, 4.97, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":4.57,\"end\":4.97}]"} {"utterance_id": "3915-98647-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0013.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.5, 4.35, "provisions"]], "response_mode": "offline", "answer": "[{\"word\":\"provisions\",\"start\":3.5,\"end\":4.35}]"} {"utterance_id": "8254-115543-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0035.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[2.03, 2.43, "brahmin"], [12.44, 12.82, "servants"]], "response_mode": "offline", "answer": "[{\"word\":\"brahmin\",\"start\":2.03,\"end\":2.43},{\"word\":\"servants\",\"start\":12.44,\"end\":12.82}]"} {"utterance_id": "8254-115543-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0016.flac", "task_type": "lexical_detection", "query": "fearfully", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fearfully\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.71, 3.35, "fearfully"]], "response_mode": "offline", "answer": "[{\"word\":\"fearfully\",\"start\":2.71,\"end\":3.35}]"} {"utterance_id": "7697-105815-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0036.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.41, 4.83, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":4.41,\"end\":4.83}]"} {"utterance_id": "8173-294714-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.83, 9.42, "child"]], "response_mode": "offline", "answer": "[{\"word\":\"child\",\"start\":8.83,\"end\":9.42}]"} {"utterance_id": "4831-18525-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0020.flac", "task_type": "lexical_detection", "query": "got", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"got\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.2, 2.39, "got"]], "response_mode": "offline", "answer": "[{\"word\":\"got\",\"start\":2.2,\"end\":2.39}]"} {"utterance_id": "2506-11278-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0020.flac", "task_type": "lexical_detection", "query": "excuse", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"excuse\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.9, "excuse"]], "response_mode": "offline", "answer": "[{\"word\":\"excuse\",\"start\":0.39,\"end\":0.9}]"} {"utterance_id": "4323-55228-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.63, 4.0, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":3.63,\"end\":4.0}]"} {"utterance_id": "4515-11057-0101", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0101.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 0.77, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":0.35,\"end\":0.77}]"} {"utterance_id": "3663-172528-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0053.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 2.88, "one"], [7.04, 7.68, "two hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.61,\"end\":2.88},{\"word\":\"two hundred\",\"start\":7.04,\"end\":7.68}]"} {"utterance_id": "3660-172182-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0020.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.79, 8.31, "nephew"]], "response_mode": "offline", "answer": "[{\"word\":\"nephew\",\"start\":7.79,\"end\":8.31}]"} {"utterance_id": "5849-50873-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.8, 6.49, "king william's"], [8.82, 9.51, "andros"]], "response_mode": "offline", "answer": "[{\"word\":\"king william's\",\"start\":5.8,\"end\":6.49},{\"word\":\"andros\",\"start\":8.82,\"end\":9.51}]"} {"utterance_id": "7641-96684-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.62, 5.83, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":5.62,\"end\":5.83}]"} {"utterance_id": "8173-294714-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0043.flac", "task_type": "lexical_detection", "query": "revealed", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"revealed\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.72, 3.21, "revealed"]], "response_mode": "offline", "answer": "[{\"word\":\"revealed\",\"start\":2.72,\"end\":3.21}]"} {"utterance_id": "6467-97061-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0011.flac", "task_type": "lexical_detection", "query": "awaiting", "prompt": "After listening to the whole recording, output every instance of the word \"awaiting\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.48, "awaiting"]], "response_mode": "offline", "answer": "[{\"word\":\"awaiting\",\"start\":2.1,\"end\":2.48}]"} {"utterance_id": "1585-131718-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0028.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Do not answer until the audio ends. Then list each instance of a country name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.85, 9.32, "chinese"]], "response_mode": "offline", "answer": "[{\"word\":\"chinese\",\"start\":8.85,\"end\":9.32}]"} {"utterance_id": "8254-84205-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0054.flac", "task_type": "lexical_detection", "query": "stores", "prompt": "After the audio ends, extract all occurrences of the word \"stores\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 3.02, "stores"]], "response_mode": "offline", "answer": "[{\"word\":\"stores\",\"start\":2.58,\"end\":3.02}]"} {"utterance_id": "2506-13150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0000.flac", "task_type": "lexical_detection", "query": "overtaken", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"overtaken\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.42, 7.0, "overtaken"]], "response_mode": "offline", "answer": "[{\"word\":\"overtaken\",\"start\":6.42,\"end\":7.0}]"} {"utterance_id": "3663-172528-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.34, 1.89, "maestro agostino"]], "response_mode": "offline", "answer": "[{\"word\":\"maestro agostino\",\"start\":0.34,\"end\":1.89}]"} {"utterance_id": "7641-96684-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.25, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":0.6,\"end\":1.25}]"} {"utterance_id": "5849-50963-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.62, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":3.06,\"end\":3.62}]"} {"utterance_id": "6267-65525-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0040.flac", "task_type": "lexical_detection", "query": "can't", "prompt": "Once the audio has finished, identify every instance of the word \"can't\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.86, 3.24, "can't"]], "response_mode": "offline", "answer": "[{\"word\":\"can't\",\"start\":2.86,\"end\":3.24}]"} {"utterance_id": "1701-141759-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0023.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.5, 1.03, "one"], [2.19, 2.75, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":1.03},{\"word\":\"two\",\"start\":2.19,\"end\":2.75}]"} {"utterance_id": "116-288048-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0006.flac", "task_type": "lexical_detection", "query": "equinox", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"equinox\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 3.14, "equinox"], [4.88, 5.44, "equinox"]], "response_mode": "offline", "answer": "[{\"word\":\"equinox\",\"start\":2.52,\"end\":3.14},{\"word\":\"equinox\",\"start\":4.88,\"end\":5.44}]"} {"utterance_id": "3663-172528-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0019.flac", "task_type": "lexical_detection", "query": "precautions", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"precautions\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.61, 4.51, "precautions"]], "response_mode": "offline", "answer": "[{\"word\":\"precautions\",\"start\":3.61,\"end\":4.51}]"} {"utterance_id": "8288-274162-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.38, 3.83, "ears"]], "response_mode": "offline", "answer": "[{\"word\":\"ears\",\"start\":3.38,\"end\":3.83}]"} {"utterance_id": "6841-88294-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0063.flac", "task_type": "lexical_counting", "query": "money", "prompt": "Once the audio has finished, list each occurrence of the word \"money\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.45, 0.75, "one"], [2.07, 2.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.45,\"end\":0.75},{\"word\":\"two\",\"start\":2.07,\"end\":2.36}]"} {"utterance_id": "6841-88294-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0043.flac", "task_type": "lexical_detection", "query": "leaving", "prompt": "After listening to the whole recording, output every instance of the word \"leaving\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.51, 2.8, "leaving"]], "response_mode": "offline", "answer": "[{\"word\":\"leaving\",\"start\":2.51,\"end\":2.8}]"} {"utterance_id": "1630-96099-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0021.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.67, "lire"], [4.17, 4.99, "francs"], [7.83, 8.48, "francs"]], "response_mode": "offline", "answer": "[{\"word\":\"lire\",\"start\":2.07,\"end\":2.67},{\"word\":\"francs\",\"start\":4.17,\"end\":4.99},{\"word\":\"francs\",\"start\":7.83,\"end\":8.48}]"} {"utterance_id": "1651-136854-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.87, 6.41, "hearts"]], "response_mode": "offline", "answer": "[{\"word\":\"hearts\",\"start\":5.87,\"end\":6.41}]"} {"utterance_id": "1701-141760-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0030.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[5.84, 6.19, "one"], [9.04, 9.36, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.84,\"end\":6.19},{\"word\":\"two\",\"start\":9.04,\"end\":9.36}]"} {"utterance_id": "1686-142278-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0038.flac", "task_type": "lexical_detection", "query": "forgive", "prompt": "After listening to the whole recording, output every instance of the word \"forgive\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.18, "forgive"]], "response_mode": "offline", "answer": "[{\"word\":\"forgive\",\"start\":0.79,\"end\":1.18}]"} {"utterance_id": "8288-274162-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0041.flac", "task_type": "lexical_detection", "query": "met", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"met\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.42, "met"]], "response_mode": "offline", "answer": "[{\"word\":\"met\",\"start\":2.14,\"end\":2.42}]"} {"utterance_id": "1701-141760-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.97, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"rostov\",\"start\":0.47,\"end\":0.97}]"} {"utterance_id": "1686-142278-0078", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0078.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.78, 4.59, "manufacturers"]], "response_mode": "offline", "answer": "[{\"word\":\"manufacturers\",\"start\":3.78,\"end\":4.59}]"} {"utterance_id": "3663-172005-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.48, "fawn"]], "response_mode": "offline", "answer": "[{\"word\":\"fawn\",\"start\":1.1,\"end\":1.48}]"} {"utterance_id": "116-288045-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0019.flac", "task_type": "lexical_detection", "query": "apollo", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"apollo\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.09, 2.52, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":2.09,\"end\":2.52}]"} {"utterance_id": "3660-6517-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0029.flac", "task_type": "lexical_detection", "query": "appeared", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"appeared\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.27, "appeared"]], "response_mode": "offline", "answer": "[{\"word\":\"appeared\",\"start\":0.91,\"end\":1.27}]"} {"utterance_id": "4831-29134-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0003.flac", "task_type": "semantic_counting", "query": "date", "prompt": "Process the full audio offline. For every occurrence of a date, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.25, "one"], [6.79, 7.15, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.91,\"end\":1.25},{\"word\":\"two\",\"start\":6.79,\"end\":7.15}]"} {"utterance_id": "7641-96670-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.35, 4.84, "medea"]], "response_mode": "offline", "answer": "[{\"word\":\"medea\",\"start\":4.35,\"end\":4.84}]"} {"utterance_id": "4323-55228-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0004.flac", "task_type": "lexical_detection", "query": "hannah", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"hannah\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.91, "hannah"]], "response_mode": "offline", "answer": "[{\"word\":\"hannah\",\"start\":0.66,\"end\":0.91}]"} {"utterance_id": "1585-157660-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0010.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.88, "one"], [7.29, 7.75, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.49,\"end\":0.88},{\"word\":\"two\",\"start\":7.29,\"end\":7.75}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.95, 3.36, "artist"]], "response_mode": "offline", "answer": "[{\"word\":\"artist\",\"start\":2.95,\"end\":3.36}]"} {"utterance_id": "6267-65525-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.36, 4.94, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":4.36,\"end\":4.94}]"} {"utterance_id": "1701-141760-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0053.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.68, 4.33, "adjutant"]], "response_mode": "offline", "answer": "[{\"word\":\"adjutant\",\"start\":3.68,\"end\":4.33}]"} {"utterance_id": "3660-6517-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0019.flac", "task_type": "lexical_detection", "query": "stern", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"stern\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.77, 2.28, "stern"]], "response_mode": "offline", "answer": "[{\"word\":\"stern\",\"start\":1.77,\"end\":2.28}]"} {"utterance_id": "4323-55228-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0033.flac", "task_type": "lexical_detection", "query": "demanded", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"demanded\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.69, 3.11, "demanded"]], "response_mode": "offline", "answer": "[{\"word\":\"demanded\",\"start\":2.69,\"end\":3.11}]"} {"utterance_id": "5849-50873-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0042.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Process the full audio offline. Find every instance of a state or province name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.24, 4.91, "new england"]], "response_mode": "offline", "answer": "[{\"word\":\"new england\",\"start\":4.24,\"end\":4.91}]"} {"utterance_id": "1686-142278-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0048.flac", "task_type": "lexical_detection", "query": "pain", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"pain\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.2, 4.6, "pain"]], "response_mode": "offline", "answer": "[{\"word\":\"pain\",\"start\":4.2,\"end\":4.6}]"} {"utterance_id": "6455-66379-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0001.flac", "task_type": "lexical_detection", "query": "charge", "prompt": "Once the audio has finished, identify every instance of the word \"charge\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.42, 1.68, "charge"]], "response_mode": "offline", "answer": "[{\"word\":\"charge\",\"start\":1.42,\"end\":1.68}]"} {"utterance_id": "6267-53049-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0029.flac", "task_type": "lexical_detection", "query": "nobody", "prompt": "When the utterance is complete, report all instances of the word \"nobody\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.31, 3.87, "nobody"]], "response_mode": "offline", "answer": "[{\"word\":\"nobody\",\"start\":3.31,\"end\":3.87}]"} {"utterance_id": "4515-11057-0107", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0107.flac", "task_type": "lexical_detection", "query": "watch", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"watch\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.35, 1.68, "watch"]], "response_mode": "offline", "answer": "[{\"word\":\"watch\",\"start\":1.35,\"end\":1.68}]"} {"utterance_id": "4572-112381-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0006.flac", "task_type": "lexical_detection", "query": "aid", "prompt": "When the utterance is complete, report all instances of the word \"aid\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.6, 2.89, "aid"]], "response_mode": "offline", "answer": "[{\"word\":\"aid\",\"start\":2.6,\"end\":2.89}]"} {"utterance_id": "6841-88294-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0049.flac", "task_type": "lexical_detection", "query": "drunk", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"drunk\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.23, "drunk"]], "response_mode": "offline", "answer": "[{\"word\":\"drunk\",\"start\":1.88,\"end\":2.23}]"} {"utterance_id": "700-122866-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0011.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.11, "one"], [2.84, 3.05, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":1.11},{\"word\":\"two\",\"start\":2.84,\"end\":3.05}]"} {"utterance_id": "3660-172183-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.63, 3.91, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":3.63,\"end\":3.91}]"} {"utterance_id": "4831-18525-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0026.flac", "task_type": "lexical_detection", "query": "accomplish", "prompt": "Once the audio has finished, identify every instance of the word \"accomplish\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.91, 1.34, "accomplish"]], "response_mode": "offline", "answer": "[{\"word\":\"accomplish\",\"start\":0.91,\"end\":1.34}]"} {"utterance_id": "3660-172183-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.51, 3.98, "steeds"]], "response_mode": "offline", "answer": "[{\"word\":\"steeds\",\"start\":3.51,\"end\":3.98}]"} {"utterance_id": "6599-38590-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0008.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.99, 1.44, "cambrai"]], "response_mode": "offline", "answer": "[{\"word\":\"cambrai\",\"start\":0.99,\"end\":1.44}]"} {"utterance_id": "4831-29134-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.89, "children"]], "response_mode": "offline", "answer": "[{\"word\":\"children\",\"start\":1.46,\"end\":1.89}]"} {"utterance_id": "1686-142278-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0061.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.1, "mister hale"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale\",\"start\":0.5,\"end\":1.1}]"} {"utterance_id": "4153-186222-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.52, 1.15, "silas"]], "response_mode": "offline", "answer": "[{\"word\":\"silas\",\"start\":0.52,\"end\":1.15}]"} {"utterance_id": "8173-294714-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0008.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.04, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":2.77,\"end\":3.04}]"} {"utterance_id": "700-122866-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0014.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.08, 1.8, "one"], [3.09, 3.27, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.08,\"end\":1.8},{\"word\":\"two\",\"start\":3.09,\"end\":3.27}]"} {"utterance_id": "6123-59150-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.91, "novelist"]], "response_mode": "offline", "answer": "[{\"word\":\"novelist\",\"start\":1.48,\"end\":1.91}]"} {"utterance_id": "8173-294714-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0015.flac", "task_type": "lexical_detection", "query": "mine", "prompt": "Process the full audio offline. Find every instance of the word \"mine\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.87, "mine"]], "response_mode": "offline", "answer": "[{\"word\":\"mine\",\"start\":1.46,\"end\":1.87}]"} {"utterance_id": "3660-172183-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0026.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[1.92, 2.28, "one"], [4.0, 4.3, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.92,\"end\":2.28},{\"word\":\"two\",\"start\":4.0,\"end\":4.3}]"} {"utterance_id": "6267-53049-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.32, 2.21, "grandniece"]], "response_mode": "offline", "answer": "[{\"word\":\"grandniece\",\"start\":1.32,\"end\":2.21}]"} {"utterance_id": "8254-84205-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.71, 4.1, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":3.71,\"end\":4.1}]"} {"utterance_id": "6123-59186-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0025.flac", "task_type": "lexical_detection", "query": "years", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"years\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.65, 2.88, "years"]], "response_mode": "offline", "answer": "[{\"word\":\"years\",\"start\":2.65,\"end\":2.88}]"} {"utterance_id": "5849-50873-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0023.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the entire clip, then report the count sequence for a number as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.17, "one"], [5.78, 6.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.98,\"end\":3.17},{\"word\":\"two\",\"start\":5.78,\"end\":6.02}]"} {"utterance_id": "1686-142278-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0075.flac", "task_type": "lexical_detection", "query": "property", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"property\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.71, "property"]], "response_mode": "offline", "answer": "[{\"word\":\"property\",\"start\":1.26,\"end\":1.71}]"} {"utterance_id": "3663-172528-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0025.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen carefully to the entire clip, then report every instance of a city name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.09, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":1.77,\"end\":2.09}]"} {"utterance_id": "8288-274162-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0002.flac", "task_type": "lexical_detection", "query": "recalled", "prompt": "After the audio ends, extract all occurrences of the word \"recalled\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.26, 5.7, "recalled"]], "response_mode": "offline", "answer": "[{\"word\":\"recalled\",\"start\":5.26,\"end\":5.7}]"} {"utterance_id": "700-122867-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0015.flac", "task_type": "lexical_detection", "query": "wicked", "prompt": "After the audio ends, extract all occurrences of the word \"wicked\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.65, "wicked"]], "response_mode": "offline", "answer": "[{\"word\":\"wicked\",\"start\":2.12,\"end\":2.65}]"} {"utterance_id": "8173-294714-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.03, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":1.74,\"end\":2.03}]"} {"utterance_id": "6455-66379-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.76, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":0.4,\"end\":0.76}]"} {"utterance_id": "7641-96684-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0010.flac", "task_type": "lexical_counting", "query": "underworld", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"underworld\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.24, 2.87, "one"], [7.83, 8.33, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.24,\"end\":2.87},{\"word\":\"two\",\"start\":7.83,\"end\":8.33}]"} {"utterance_id": "1630-73710-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0003.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[4.03, 4.49, "one"], [10.48, 10.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.03,\"end\":4.49},{\"word\":\"two\",\"start\":10.48,\"end\":10.91}]"} {"utterance_id": "700-122868-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.46, 4.3, "jane andrews"]], "response_mode": "offline", "answer": "[{\"word\":\"jane andrews\",\"start\":3.46,\"end\":4.3}]"} {"utterance_id": "6455-66379-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0002.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a profession with word/start/end. If none, reply [].", "labels": [[5.62, 6.16, "one"], [8.23, 8.52, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.62,\"end\":6.16},{\"word\":\"two\",\"start\":8.23,\"end\":8.52}]"} {"utterance_id": "4572-64670-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0003.flac", "task_type": "lexical_detection", "query": "frame", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"frame\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.94, 2.44, "frame"]], "response_mode": "offline", "answer": "[{\"word\":\"frame\",\"start\":1.94,\"end\":2.44}]"} {"utterance_id": "6841-88291-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0040.flac", "task_type": "lexical_detection", "query": "fail", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"fail\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.36, 5.84, "fail"]], "response_mode": "offline", "answer": "[{\"word\":\"fail\",\"start\":5.36,\"end\":5.84}]"} {"utterance_id": "4515-11057-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0056.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.66, 4.11, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":3.66,\"end\":4.11}]"} {"utterance_id": "2506-11278-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0027.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.51, 2.79, "middle"]], "response_mode": "offline", "answer": "[{\"word\":\"middle\",\"start\":2.51,\"end\":2.79}]"} {"utterance_id": "4831-29134-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0017.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.54, "children"]], "response_mode": "offline", "answer": "[{\"word\":\"children\",\"start\":1.12,\"end\":1.54}]"} {"utterance_id": "5543-27761-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0022.flac", "task_type": "lexical_detection", "query": "bliss", "prompt": "After listening to the whole recording, output every instance of the word \"bliss\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.07, "bliss"]], "response_mode": "offline", "answer": "[{\"word\":\"bliss\",\"start\":1.68,\"end\":2.07}]"} {"utterance_id": "4515-11057-0095", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0095.flac", "task_type": "lexical_detection", "query": "despise", "prompt": "Once the audio has finished, identify every instance of the word \"despise\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.04, 2.71, "despise"]], "response_mode": "offline", "answer": "[{\"word\":\"despise\",\"start\":2.04,\"end\":2.71}]"} {"utterance_id": "1686-142278-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0044.flac", "task_type": "lexical_detection", "query": "trial", "prompt": "Process the full audio offline. Find every instance of the word \"trial\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.65, "trial"]], "response_mode": "offline", "answer": "[{\"word\":\"trial\",\"start\":1.19,\"end\":1.65}]"} {"utterance_id": "4831-25894-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.76, 2.09, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":1.76,\"end\":2.09}]"} {"utterance_id": "4153-61735-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.41, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.19,\"end\":1.41}]"} {"utterance_id": "1686-142278-0069", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0069.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 1.12, "mister hale"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale\",\"start\":0.49,\"end\":1.12}]"} {"utterance_id": "6123-59186-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0001.flac", "task_type": "lexical_detection", "query": "midnight", "prompt": "When the utterance is complete, report all instances of the word \"midnight\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.29, "midnight"]], "response_mode": "offline", "answer": "[{\"word\":\"midnight\",\"start\":4.82,\"end\":5.29}]"} {"utterance_id": "4515-11057-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0044.flac", "task_type": "lexical_detection", "query": "approached", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"approached\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.59, 7.1, "approached"]], "response_mode": "offline", "answer": "[{\"word\":\"approached\",\"start\":6.59,\"end\":7.1}]"} {"utterance_id": "7697-105815-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 1.09, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.36,\"end\":1.09}]"} {"utterance_id": "6841-88291-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0028.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "When the utterance is complete, report all instances of a state or province name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.75, "arizona"]], "response_mode": "offline", "answer": "[{\"word\":\"arizona\",\"start\":3.15,\"end\":3.75}]"} {"utterance_id": "4323-13259-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0005.flac", "task_type": "lexical_detection", "query": "president", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"president\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.74, "president"]], "response_mode": "offline", "answer": "[{\"word\":\"president\",\"start\":0.37,\"end\":0.74}]"} {"utterance_id": "4515-11057-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0020.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.91, "foremost"]], "response_mode": "offline", "answer": "[{\"word\":\"foremost\",\"start\":1.38,\"end\":1.91}]"} {"utterance_id": "6267-65525-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0035.flac", "task_type": "lexical_detection", "query": "visible", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"visible\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.9, 4.4, "visible"]], "response_mode": "offline", "answer": "[{\"word\":\"visible\",\"start\":3.9,\"end\":4.4}]"} {"utterance_id": "7697-105815-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0047.flac", "task_type": "lexical_detection", "query": "charity", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"charity\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.89, "charity"]], "response_mode": "offline", "answer": "[{\"word\":\"charity\",\"start\":0.44,\"end\":0.89}]"} {"utterance_id": "1630-141772-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0019.flac", "task_type": "lexical_counting", "query": "cannot", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"cannot\" with word/start/end. If none, reply [].", "labels": [[5.09, 5.43, "one"], [6.69, 7.11, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.09,\"end\":5.43},{\"word\":\"two\",\"start\":6.69,\"end\":7.11}]"} {"utterance_id": "3663-172528-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.6, 7.22, "doctors"]], "response_mode": "offline", "answer": "[{\"word\":\"doctors\",\"start\":6.6,\"end\":7.22}]"} {"utterance_id": "1585-131718-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0009.flac", "task_type": "lexical_detection", "query": "ultimate", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"ultimate\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[15.4, 15.89, "ultimate"]], "response_mode": "offline", "answer": "[{\"word\":\"ultimate\",\"start\":15.4,\"end\":15.89}]"} {"utterance_id": "4572-64670-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.65, 5.15, "heart"], [7.72, 8.13, "mind"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.65,\"end\":5.15},{\"word\":\"mind\",\"start\":7.72,\"end\":8.13}]"} {"utterance_id": "5849-50963-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0013.flac", "task_type": "lexical_detection", "query": "awkward", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"awkward\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.81, 8.26, "awkward"]], "response_mode": "offline", "answer": "[{\"word\":\"awkward\",\"start\":7.81,\"end\":8.26}]"} {"utterance_id": "3663-172528-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0021.flac", "task_type": "lexical_detection", "query": "commit", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"commit\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.87, 6.41, "commit"]], "response_mode": "offline", "answer": "[{\"word\":\"commit\",\"start\":5.87,\"end\":6.41}]"} {"utterance_id": "8254-84205-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0024.flac", "task_type": "lexical_detection", "query": "ned", "prompt": "When the utterance is complete, report all instances of the word \"ned\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.45, "ned"]], "response_mode": "offline", "answer": "[{\"word\":\"ned\",\"start\":2.26,\"end\":2.45}]"} {"utterance_id": "1701-141760-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0032.flac", "task_type": "lexical_detection", "query": "saying", "prompt": "Process the full audio offline. Find every instance of the word \"saying\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.0, 3.35, "saying"]], "response_mode": "offline", "answer": "[{\"word\":\"saying\",\"start\":3.0,\"end\":3.35}]"} {"utterance_id": "6467-94831-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0041.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.13, "donkey's"]], "response_mode": "offline", "answer": "[{\"word\":\"donkey's\",\"start\":0.59,\"end\":1.13}]"} {"utterance_id": "2506-11267-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.82, 5.37, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.82,\"end\":5.37}]"} {"utterance_id": "1255-90413-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0024.flac", "task_type": "lexical_detection", "query": "retreated", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"retreated\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.69, 3.16, "retreated"]], "response_mode": "offline", "answer": "[{\"word\":\"retreated\",\"start\":2.69,\"end\":3.16}]"} {"utterance_id": "1255-74899-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0001.flac", "task_type": "lexical_detection", "query": "countess", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"countess\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.36, "countess"]], "response_mode": "offline", "answer": "[{\"word\":\"countess\",\"start\":0.75,\"end\":1.36}]"} {"utterance_id": "116-288046-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0010.flac", "task_type": "lexical_detection", "query": "irreverent", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"irreverent\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.88, 3.65, "irreverent"]], "response_mode": "offline", "answer": "[{\"word\":\"irreverent\",\"start\":2.88,\"end\":3.65}]"} {"utterance_id": "4153-186223-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.0, "aunt jane's"], [2.4, 2.94, "aunt jane"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt jane's\",\"start\":0.4,\"end\":1.0},{\"word\":\"aunt jane\",\"start\":2.4,\"end\":2.94}]"} {"utterance_id": "6841-88291-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0019.flac", "task_type": "lexical_detection", "query": "knocked", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"knocked\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 1.93, "knocked"]], "response_mode": "offline", "answer": "[{\"word\":\"knocked\",\"start\":1.67,\"end\":1.93}]"} {"utterance_id": "4572-112381-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0001.flac", "task_type": "lexical_detection", "query": "fed", "prompt": "Process the full audio offline. Find every instance of the word \"fed\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.55, "fed"]], "response_mode": "offline", "answer": "[{\"word\":\"fed\",\"start\":3.22,\"end\":3.55}]"} {"utterance_id": "6467-94831-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0015.flac", "task_type": "lexical_detection", "query": "tom", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"tom\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.35, 2.68, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":2.35,\"end\":2.68}]"} {"utterance_id": "8254-84205-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0051.flac", "task_type": "lexical_detection", "query": "seem", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"seem\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.93, "seem"]], "response_mode": "offline", "answer": "[{\"word\":\"seem\",\"start\":0.66,\"end\":0.93}]"} {"utterance_id": "5543-27761-0095", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0095.flac", "task_type": "lexical_detection", "query": "dry", "prompt": "After listening to the whole recording, output every instance of the word \"dry\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.78, "dry"]], "response_mode": "offline", "answer": "[{\"word\":\"dry\",\"start\":2.31,\"end\":2.78}]"} {"utterance_id": "1701-141759-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0027.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.03, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.67,\"end\":1.03}]"} {"utterance_id": "1686-142278-0082", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0082.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.47, 14.94, "helstone"]], "response_mode": "offline", "answer": "[{\"word\":\"helstone\",\"start\":14.47,\"end\":14.94}]"} {"utterance_id": "5849-50873-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0038.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.63, "fruit"]], "response_mode": "offline", "answer": "[{\"word\":\"fruit\",\"start\":2.27,\"end\":2.63}]"} {"utterance_id": "6467-62797-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.65, "blood"], [3.63, 4.15, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"blood\",\"start\":0.32,\"end\":0.65},{\"word\":\"eyes\",\"start\":3.63,\"end\":4.15}]"} {"utterance_id": "1701-141759-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0019.flac", "task_type": "lexical_detection", "query": "officer", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"officer\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 2.06, "officer"]], "response_mode": "offline", "answer": "[{\"word\":\"officer\",\"start\":1.47,\"end\":2.06}]"} {"utterance_id": "1630-96099-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0020.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.59, 1.96, "ass"]], "response_mode": "offline", "answer": "[{\"word\":\"ass\",\"start\":1.59,\"end\":1.96}]"} {"utterance_id": "4831-29134-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0012.flac", "task_type": "lexical_detection", "query": "children", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"children\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.46, 1.89, "children"]], "response_mode": "offline", "answer": "[{\"word\":\"children\",\"start\":1.46,\"end\":1.89}]"} {"utterance_id": "7697-105815-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0031.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.92, 9.01, "foot"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":8.92,\"end\":9.01}]"} {"utterance_id": "700-122868-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0024.flac", "task_type": "lexical_detection", "query": "exclaimed", "prompt": "When the utterance is complete, report all instances of the word \"exclaimed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.94, 2.61, "exclaimed"]], "response_mode": "offline", "answer": "[{\"word\":\"exclaimed\",\"start\":1.94,\"end\":2.61}]"} {"utterance_id": "3663-172528-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0054.flac", "task_type": "lexical_detection", "query": "haste", "prompt": "After the audio ends, extract all occurrences of the word \"haste\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.48, 6.94, "haste"]], "response_mode": "offline", "answer": "[{\"word\":\"haste\",\"start\":6.48,\"end\":6.94}]"} {"utterance_id": "6123-59150-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0013.flac", "task_type": "lexical_counting", "query": "took", "prompt": "Listen to the entire clip, then report the count sequence for the word \"took\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.27, 0.48, "one"], [5.23, 5.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.27,\"end\":0.48},{\"word\":\"two\",\"start\":5.23,\"end\":5.49}]"} {"utterance_id": "4323-55228-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0046.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.93, "aunt hannah's"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt hannah's\",\"start\":0.46,\"end\":0.93}]"} {"utterance_id": "4153-186223-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0006.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.94, 6.3, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":5.94,\"end\":6.3}]"} {"utterance_id": "700-122866-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.28, "diana"]], "response_mode": "offline", "answer": "[{\"word\":\"diana\",\"start\":0.7,\"end\":1.28}]"} {"utterance_id": "2506-11267-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0001.flac", "task_type": "lexical_detection", "query": "beach", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"beach\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.05, 6.52, "beach"]], "response_mode": "offline", "answer": "[{\"word\":\"beach\",\"start\":6.05,\"end\":6.52}]"} {"utterance_id": "1686-142278-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0059.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.41, 3.92, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":3.41,\"end\":3.92}]"} {"utterance_id": "116-288047-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0017.flac", "task_type": "lexical_detection", "query": "religion", "prompt": "After listening to the whole recording, output every instance of the word \"religion\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.71, "religion"]], "response_mode": "offline", "answer": "[{\"word\":\"religion\",\"start\":0.33,\"end\":0.71}]"} {"utterance_id": "6467-94831-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0044.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.68, "peleg"]], "response_mode": "offline", "answer": "[{\"word\":\"peleg\",\"start\":0.31,\"end\":0.68}]"} {"utterance_id": "1630-73710-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0017.flac", "task_type": "lexical_detection", "query": "engagement", "prompt": "Once the audio has finished, identify every instance of the word \"engagement\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.49, 6.2, "engagement"]], "response_mode": "offline", "answer": "[{\"word\":\"engagement\",\"start\":5.49,\"end\":6.2}]"} {"utterance_id": "6455-66379-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0010.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.58, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.18,\"end\":1.58}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the entire clip, then report the count sequence for a number as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.5, "one"], [1.69, 1.88, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.16,\"end\":0.5},{\"word\":\"two\",\"start\":1.69,\"end\":1.88}]"} {"utterance_id": "3663-172528-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0018.flac", "task_type": "lexical_detection", "query": "departed", "prompt": "When the utterance is complete, report all instances of the word \"departed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.96, "departed"]], "response_mode": "offline", "answer": "[{\"word\":\"departed\",\"start\":3.39,\"end\":3.96}]"} {"utterance_id": "1686-142278-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.05, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.5,\"end\":1.05}]"} {"utterance_id": "4515-11057-0099", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0099.flac", "task_type": "lexical_detection", "query": "seems", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"seems\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.69, 1.01, "seems"]], "response_mode": "offline", "answer": "[{\"word\":\"seems\",\"start\":0.69,\"end\":1.01}]"} {"utterance_id": "1686-142278-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0079.flac", "task_type": "lexical_detection", "query": "fellows", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"fellows\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.86, 5.43, "fellows"]], "response_mode": "offline", "answer": "[{\"word\":\"fellows\",\"start\":4.86,\"end\":5.43}]"} {"utterance_id": "700-122866-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0009.flac", "task_type": "lexical_detection", "query": "scarcely", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"scarcely\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.72, 1.34, "scarcely"]], "response_mode": "offline", "answer": "[{\"word\":\"scarcely\",\"start\":0.72,\"end\":1.34}]"} {"utterance_id": "6267-65525-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0019.flac", "task_type": "lexical_detection", "query": "lincoln", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"lincoln\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.51, "lincoln"]], "response_mode": "offline", "answer": "[{\"word\":\"lincoln\",\"start\":2.0,\"end\":2.51}]"} {"utterance_id": "1701-141760-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0013.flac", "task_type": "lexical_detection", "query": "expect", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"expect\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.17, "expect"]], "response_mode": "offline", "answer": "[{\"word\":\"expect\",\"start\":0.75,\"end\":1.17}]"} {"utterance_id": "1701-141759-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.23, "natasha"]], "response_mode": "offline", "answer": "[{\"word\":\"natasha\",\"start\":2.7,\"end\":3.23}]"} {"utterance_id": "6467-94831-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 2.05, "teacher"]], "response_mode": "offline", "answer": "[{\"word\":\"teacher\",\"start\":1.63,\"end\":2.05}]"} {"utterance_id": "1651-136854-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0001.flac", "task_type": "lexical_detection", "query": "chose", "prompt": "After the audio ends, extract all occurrences of the word \"chose\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.01, 1.45, "chose"]], "response_mode": "offline", "answer": "[{\"word\":\"chose\",\"start\":1.01,\"end\":1.45}]"} {"utterance_id": "4570-56594-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0009.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After listening to the whole recording, output timed count words for a body part (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.06, 1.56, "one"], [7.11, 7.59, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.06,\"end\":1.56},{\"word\":\"two\",\"start\":7.11,\"end\":7.59}]"} {"utterance_id": "1255-74899-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0019.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a color name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.79, 1.19, "pink"]], "response_mode": "offline", "answer": "[{\"word\":\"pink\",\"start\":0.79,\"end\":1.19}]"} {"utterance_id": "2506-11267-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.62, 2.11, "carabas"]], "response_mode": "offline", "answer": "[{\"word\":\"carabas\",\"start\":1.62,\"end\":2.11}]"} {"utterance_id": "7601-175351-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0019.flac", "task_type": "lexical_detection", "query": "accustomed", "prompt": "After the audio ends, extract all occurrences of the word \"accustomed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.88, "accustomed"], [11.74, 12.36, "accustomed"]], "response_mode": "offline", "answer": "[{\"word\":\"accustomed\",\"start\":2.14,\"end\":2.88},{\"word\":\"accustomed\",\"start\":11.74,\"end\":12.36}]"} {"utterance_id": "3915-57461-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0030.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Once the audio has finished, list each occurrence of a body part as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[6.65, 7.04, "one"], [9.99, 10.45, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.65,\"end\":7.04},{\"word\":\"two\",\"start\":9.99,\"end\":10.45}]"} {"utterance_id": "7697-105815-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.98, 3.35, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":2.98,\"end\":3.35}]"} {"utterance_id": "6267-65525-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.2, 0.7, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":0.2,\"end\":0.7}]"} {"utterance_id": "1686-142278-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0050.flac", "task_type": "lexical_detection", "query": "perhaps", "prompt": "When the utterance is complete, report all instances of the word \"perhaps\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.95, "perhaps"]], "response_mode": "offline", "answer": "[{\"word\":\"perhaps\",\"start\":0.34,\"end\":0.95}]"} {"utterance_id": "6123-59186-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.92, 5.63, "mister hardy's"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy's\",\"start\":4.92,\"end\":5.63}]"} {"utterance_id": "4153-185072-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.91, 2.31, "allen"]], "response_mode": "offline", "answer": "[{\"word\":\"allen\",\"start\":1.91,\"end\":2.31}]"} {"utterance_id": "3660-172183-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0003.flac", "task_type": "lexical_counting", "query": "unto", "prompt": "After the audio ends, tally every instance of the word \"unto\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[2.13, 2.41, "one"], [4.81, 4.99, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.13,\"end\":2.41},{\"word\":\"two\",\"start\":4.81,\"end\":4.99}]"} {"utterance_id": "1650-173552-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0005.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "Listen to the audio in full, then emit a JSON array of count events for an organization name (word, start, end). If none, reply [].", "labels": [[1.48, 2.15, "one"], [3.98, 4.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.48,\"end\":2.15},{\"word\":\"two\",\"start\":3.98,\"end\":4.63}]"} {"utterance_id": "4323-18416-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0018.flac", "task_type": "lexical_detection", "query": "intend", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"intend\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.69, 5.99, "intend"]], "response_mode": "offline", "answer": "[{\"word\":\"intend\",\"start\":5.69,\"end\":5.99}]"} {"utterance_id": "1255-74899-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.31, 4.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.31,\"end\":4.63}]"} {"utterance_id": "6455-67803-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.56, 1.14, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":0.56,\"end\":1.14}]"} {"utterance_id": "4515-11057-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0006.flac", "task_type": "lexical_detection", "query": "wrote", "prompt": "Once the audio has finished, identify every instance of the word \"wrote\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.71, 0.99, "wrote"]], "response_mode": "offline", "answer": "[{\"word\":\"wrote\",\"start\":0.71,\"end\":0.99}]"} {"utterance_id": "6841-88294-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0039.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 1.75, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.5,\"end\":1.75}]"} {"utterance_id": "4570-102353-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0008.flac", "task_type": "lexical_detection", "query": "assured", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"assured\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.83, "assured"]], "response_mode": "offline", "answer": "[{\"word\":\"assured\",\"start\":0.46,\"end\":0.83}]"} {"utterance_id": "6455-66379-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0007.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "When the utterance is complete, report all instances of a weather term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.61, 7.93, "fog"]], "response_mode": "offline", "answer": "[{\"word\":\"fog\",\"start\":7.61,\"end\":7.93}]"} {"utterance_id": "4572-112375-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.38, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":2.76,\"end\":3.38}]"} {"utterance_id": "1701-141759-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.62, "sonya"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":1.08,\"end\":1.62}]"} {"utterance_id": "700-122866-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0000.flac", "task_type": "lexical_detection", "query": "avonlea", "prompt": "Once the audio has finished, identify every instance of the word \"avonlea\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.76, 1.39, "avonlea"]], "response_mode": "offline", "answer": "[{\"word\":\"avonlea\",\"start\":0.76,\"end\":1.39}]"} {"utterance_id": "6841-88291-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0043.flac", "task_type": "lexical_detection", "query": "dust", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"dust\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.74, "dust"]], "response_mode": "offline", "answer": "[{\"word\":\"dust\",\"start\":0.32,\"end\":0.74}]"} {"utterance_id": "1651-136854-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0030.flac", "task_type": "lexical_detection", "query": "latin", "prompt": "When the utterance is complete, report all instances of the word \"latin\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.92, 7.36, "latin"]], "response_mode": "offline", "answer": "[{\"word\":\"latin\",\"start\":6.92,\"end\":7.36}]"} {"utterance_id": "5543-27761-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0058.flac", "task_type": "lexical_detection", "query": "ordinary", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ordinary\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.58, 5.13, "ordinary"]], "response_mode": "offline", "answer": "[{\"word\":\"ordinary\",\"start\":4.58,\"end\":5.13}]"} {"utterance_id": "4515-11057-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0035.flac", "task_type": "lexical_detection", "query": "easily", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"easily\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.44, 1.93, "easily"]], "response_mode": "offline", "answer": "[{\"word\":\"easily\",\"start\":1.44,\"end\":1.93}]"} {"utterance_id": "1650-167613-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0038.flac", "task_type": "lexical_detection", "query": "rules", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"rules\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.48, 1.87, "rules"]], "response_mode": "offline", "answer": "[{\"word\":\"rules\",\"start\":1.48,\"end\":1.87}]"} {"utterance_id": "4323-18416-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0009.flac", "task_type": "lexical_detection", "query": "shutting", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"shutting\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.6, "shutting"]], "response_mode": "offline", "answer": "[{\"word\":\"shutting\",\"start\":2.35,\"end\":2.6}]"} {"utterance_id": "4570-24733-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0005.flac", "task_type": "lexical_counting", "query": "existed", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"existed\" with word/start/end. If none, reply [].", "labels": [[7.27, 8.06, "one"], [14.81, 15.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.27,\"end\":8.06},{\"word\":\"two\",\"start\":14.81,\"end\":15.49}]"} {"utterance_id": "116-288046-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0007.flac", "task_type": "lexical_detection", "query": "bruno", "prompt": "Process the full audio offline. Find every instance of the word \"bruno\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.84, "bruno"]], "response_mode": "offline", "answer": "[{\"word\":\"bruno\",\"start\":0.49,\"end\":0.84}]"} {"utterance_id": "1650-167613-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0020.flac", "task_type": "lexical_detection", "query": "relative", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"relative\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.57, 1.24, "relative"]], "response_mode": "offline", "answer": "[{\"word\":\"relative\",\"start\":0.57,\"end\":1.24}]"} {"utterance_id": "7697-105815-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0040.flac", "task_type": "lexical_detection", "query": "rest", "prompt": "Once the audio has finished, identify every instance of the word \"rest\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.28, 3.64, "rest"]], "response_mode": "offline", "answer": "[{\"word\":\"rest\",\"start\":3.28,\"end\":3.64}]"} {"utterance_id": "6455-67803-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.09, "lucia"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia\",\"start\":0.54,\"end\":1.09}]"} {"utterance_id": "700-122866-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0036.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.71, 2.35, "missus allan"]], "response_mode": "offline", "answer": "[{\"word\":\"missus allan\",\"start\":1.71,\"end\":2.35}]"} {"utterance_id": "7697-245712-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0000.flac", "task_type": "lexical_detection", "query": "state", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"state\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.71, 2.08, "state"]], "response_mode": "offline", "answer": "[{\"word\":\"state\",\"start\":1.71,\"end\":2.08}]"} {"utterance_id": "700-122866-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 1.78, "miss josephine barry"]], "response_mode": "offline", "answer": "[{\"word\":\"miss josephine barry\",\"start\":0.45,\"end\":1.78}]"} {"utterance_id": "6467-94831-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0022.flac", "task_type": "lexical_detection", "query": "brow", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"brow\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.33, "brow"]], "response_mode": "offline", "answer": "[{\"word\":\"brow\",\"start\":0.96,\"end\":1.33}]"} {"utterance_id": "3663-172528-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0010.flac", "task_type": "lexical_detection", "query": "francesco", "prompt": "When the utterance is complete, report all instances of the word \"francesco\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.67, "francesco"]], "response_mode": "offline", "answer": "[{\"word\":\"francesco\",\"start\":0.86,\"end\":1.67}]"} {"utterance_id": "6123-59150-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0042.flac", "task_type": "lexical_detection", "query": "described", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"described\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.72, 4.26, "described"]], "response_mode": "offline", "answer": "[{\"word\":\"described\",\"start\":3.72,\"end\":4.26}]"} {"utterance_id": "6467-56885-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0010.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.51, "face"], [13.42, 13.89, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":2.0,\"end\":2.51},{\"word\":\"eyes\",\"start\":13.42,\"end\":13.89}]"} {"utterance_id": "2506-11267-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0005.flac", "task_type": "lexical_detection", "query": "terror", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"terror\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.59, 6.05, "terror"]], "response_mode": "offline", "answer": "[{\"word\":\"terror\",\"start\":5.59,\"end\":6.05}]"} {"utterance_id": "6267-53049-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0026.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a country name with fields word, start, end (seconds). If none, reply [].", "labels": [[9.7, 10.24, "thibet"]], "response_mode": "offline", "answer": "[{\"word\":\"thibet\",\"start\":9.7,\"end\":10.24}]"} {"utterance_id": "8254-84205-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0060.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 0.88, "provisions"]], "response_mode": "offline", "answer": "[{\"word\":\"provisions\",\"start\":0.35,\"end\":0.88}]"} {"utterance_id": "3915-57461-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0024.flac", "task_type": "lexical_detection", "query": "height", "prompt": "Once the audio has finished, identify every instance of the word \"height\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.44, 2.81, "height"]], "response_mode": "offline", "answer": "[{\"word\":\"height\",\"start\":2.44,\"end\":2.81}]"} {"utterance_id": "1701-141760-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0036.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.85, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":1.44,\"end\":1.85}]"} {"utterance_id": "700-122867-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0032.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Once the audio has finished, identify every instance of a date. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.12, 3.49, "monday"]], "response_mode": "offline", "answer": "[{\"word\":\"monday\",\"start\":3.12,\"end\":3.49}]"} {"utterance_id": "6123-59150-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0036.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.87, "dostoevsky"], [3.02, 3.72, "caliban"], [6.16, 6.97, "setebos"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":1.04,\"end\":1.87},{\"word\":\"caliban\",\"start\":3.02,\"end\":3.72},{\"word\":\"setebos\",\"start\":6.16,\"end\":6.97}]"} {"utterance_id": "4572-112383-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0003.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a food name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.13, 2.78, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":2.13,\"end\":2.78}]"} {"utterance_id": "5849-50964-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0007.flac", "task_type": "lexical_detection", "query": "engine", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"engine\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.0, 3.3, "engine"]], "response_mode": "offline", "answer": "[{\"word\":\"engine\",\"start\":3.0,\"end\":3.3}]"} {"utterance_id": "3660-172182-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0011.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.01, 7.62, "dwarf"]], "response_mode": "offline", "answer": "[{\"word\":\"dwarf\",\"start\":7.01,\"end\":7.62}]"} {"utterance_id": "6455-67804-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0005.flac", "task_type": "lexical_counting", "query": "time", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"time\" (word, start, end). If none, reply [].", "labels": [[0.63, 0.86, "one"], [2.86, 3.04, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.63,\"end\":0.86},{\"word\":\"two\",\"start\":2.86,\"end\":3.04}]"} {"utterance_id": "3915-57461-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0010.flac", "task_type": "lexical_detection", "query": "avoid", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"avoid\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 2.95, "avoid"]], "response_mode": "offline", "answer": "[{\"word\":\"avoid\",\"start\":2.57,\"end\":2.95}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "lexical_counting", "query": "plague", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"plague\" (word, start, end). If none, reply [].", "labels": [[4.81, 5.32, "one"], [27.75, 28.21, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.81,\"end\":5.32},{\"word\":\"two\",\"start\":27.75,\"end\":28.21}]"} {"utterance_id": "6455-67804-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0004.flac", "task_type": "lexical_counting", "query": "find", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"find\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.44, 0.62, "one"], [4.36, 4.61, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.44,\"end\":0.62},{\"word\":\"two\",\"start\":4.36,\"end\":4.61}]"} {"utterance_id": "6841-88294-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0042.flac", "task_type": "lexical_detection", "query": "died", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"died\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.92, "died"]], "response_mode": "offline", "answer": "[{\"word\":\"died\",\"start\":0.63,\"end\":0.92}]"} {"utterance_id": "1255-90407-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0006.flac", "task_type": "lexical_detection", "query": "he's", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"he's\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.53, 0.89, "he's"]], "response_mode": "offline", "answer": "[{\"word\":\"he's\",\"start\":0.53,\"end\":0.89}]"} {"utterance_id": "6123-59150-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0043.flac", "task_type": "lexical_detection", "query": "sworn", "prompt": "Process the full audio offline. Find every instance of the word \"sworn\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.73, 4.18, "sworn"]], "response_mode": "offline", "answer": "[{\"word\":\"sworn\",\"start\":3.73,\"end\":4.18}]"} {"utterance_id": "2506-13150-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0009.flac", "task_type": "lexical_detection", "query": "pains", "prompt": "Once the audio has finished, identify every instance of the word \"pains\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.83, 4.65, "pains"]], "response_mode": "offline", "answer": "[{\"word\":\"pains\",\"start\":3.83,\"end\":4.65}]"} {"utterance_id": "4515-11057-0073", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0073.flac", "task_type": "lexical_detection", "query": "need", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"need\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.18, "need"]], "response_mode": "offline", "answer": "[{\"word\":\"need\",\"start\":0.85,\"end\":1.18}]"} {"utterance_id": "1585-131718-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0018.flac", "task_type": "lexical_detection", "query": "conspicuous", "prompt": "When the utterance is complete, report all instances of the word \"conspicuous\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.39, "conspicuous"]], "response_mode": "offline", "answer": "[{\"word\":\"conspicuous\",\"start\":0.71,\"end\":1.39}]"} {"utterance_id": "6123-59150-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0038.flac", "task_type": "lexical_counting", "query": "night", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"night\" with word/start/end. If none, reply [].", "labels": [[1.24, 1.46, "one"], [3.06, 3.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.24,\"end\":1.46},{\"word\":\"two\",\"start\":3.06,\"end\":3.53}]"} {"utterance_id": "1585-131718-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0011.flac", "task_type": "lexical_detection", "query": "understood", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"understood\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.71, "understood"]], "response_mode": "offline", "answer": "[{\"word\":\"understood\",\"start\":1.26,\"end\":1.71}]"} {"utterance_id": "4570-24733-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0000.flac", "task_type": "lexical_detection", "query": "infinite", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"infinite\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.44, 9.9, "infinite"], [14.13, 14.62, "infinite"]], "response_mode": "offline", "answer": "[{\"word\":\"infinite\",\"start\":9.44,\"end\":9.9},{\"word\":\"infinite\",\"start\":14.13,\"end\":14.62}]"} {"utterance_id": "5849-50873-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0004.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.07, 7.72, "hartford"]], "response_mode": "offline", "answer": "[{\"word\":\"hartford\",\"start\":7.07,\"end\":7.72}]"} {"utterance_id": "6599-38591-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0008.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.2, 4.48, "sister"]], "response_mode": "offline", "answer": "[{\"word\":\"sister\",\"start\":4.2,\"end\":4.48}]"} {"utterance_id": "6267-53049-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0016.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.4, "tea"]], "response_mode": "offline", "answer": "[{\"word\":\"tea\",\"start\":2.07,\"end\":2.4}]"} {"utterance_id": "1701-141760-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.84, 6.19, "hand"], [9.04, 9.36, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":5.84,\"end\":6.19},{\"word\":\"face\",\"start\":9.04,\"end\":9.36}]"} {"utterance_id": "6841-88294-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0040.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.17, "sailor's"]], "response_mode": "offline", "answer": "[{\"word\":\"sailor's\",\"start\":2.75,\"end\":3.17}]"} {"utterance_id": "1701-141759-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0001.flac", "task_type": "lexical_detection", "query": "handed", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"handed\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.7, 3.05, "handed"]], "response_mode": "offline", "answer": "[{\"word\":\"handed\",\"start\":2.7,\"end\":3.05}]"} {"utterance_id": "700-122868-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.83, "anne's"]], "response_mode": "offline", "answer": "[{\"word\":\"anne's\",\"start\":0.52,\"end\":0.83}]"} {"utterance_id": "6455-66379-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.37, 3.03, "yellow face's"]], "response_mode": "offline", "answer": "[{\"word\":\"yellow face's\",\"start\":2.37,\"end\":3.03}]"} {"utterance_id": "4323-55228-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0006.flac", "task_type": "lexical_detection", "query": "glance", "prompt": "When the utterance is complete, report all instances of the word \"glance\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 2.87, "glance"]], "response_mode": "offline", "answer": "[{\"word\":\"glance\",\"start\":2.58,\"end\":2.87}]"} {"utterance_id": "1255-90407-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0030.flac", "task_type": "lexical_detection", "query": "how", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"how\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 0.85, "how"]], "response_mode": "offline", "answer": "[{\"word\":\"how\",\"start\":0.7,\"end\":0.85}]"} {"utterance_id": "4570-56594-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.19, 8.86, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":8.19,\"end\":8.86}]"} {"utterance_id": "4153-61735-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0010.flac", "task_type": "lexical_detection", "query": "compromised", "prompt": "After the audio ends, extract all occurrences of the word \"compromised\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.34, 8.1, "compromised"]], "response_mode": "offline", "answer": "[{\"word\":\"compromised\",\"start\":7.34,\"end\":8.1}]"} {"utterance_id": "1650-173552-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0006.flac", "task_type": "lexical_detection", "query": "false", "prompt": "When the utterance is complete, report all instances of the word \"false\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.72, 8.17, "false"]], "response_mode": "offline", "answer": "[{\"word\":\"false\",\"start\":7.72,\"end\":8.17}]"} {"utterance_id": "700-122868-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0009.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After the audio ends, extract all occurrences of a color name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 0.88, "black"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":0.61,\"end\":0.88}]"} {"utterance_id": "3660-172183-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0001.flac", "task_type": "lexical_detection", "query": "shout", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"shout\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.85, 6.1, "shout"]], "response_mode": "offline", "answer": "[{\"word\":\"shout\",\"start\":5.85,\"end\":6.1}]"} {"utterance_id": "1701-141760-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0035.flac", "task_type": "lexical_detection", "query": "mentioned", "prompt": "Once the audio has finished, identify every instance of the word \"mentioned\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.54, 1.92, "mentioned"]], "response_mode": "offline", "answer": "[{\"word\":\"mentioned\",\"start\":1.54,\"end\":1.92}]"} {"utterance_id": "4570-14911-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0000.flac", "task_type": "lexical_detection", "query": "harrier", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"harrier\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.66, 3.12, "harrier"]], "response_mode": "offline", "answer": "[{\"word\":\"harrier\",\"start\":2.66,\"end\":3.12}]"} {"utterance_id": "1650-157641-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0008.flac", "task_type": "lexical_detection", "query": "comment", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"comment\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.11, "comment"]], "response_mode": "offline", "answer": "[{\"word\":\"comment\",\"start\":2.69,\"end\":3.11}]"} {"utterance_id": "2506-169427-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0003.flac", "task_type": "lexical_counting", "query": "often", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"often\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[3.1, 3.41, "one"], [8.63, 8.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.1,\"end\":3.41},{\"word\":\"two\",\"start\":8.63,\"end\":8.91}]"} {"utterance_id": "6841-88291-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.64, 2.91, "rib"]], "response_mode": "offline", "answer": "[{\"word\":\"rib\",\"start\":2.64,\"end\":2.91}]"} {"utterance_id": "4323-13259-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0010.flac", "task_type": "lexical_detection", "query": "glowed", "prompt": "When the utterance is complete, report all instances of the word \"glowed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.76, 1.99, "glowed"]], "response_mode": "offline", "answer": "[{\"word\":\"glowed\",\"start\":1.76,\"end\":1.99}]"} {"utterance_id": "4153-61735-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0009.flac", "task_type": "lexical_detection", "query": "within", "prompt": "Once the audio has finished, identify every instance of the word \"within\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.69, 2.01, "within"]], "response_mode": "offline", "answer": "[{\"word\":\"within\",\"start\":1.69,\"end\":2.01}]"} {"utterance_id": "2506-13150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.43, 4.86, "clerk"]], "response_mode": "offline", "answer": "[{\"word\":\"clerk\",\"start\":4.43,\"end\":4.86}]"} {"utterance_id": "700-122868-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0023.flac", "task_type": "lexical_detection", "query": "shirley", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"shirley\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.79, 1.27, "shirley"]], "response_mode": "offline", "answer": "[{\"word\":\"shirley\",\"start\":0.79,\"end\":1.27}]"} {"utterance_id": "4831-18525-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[3.87, 4.24, "polly"], [8.5, 8.81, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":3.87,\"end\":4.24},{\"word\":\"tom\",\"start\":8.5,\"end\":8.81}]"} {"utterance_id": "8254-84205-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0037.flac", "task_type": "lexical_detection", "query": "ponies", "prompt": "After listening to the whole recording, output every instance of the word \"ponies\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 2.25, "ponies"]], "response_mode": "offline", "answer": "[{\"word\":\"ponies\",\"start\":1.81,\"end\":2.25}]"} {"utterance_id": "6599-38590-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0006.flac", "task_type": "lexical_detection", "query": "hast", "prompt": "After listening to the whole recording, output every instance of the word \"hast\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.07, "hast"]], "response_mode": "offline", "answer": "[{\"word\":\"hast\",\"start\":1.84,\"end\":2.07}]"} {"utterance_id": "4153-186223-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.38, 0.74, "kenneth"]], "response_mode": "offline", "answer": "[{\"word\":\"kenneth\",\"start\":0.38,\"end\":0.74}]"} {"utterance_id": "7641-96684-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.0, 2.34, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":2.0,\"end\":2.34}]"} {"utterance_id": "4515-11057-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0023.flac", "task_type": "lexical_detection", "query": "work", "prompt": "After listening to the whole recording, output every instance of the word \"work\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.07, 6.45, "work"]], "response_mode": "offline", "answer": "[{\"word\":\"work\",\"start\":6.07,\"end\":6.45}]"} {"utterance_id": "2506-11278-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0009.flac", "task_type": "lexical_detection", "query": "honourable", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"honourable\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.8, 3.3, "honourable"]], "response_mode": "offline", "answer": "[{\"word\":\"honourable\",\"start\":2.8,\"end\":3.3}]"} {"utterance_id": "3660-6517-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0033.flac", "task_type": "lexical_detection", "query": "rail", "prompt": "When the utterance is complete, report all instances of the word \"rail\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.4, "rail"]], "response_mode": "offline", "answer": "[{\"word\":\"rail\",\"start\":3.15,\"end\":3.4}]"} {"utterance_id": "6455-67803-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.24, "hand"], [3.73, 4.18, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":1.82,\"end\":2.24},{\"word\":\"face\",\"start\":3.73,\"end\":4.18}]"} {"utterance_id": "4831-25894-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0028.flac", "task_type": "lexical_detection", "query": "thank", "prompt": "After listening to the whole recording, output every instance of the word \"thank\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.44, "thank"]], "response_mode": "offline", "answer": "[{\"word\":\"thank\",\"start\":3.1,\"end\":3.44}]"} {"utterance_id": "700-122867-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 1.26, "anne shirley"]], "response_mode": "offline", "answer": "[{\"word\":\"anne shirley\",\"start\":0.33,\"end\":1.26}]"} {"utterance_id": "116-288047-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0004.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Do not answer until the audio ends. Then list each instance of a country name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.07, 2.45, "india"]], "response_mode": "offline", "answer": "[{\"word\":\"india\",\"start\":2.07,\"end\":2.45}]"} {"utterance_id": "4515-11057-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0064.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.59, 2.1, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":1.59,\"end\":2.1}]"} {"utterance_id": "4153-186222-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0013.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.18, 3.55, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":3.18,\"end\":3.55}]"} {"utterance_id": "116-288046-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0011.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.35, 4.71, "bee"]], "response_mode": "offline", "answer": "[{\"word\":\"bee\",\"start\":4.35,\"end\":4.71}]"} {"utterance_id": "1701-141759-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0004.flac", "task_type": "lexical_detection", "query": "careful", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"careful\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.73, 2.29, "careful"]], "response_mode": "offline", "answer": "[{\"word\":\"careful\",\"start\":1.73,\"end\":2.29}]"} {"utterance_id": "3663-172528-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0053.flac", "task_type": "lexical_detection", "query": "shoot", "prompt": "When the utterance is complete, report all instances of the word \"shoot\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.31, 1.75, "shoot"]], "response_mode": "offline", "answer": "[{\"word\":\"shoot\",\"start\":1.31,\"end\":1.75}]"} {"utterance_id": "8288-274162-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0009.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.99, "chaffinch"]], "response_mode": "offline", "answer": "[{\"word\":\"chaffinch\",\"start\":3.22,\"end\":3.99}]"} {"utterance_id": "6123-59150-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0040.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.12, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":1.57,\"end\":2.12}]"} {"utterance_id": "700-122867-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.76, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":0.51,\"end\":0.76}]"} {"utterance_id": "4572-64670-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0006.flac", "task_type": "lexical_detection", "query": "mourned", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"mourned\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.53, 10.03, "mourned"]], "response_mode": "offline", "answer": "[{\"word\":\"mourned\",\"start\":9.53,\"end\":10.03}]"} {"utterance_id": "4153-185072-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0009.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.87, 2.55, "refreshment"]], "response_mode": "offline", "answer": "[{\"word\":\"refreshment\",\"start\":1.87,\"end\":2.55}]"} {"utterance_id": "6841-88294-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0006.flac", "task_type": "lexical_detection", "query": "mostly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"mostly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.97, "mostly"]], "response_mode": "offline", "answer": "[{\"word\":\"mostly\",\"start\":1.59,\"end\":1.97}]"} {"utterance_id": "6467-94831-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.41, "aleck"]], "response_mode": "offline", "answer": "[{\"word\":\"aleck\",\"start\":0.98,\"end\":1.41}]"} {"utterance_id": "7697-105815-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0024.flac", "task_type": "lexical_detection", "query": "limbs", "prompt": "After listening to the whole recording, output every instance of the word \"limbs\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.43, 12.04, "limbs"]], "response_mode": "offline", "answer": "[{\"word\":\"limbs\",\"start\":11.43,\"end\":12.04}]"} {"utterance_id": "116-288046-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0010.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.71, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.37,\"end\":0.71}]"} {"utterance_id": "1650-157641-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0015.flac", "task_type": "lexical_detection", "query": "consort", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"consort\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.56, 3.12, "consort"]], "response_mode": "offline", "answer": "[{\"word\":\"consort\",\"start\":2.56,\"end\":3.12}]"} {"utterance_id": "2506-169427-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0000.flac", "task_type": "lexical_detection", "query": "pop", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pop\" with fields word, start, end (seconds). If none, reply [].", "labels": [[13.69, 14.07, "pop"]], "response_mode": "offline", "answer": "[{\"word\":\"pop\",\"start\":13.69,\"end\":14.07}]"} {"utterance_id": "1585-131718-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0054.flac", "task_type": "lexical_detection", "query": "active", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"active\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.61, 1.07, "active"]], "response_mode": "offline", "answer": "[{\"word\":\"active\",\"start\":0.61,\"end\":1.07}]"} {"utterance_id": "6455-67804-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0012.flac", "task_type": "lexical_detection", "query": "subject", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"subject\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 1.93, "subject"]], "response_mode": "offline", "answer": "[{\"word\":\"subject\",\"start\":1.57,\"end\":1.93}]"} {"utterance_id": "8173-294714-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0044.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.75, "eye"]], "response_mode": "offline", "answer": "[{\"word\":\"eye\",\"start\":0.54,\"end\":0.75}]"} {"utterance_id": "6467-56885-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.21, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.82,\"end\":4.21}]"} {"utterance_id": "6841-88291-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0036.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.57, 1.94, "foot"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":1.57,\"end\":1.94}]"} {"utterance_id": "6467-94831-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 2.81, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":2.46,\"end\":2.81}]"} {"utterance_id": "8254-115543-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0019.flac", "task_type": "lexical_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"animal\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 2.21, "animal"]], "response_mode": "offline", "answer": "[{\"word\":\"animal\",\"start\":1.81,\"end\":2.21}]"} {"utterance_id": "6455-66379-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0015.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.27, 2.64, "six"]], "response_mode": "offline", "answer": "[{\"word\":\"six\",\"start\":2.27,\"end\":2.64}]"} {"utterance_id": "1686-142278-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0037.flac", "task_type": "lexical_detection", "query": "quietly", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"quietly\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.58, 8.03, "quietly"]], "response_mode": "offline", "answer": "[{\"word\":\"quietly\",\"start\":7.58,\"end\":8.03}]"} {"utterance_id": "1686-142278-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0062.flac", "task_type": "lexical_detection", "query": "burst", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"burst\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.27, 2.59, "burst"]], "response_mode": "offline", "answer": "[{\"word\":\"burst\",\"start\":2.27,\"end\":2.59}]"} {"utterance_id": "8288-274162-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.28, 0.9, "d'artagnan"]], "response_mode": "offline", "answer": "[{\"word\":\"d'artagnan\",\"start\":0.28,\"end\":0.9}]"} {"utterance_id": "4572-64670-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0022.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.04, 3.35, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":3.04,\"end\":3.35}]"} {"utterance_id": "1255-74899-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0003.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.66, 1.97, "wine"]], "response_mode": "offline", "answer": "[{\"word\":\"wine\",\"start\":1.66,\"end\":1.97}]"} {"utterance_id": "7641-96252-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0015.flac", "task_type": "lexical_detection", "query": "reason", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"reason\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.32, "reason"]], "response_mode": "offline", "answer": "[{\"word\":\"reason\",\"start\":2.91,\"end\":3.32}]"} {"utterance_id": "6841-88294-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.83, 4.25, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":3.83,\"end\":4.25}]"} {"utterance_id": "7641-96252-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0017.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.87, 7.2, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":6.87,\"end\":7.2}]"} {"utterance_id": "3660-172182-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0005.flac", "task_type": "lexical_counting", "query": "know", "prompt": "When the utterance is complete, report how the word \"know\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.89, "one"], [3.46, 3.74, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.64,\"end\":0.89},{\"word\":\"two\",\"start\":3.46,\"end\":3.74}]"} {"utterance_id": "6455-67804-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0017.flac", "task_type": "lexical_detection", "query": "garden", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"garden\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.94, 3.24, "garden"]], "response_mode": "offline", "answer": "[{\"word\":\"garden\",\"start\":2.94,\"end\":3.24}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "lexical_counting", "query": "three", "prompt": "When the utterance is complete, report how the word \"three\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.41, "one"], [5.15, 5.37, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.21,\"end\":0.41},{\"word\":\"two\",\"start\":5.15,\"end\":5.37}]"} {"utterance_id": "7601-175351-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0025.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.14, "one"], [7.88, 8.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.82,\"end\":5.14},{\"word\":\"two\",\"start\":7.88,\"end\":8.44}]"} {"utterance_id": "3660-172182-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.5, "arthur"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur\",\"start\":1.11,\"end\":1.5}]"} {"utterance_id": "4570-24733-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.52, 11.9, "god"], [26.97, 27.57, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":11.52,\"end\":11.9},{\"word\":\"god\",\"start\":26.97,\"end\":27.57}]"} {"utterance_id": "116-288047-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0016.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an organization name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.27, 0.75, "school"]], "response_mode": "offline", "answer": "[{\"word\":\"school\",\"start\":0.27,\"end\":0.75}]"} {"utterance_id": "4515-11057-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0055.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.63, 3.05, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.63,\"end\":3.05}]"} {"utterance_id": "3660-172182-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0014.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Process the full audio offline. Find every instance of a disease name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.45, 4.94, "wounds"]], "response_mode": "offline", "answer": "[{\"word\":\"wounds\",\"start\":4.45,\"end\":4.94}]"} {"utterance_id": "6841-88294-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0009.flac", "task_type": "lexical_detection", "query": "barbaree", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"barbaree\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.79, 12.44, "barbaree"]], "response_mode": "offline", "answer": "[{\"word\":\"barbaree\",\"start\":11.79,\"end\":12.44}]"} {"utterance_id": "1630-102884-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0005.flac", "task_type": "lexical_counting", "query": "walls", "prompt": "Process the full audio offline. For every occurrence of the word \"walls\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.4, 6.88, "one"], [9.66, 10.08, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.4,\"end\":6.88},{\"word\":\"two\",\"start\":9.66,\"end\":10.08}]"} {"utterance_id": "7601-175351-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0012.flac", "task_type": "lexical_detection", "query": "warren", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"warren\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.21, 3.82, "warren"]], "response_mode": "offline", "answer": "[{\"word\":\"warren\",\"start\":3.21,\"end\":3.82}]"} {"utterance_id": "5543-27761-0079", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0079.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.69, 1.07, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":0.69,\"end\":1.07}]"} {"utterance_id": "116-288047-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0018.flac", "task_type": "lexical_detection", "query": "discussion", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"discussion\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 1.14, "discussion"]], "response_mode": "offline", "answer": "[{\"word\":\"discussion\",\"start\":0.64,\"end\":1.14}]"} {"utterance_id": "1255-74899-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0013.flac", "task_type": "lexical_detection", "query": "legislative", "prompt": "Once the audio has finished, identify every instance of the word \"legislative\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.65, 2.5, "legislative"]], "response_mode": "offline", "answer": "[{\"word\":\"legislative\",\"start\":1.65,\"end\":2.5}]"} {"utterance_id": "6455-66379-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0003.flac", "task_type": "lexical_detection", "query": "thompson", "prompt": "When the utterance is complete, report all instances of the word \"thompson\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.52, "thompson"]], "response_mode": "offline", "answer": "[{\"word\":\"thompson\",\"start\":3.15,\"end\":3.52}]"} {"utterance_id": "7641-96684-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0008.flac", "task_type": "lexical_detection", "query": "voice", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"voice\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.35, "voice"], [2.87, 3.23, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"voice\",\"start\":1.06,\"end\":1.35},{\"word\":\"voice\",\"start\":2.87,\"end\":3.23}]"} {"utterance_id": "6455-66379-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.6, 1.83, "six"]], "response_mode": "offline", "answer": "[{\"word\":\"six\",\"start\":1.6,\"end\":1.83}]"} {"utterance_id": "4831-18525-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0017.flac", "task_type": "lexical_detection", "query": "pen", "prompt": "After listening to the whole recording, output every instance of the word \"pen\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.97, 3.22, "pen"]], "response_mode": "offline", "answer": "[{\"word\":\"pen\",\"start\":2.97,\"end\":3.22}]"} {"utterance_id": "6467-94831-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0022.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.96, 1.33, "brow"]], "response_mode": "offline", "answer": "[{\"word\":\"brow\",\"start\":0.96,\"end\":1.33}]"} {"utterance_id": "3660-6517-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0007.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.46, "fifteen"]], "response_mode": "offline", "answer": "[{\"word\":\"fifteen\",\"start\":2.0,\"end\":2.46}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "After the audio ends, tally every instance of a number. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.21, 0.41, "one"], [1.89, 2.24, "two"], [5.15, 5.37, "three"], [10.87, 11.12, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.21,\"end\":0.41},{\"word\":\"two\",\"start\":1.89,\"end\":2.24},{\"word\":\"three\",\"start\":5.15,\"end\":5.37},{\"word\":\"four\",\"start\":10.87,\"end\":11.12}]"} {"utterance_id": "7697-245712-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0008.flac", "task_type": "lexical_detection", "query": "names", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"names\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.65, "names"]], "response_mode": "offline", "answer": "[{\"word\":\"names\",\"start\":4.32,\"end\":4.65}]"} {"utterance_id": "1686-142278-0096", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0096.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.81, 9.25, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"daughter\",\"start\":8.81,\"end\":9.25}]"} {"utterance_id": "3660-6517-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0035.flac", "task_type": "lexical_detection", "query": "saved", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"saved\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.0, 4.32, "saved"]], "response_mode": "offline", "answer": "[{\"word\":\"saved\",\"start\":4.0,\"end\":4.32}]"} {"utterance_id": "1255-74899-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0002.flac", "task_type": "lexical_detection", "query": "written", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"written\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.31, "written"]], "response_mode": "offline", "answer": "[{\"word\":\"written\",\"start\":3.06,\"end\":3.31}]"} {"utterance_id": "6841-88294-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0033.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.39, 1.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.39,\"end\":1.63}]"} {"utterance_id": "700-122866-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0041.flac", "task_type": "lexical_detection", "query": "lynde", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"lynde\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.87, "lynde"]], "response_mode": "offline", "answer": "[{\"word\":\"lynde\",\"start\":0.54,\"end\":0.87}]"} {"utterance_id": "4323-55228-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0050.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.45, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":1.22,\"end\":1.45}]"} {"utterance_id": "6123-59150-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.14, 2.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.14,\"end\":2.28}]"} {"utterance_id": "4153-186222-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0013.flac", "task_type": "lexical_detection", "query": "cent", "prompt": "When the utterance is complete, report all instances of the word \"cent\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.59, "cent"]], "response_mode": "offline", "answer": "[{\"word\":\"cent\",\"start\":1.25,\"end\":1.59}]"} {"utterance_id": "4515-11057-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0022.flac", "task_type": "lexical_detection", "query": "serious", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"serious\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.95, "serious"]], "response_mode": "offline", "answer": "[{\"word\":\"serious\",\"start\":1.15,\"end\":1.95}]"} {"utterance_id": "4323-55228-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0011.flac", "task_type": "lexical_detection", "query": "grave", "prompt": "When the utterance is complete, report all instances of the word \"grave\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.62, 1.93, "grave"]], "response_mode": "offline", "answer": "[{\"word\":\"grave\",\"start\":1.62,\"end\":1.93}]"} {"utterance_id": "8288-274162-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0016.flac", "task_type": "lexical_detection", "query": "d'artagnan", "prompt": "After listening to the whole recording, output every instance of the word \"d'artagnan\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.22, 4.84, "d'artagnan"]], "response_mode": "offline", "answer": "[{\"word\":\"d'artagnan\",\"start\":4.22,\"end\":4.84}]"} {"utterance_id": "6841-88294-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0012.flac", "task_type": "lexical_counting", "query": "he'd", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"he'd\" (word, start, end). If none, reply [].", "labels": [[0.64, 0.79, "one"], [7.71, 7.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.64,\"end\":0.79},{\"word\":\"two\",\"start\":7.71,\"end\":7.86}]"} {"utterance_id": "4572-64670-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0012.flac", "task_type": "lexical_detection", "query": "ere", "prompt": "After listening to the whole recording, output every instance of the word \"ere\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.45, 3.61, "ere"]], "response_mode": "offline", "answer": "[{\"word\":\"ere\",\"start\":3.45,\"end\":3.61}]"} {"utterance_id": "6123-59150-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0036.flac", "task_type": "lexical_detection", "query": "trying", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"trying\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.57, "trying"]], "response_mode": "offline", "answer": "[{\"word\":\"trying\",\"start\":4.23,\"end\":4.57}]"} {"utterance_id": "6841-88294-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0052.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.38, "handy solomon"]], "response_mode": "offline", "answer": "[{\"word\":\"handy solomon\",\"start\":1.68,\"end\":2.38}]"} {"utterance_id": "8254-115543-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0025.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.52, 0.92, "driver"]], "response_mode": "offline", "answer": "[{\"word\":\"driver\",\"start\":0.52,\"end\":0.92}]"} {"utterance_id": "1585-131718-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0045.flac", "task_type": "semantic_counting", "query": "clock_time", "prompt": "After listening to the whole recording, output timed count words for a time of day (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.62, 0.99, "one"], [3.14, 3.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.62,\"end\":0.99},{\"word\":\"two\",\"start\":3.14,\"end\":3.91}]"} {"utterance_id": "700-122868-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0015.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a color name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.4, 4.65, "white"]], "response_mode": "offline", "answer": "[{\"word\":\"white\",\"start\":4.4,\"end\":4.65}]"} {"utterance_id": "3660-6517-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.86, "williams"]], "response_mode": "offline", "answer": "[{\"word\":\"williams\",\"start\":0.5,\"end\":0.86}]"} {"utterance_id": "2506-11267-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0004.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Process the full audio offline. Find every instance of an organization name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.08, 10.7, "vatican"]], "response_mode": "offline", "answer": "[{\"word\":\"vatican\",\"start\":10.08,\"end\":10.7}]"} {"utterance_id": "1630-141772-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0006.flac", "task_type": "lexical_detection", "query": "during", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"during\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.48, 0.83, "during"]], "response_mode": "offline", "answer": "[{\"word\":\"during\",\"start\":0.48,\"end\":0.83}]"} {"utterance_id": "3915-57461-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[13.54, 13.96, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":13.54,\"end\":13.96}]"} {"utterance_id": "8254-84205-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0009.flac", "task_type": "lexical_detection", "query": "swinging", "prompt": "Process the full audio offline. Find every instance of the word \"swinging\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.46, "swinging"]], "response_mode": "offline", "answer": "[{\"word\":\"swinging\",\"start\":1.05,\"end\":1.46}]"} {"utterance_id": "1650-157641-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0005.flac", "task_type": "lexical_detection", "query": "tobacco", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"tobacco\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.55, "tobacco"]], "response_mode": "offline", "answer": "[{\"word\":\"tobacco\",\"start\":3.82,\"end\":4.55}]"} {"utterance_id": "4831-25894-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0028.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.65, 9.03, "tree"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":8.65,\"end\":9.03}]"} {"utterance_id": "6841-88294-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.48, 4.87, "leg"]], "response_mode": "offline", "answer": "[{\"word\":\"leg\",\"start\":4.48,\"end\":4.87}]"} {"utterance_id": "8173-294714-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0036.flac", "task_type": "lexical_detection", "query": "debts", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"debts\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.26, "debts"]], "response_mode": "offline", "answer": "[{\"word\":\"debts\",\"start\":2.91,\"end\":3.26}]"} {"utterance_id": "4323-55228-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0040.flac", "task_type": "lexical_counting", "query": "get", "prompt": "Listen to the entire clip, then report the count sequence for the word \"get\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.23, "one"], [3.66, 3.8, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.04,\"end\":1.23},{\"word\":\"two\",\"start\":3.66,\"end\":3.8}]"} {"utterance_id": "700-122866-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0040.flac", "task_type": "lexical_detection", "query": "felt", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"felt\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.34, 0.57, "felt"]], "response_mode": "offline", "answer": "[{\"word\":\"felt\",\"start\":0.34,\"end\":0.57}]"} {"utterance_id": "8254-115543-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.21, "nao"]], "response_mode": "offline", "answer": "[{\"word\":\"nao\",\"start\":1.98,\"end\":2.21}]"} {"utterance_id": "116-288048-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0012.flac", "task_type": "lexical_detection", "query": "judas", "prompt": "Once the audio has finished, identify every instance of the word \"judas\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.7, 3.27, "judas"]], "response_mode": "offline", "answer": "[{\"word\":\"judas\",\"start\":2.7,\"end\":3.27}]"} {"utterance_id": "4153-185072-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0007.flac", "task_type": "lexical_detection", "query": "woods", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"woods\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.67, "woods"]], "response_mode": "offline", "answer": "[{\"word\":\"woods\",\"start\":1.23,\"end\":1.67}]"} {"utterance_id": "6123-59186-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0011.flac", "task_type": "lexical_counting", "query": "come", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"come\" (word, start, end). If none, reply [].", "labels": [[1.03, 1.48, "one"], [20.81, 21.3, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.03,\"end\":1.48},{\"word\":\"two\",\"start\":20.81,\"end\":21.3}]"} {"utterance_id": "700-122867-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 2.67, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":2.43,\"end\":2.67}]"} {"utterance_id": "2506-11267-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.41, "davis"]], "response_mode": "offline", "answer": "[{\"word\":\"davis\",\"start\":0.95,\"end\":1.41}]"} {"utterance_id": "3915-98647-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0026.flac", "task_type": "lexical_counting", "query": "change", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"change\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[6.99, 7.52, "one"], [12.73, 13.2, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.99,\"end\":7.52},{\"word\":\"two\",\"start\":12.73,\"end\":13.2}]"} {"utterance_id": "3660-6517-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0017.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an ordinal number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.77, "third"]], "response_mode": "offline", "answer": "[{\"word\":\"third\",\"start\":0.52,\"end\":0.77}]"} {"utterance_id": "6467-97061-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0014.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After the audio ends, extract all occurrences of a plant name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.93, 7.54, "lotus flower"]], "response_mode": "offline", "answer": "[{\"word\":\"lotus flower\",\"start\":6.93,\"end\":7.54}]"} {"utterance_id": "3915-57461-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0017.flac", "task_type": "lexical_detection", "query": "hiding", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"hiding\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.36, 0.66, "hiding"]], "response_mode": "offline", "answer": "[{\"word\":\"hiding\",\"start\":0.36,\"end\":0.66}]"} {"utterance_id": "116-288048-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0002.flac", "task_type": "lexical_detection", "query": "document", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"document\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.2, 2.82, "document"]], "response_mode": "offline", "answer": "[{\"word\":\"document\",\"start\":2.2,\"end\":2.82}]"} {"utterance_id": "116-288046-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0005.flac", "task_type": "lexical_detection", "query": "preacher", "prompt": "Once the audio has finished, identify every instance of the word \"preacher\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.55, 1.0, "preacher"]], "response_mode": "offline", "answer": "[{\"word\":\"preacher\",\"start\":0.55,\"end\":1.0}]"} {"utterance_id": "4323-55228-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0022.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.19, 1.44, "one"], [2.79, 3.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.19,\"end\":1.44},{\"word\":\"two\",\"start\":2.79,\"end\":3.18}]"} {"utterance_id": "700-122866-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.11, "ruby gillis"], [2.84, 3.05, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"ruby gillis\",\"start\":0.5,\"end\":1.11},{\"word\":\"anne\",\"start\":2.84,\"end\":3.05}]"} {"utterance_id": "6467-56885-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0009.flac", "task_type": "lexical_detection", "query": "ceremony", "prompt": "Process the full audio offline. Find every instance of the word \"ceremony\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.78, 7.19, "ceremony"]], "response_mode": "offline", "answer": "[{\"word\":\"ceremony\",\"start\":6.78,\"end\":7.19}]"} {"utterance_id": "6455-67803-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0005.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Do not answer until the audio ends. Then list each instance of a time of day as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.67, 0.91, "night"], [3.04, 3.28, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":0.67,\"end\":0.91},{\"word\":\"morning\",\"start\":3.04,\"end\":3.28}]"} {"utterance_id": "1686-142278-0091", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0091.flac", "task_type": "lexical_detection", "query": "easy", "prompt": "After listening to the whole recording, output every instance of the word \"easy\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.77, 5.14, "easy"]], "response_mode": "offline", "answer": "[{\"word\":\"easy\",\"start\":4.77,\"end\":5.14}]"} {"utterance_id": "4323-13259-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0016.flac", "task_type": "lexical_detection", "query": "guided", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"guided\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.8, "guided"]], "response_mode": "offline", "answer": "[{\"word\":\"guided\",\"start\":0.48,\"end\":0.8}]"} {"utterance_id": "6123-59186-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0028.flac", "task_type": "lexical_detection", "query": "loyal", "prompt": "When the utterance is complete, report all instances of the word \"loyal\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.4, 3.79, "loyal"]], "response_mode": "offline", "answer": "[{\"word\":\"loyal\",\"start\":3.4,\"end\":3.79}]"} {"utterance_id": "8288-274162-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0009.flac", "task_type": "lexical_detection", "query": "england", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"england\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.51, "england"]], "response_mode": "offline", "answer": "[{\"word\":\"england\",\"start\":0.98,\"end\":1.51}]"} {"utterance_id": "4572-64670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0024.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Process the full audio offline. Find every instance of a plant name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.36, "tree"], [3.36, 3.85, "grass"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":1.02,\"end\":1.36},{\"word\":\"grass\",\"start\":3.36,\"end\":3.85}]"} {"utterance_id": "1701-141760-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0024.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.85, 2.26, "wine"]], "response_mode": "offline", "answer": "[{\"word\":\"wine\",\"start\":1.85,\"end\":2.26}]"} {"utterance_id": "1650-173552-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0006.flac", "task_type": "lexical_counting", "query": "war", "prompt": "When the utterance is complete, report how the word \"war\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.29, 4.74, "one"], [6.37, 6.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.29,\"end\":4.74},{\"word\":\"two\",\"start\":6.37,\"end\":6.76}]"} {"utterance_id": "5849-50873-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0015.flac", "task_type": "lexical_detection", "query": "greater", "prompt": "Once the audio has finished, identify every instance of the word \"greater\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.66, 2.05, "greater"]], "response_mode": "offline", "answer": "[{\"word\":\"greater\",\"start\":1.66,\"end\":2.05}]"} {"utterance_id": "4153-61735-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0000.flac", "task_type": "lexical_detection", "query": "dialogue", "prompt": "When the utterance is complete, report all instances of the word \"dialogue\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.03, 4.68, "dialogue"]], "response_mode": "offline", "answer": "[{\"word\":\"dialogue\",\"start\":4.03,\"end\":4.68}]"} {"utterance_id": "4323-18416-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.79, 1.13, "phronsie"], [3.07, 3.56, "mister king"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":0.79,\"end\":1.13},{\"word\":\"mister king\",\"start\":3.07,\"end\":3.56}]"} {"utterance_id": "7641-96252-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0013.flac", "task_type": "lexical_detection", "query": "concern", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"concern\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.24, 1.83, "concern"]], "response_mode": "offline", "answer": "[{\"word\":\"concern\",\"start\":1.24,\"end\":1.83}]"} {"utterance_id": "1255-90413-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0002.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After listening to the whole recording, output every instance of a time of day with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.4, 7.82, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":7.4,\"end\":7.82}]"} {"utterance_id": "1585-131718-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0023.flac", "task_type": "lexical_detection", "query": "sold", "prompt": "When the utterance is complete, report all instances of the word \"sold\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 0.6, "sold"]], "response_mode": "offline", "answer": "[{\"word\":\"sold\",\"start\":0.36,\"end\":0.6}]"} {"utterance_id": "8288-274162-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0024.flac", "task_type": "lexical_detection", "query": "curiosity", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"curiosity\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.54, "curiosity"]], "response_mode": "offline", "answer": "[{\"word\":\"curiosity\",\"start\":2.76,\"end\":3.54}]"} {"utterance_id": "7601-175351-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0016.flac", "task_type": "lexical_detection", "query": "beyond", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"beyond\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.4, 3.84, "beyond"]], "response_mode": "offline", "answer": "[{\"word\":\"beyond\",\"start\":3.4,\"end\":3.84}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.08, 1.73, "one"], [5.73, 6.38, "two"], [14.3, 14.76, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.08,\"end\":1.73},{\"word\":\"two\",\"start\":5.73,\"end\":6.38},{\"word\":\"three\",\"start\":14.3,\"end\":14.76}]"} {"utterance_id": "1630-96099-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0003.flac", "task_type": "lexical_detection", "query": "gone", "prompt": "When the utterance is complete, report all instances of the word \"gone\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.29, 0.87, "gone"]], "response_mode": "offline", "answer": "[{\"word\":\"gone\",\"start\":0.29,\"end\":0.87}]"} {"utterance_id": "2506-11278-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0021.flac", "task_type": "lexical_detection", "query": "thinking", "prompt": "When the utterance is complete, report all instances of the word \"thinking\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.83, 4.28, "thinking"]], "response_mode": "offline", "answer": "[{\"word\":\"thinking\",\"start\":3.83,\"end\":4.28}]"} {"utterance_id": "4831-18525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0022.flac", "task_type": "lexical_detection", "query": "face", "prompt": "After listening to the whole recording, output every instance of the word \"face\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 4.01, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.54,\"end\":4.01}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.03, 8.22, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.03,\"end\":8.22}]"} {"utterance_id": "3660-6517-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0016.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Listen to the full audio. After it finishes, find every instance of a date. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.59, 2.91, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":2.59,\"end\":2.91}]"} {"utterance_id": "8173-294714-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0031.flac", "task_type": "lexical_detection", "query": "future", "prompt": "After the audio ends, extract all occurrences of the word \"future\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.09, "future"]], "response_mode": "offline", "answer": "[{\"word\":\"future\",\"start\":1.55,\"end\":2.09}]"} {"utterance_id": "6123-59150-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0030.flac", "task_type": "lexical_detection", "query": "record", "prompt": "After listening to the whole recording, output every instance of the word \"record\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 2.97, "record"], [6.82, 7.21, "record"]], "response_mode": "offline", "answer": "[{\"word\":\"record\",\"start\":2.57,\"end\":2.97},{\"word\":\"record\",\"start\":6.82,\"end\":7.21}]"} {"utterance_id": "6267-53049-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0006.flac", "task_type": "lexical_detection", "query": "listen", "prompt": "When the utterance is complete, report all instances of the word \"listen\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.68, "listen"]], "response_mode": "offline", "answer": "[{\"word\":\"listen\",\"start\":0.23,\"end\":0.68}]"} {"utterance_id": "4323-13259-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0001.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.43, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":2.07,\"end\":2.43}]"} {"utterance_id": "5849-50873-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0033.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.6, "governor"]], "response_mode": "offline", "answer": "[{\"word\":\"governor\",\"start\":2.13,\"end\":2.6}]"} {"utterance_id": "8288-274150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0001.flac", "task_type": "lexical_detection", "query": "possession", "prompt": "After the audio ends, extract all occurrences of the word \"possession\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.43, 12.14, "possession"]], "response_mode": "offline", "answer": "[{\"word\":\"possession\",\"start\":11.43,\"end\":12.14}]"} {"utterance_id": "1585-131718-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.56, 4.88, "edison"]], "response_mode": "offline", "answer": "[{\"word\":\"edison\",\"start\":4.56,\"end\":4.88}]"} {"utterance_id": "116-288047-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.47, 3.14, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":2.47,\"end\":3.14}]"} {"utterance_id": "1630-141772-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0007.flac", "task_type": "lexical_detection", "query": "prisoners", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"prisoners\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 2.12, "prisoners"]], "response_mode": "offline", "answer": "[{\"word\":\"prisoners\",\"start\":1.41,\"end\":2.12}]"} {"utterance_id": "4570-24733-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0006.flac", "task_type": "lexical_detection", "query": "proposition", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"proposition\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.34, 3.17, "proposition"]], "response_mode": "offline", "answer": "[{\"word\":\"proposition\",\"start\":2.34,\"end\":3.17}]"} {"utterance_id": "1686-142278-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.65, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":1.12,\"end\":1.65}]"} {"utterance_id": "6841-88291-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0001.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.67, 5.96, "black"], [7.39, 7.78, "silvery"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":5.67,\"end\":5.96},{\"word\":\"silvery\",\"start\":7.39,\"end\":7.78}]"} {"utterance_id": "7641-96684-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.48, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":1.05,\"end\":1.48}]"} {"utterance_id": "1686-142278-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0066.flac", "task_type": "lexical_detection", "query": "serving", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"serving\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 2.15, "serving"]], "response_mode": "offline", "answer": "[{\"word\":\"serving\",\"start\":1.71,\"end\":2.15}]"} {"utterance_id": "1585-157660-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0010.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.29, 2.71, "dealer"]], "response_mode": "offline", "answer": "[{\"word\":\"dealer\",\"start\":2.29,\"end\":2.71}]"} {"utterance_id": "1585-131718-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0043.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 2.95, "one"], [6.75, 7.39, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.52,\"end\":2.95},{\"word\":\"two\",\"start\":6.75,\"end\":7.39}]"} {"utterance_id": "4153-185072-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0001.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[1.34, 1.93, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":1.34,\"end\":1.93}]"} {"utterance_id": "1630-73710-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.51, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.18,\"end\":1.51}]"} {"utterance_id": "1701-141759-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.7, "sonya"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":1.3,\"end\":1.7}]"} {"utterance_id": "4323-18416-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0001.flac", "task_type": "lexical_detection", "query": "laid", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"laid\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.26, 1.43, "laid"]], "response_mode": "offline", "answer": "[{\"word\":\"laid\",\"start\":1.26,\"end\":1.43}]"} {"utterance_id": "6467-94831-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 1.48, "peleg snuggers"]], "response_mode": "offline", "answer": "[{\"word\":\"peleg snuggers\",\"start\":0.51,\"end\":1.48}]"} {"utterance_id": "3663-172528-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0016.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.01, 4.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":4.01,\"end\":4.28}]"} {"utterance_id": "6123-59186-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0027.flac", "task_type": "lexical_counting", "query": "souls", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"souls\" with word/start/end. If none, reply [].", "labels": [[7.64, 8.05, "one"], [9.86, 10.22, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.64,\"end\":8.05},{\"word\":\"two\",\"start\":9.86,\"end\":10.22}]"} {"utterance_id": "7601-175351-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0016.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 1.04, "one"], [6.26, 7.67, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.39,\"end\":1.04},{\"word\":\"two\",\"start\":6.26,\"end\":7.67}]"} {"utterance_id": "700-122866-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0015.flac", "task_type": "lexical_detection", "query": "since", "prompt": "Once the audio has finished, identify every instance of the word \"since\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.83, 3.04, "since"]], "response_mode": "offline", "answer": "[{\"word\":\"since\",\"start\":2.83,\"end\":3.04}]"} {"utterance_id": "5849-50964-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.61, 4.14, "downy"], [5.21, 5.54, "snoop"]], "response_mode": "offline", "answer": "[{\"word\":\"downy\",\"start\":3.61,\"end\":4.14},{\"word\":\"snoop\",\"start\":5.21,\"end\":5.54}]"} {"utterance_id": "3663-172528-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0007.flac", "task_type": "lexical_detection", "query": "helping", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"helping\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.76, 4.23, "helping"]], "response_mode": "offline", "answer": "[{\"word\":\"helping\",\"start\":3.76,\"end\":4.23}]"} {"utterance_id": "4572-112381-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0000.flac", "task_type": "lexical_detection", "query": "comrades", "prompt": "After the audio ends, extract all occurrences of the word \"comrades\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.32, "comrades"]], "response_mode": "offline", "answer": "[{\"word\":\"comrades\",\"start\":0.77,\"end\":1.32}]"} {"utterance_id": "1701-141759-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0020.flac", "task_type": "lexical_detection", "query": "don't", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"don't\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.92, "don't"]], "response_mode": "offline", "answer": "[{\"word\":\"don't\",\"start\":0.68,\"end\":0.92}]"} {"utterance_id": "7641-96252-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.84, 8.24, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.84,\"end\":8.24}]"} {"utterance_id": "6599-38590-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.46, 3.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":3.46,\"end\":3.63}]"} {"utterance_id": "4831-29134-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0001.flac", "task_type": "lexical_detection", "query": "spit", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"spit\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.34, 5.65, "spit"]], "response_mode": "offline", "answer": "[{\"word\":\"spit\",\"start\":5.34,\"end\":5.65}]"} {"utterance_id": "1701-141759-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 1.34, "anna mikhaylovna"]], "response_mode": "offline", "answer": "[{\"word\":\"anna mikhaylovna\",\"start\":0.37,\"end\":1.34}]"} {"utterance_id": "4323-13259-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.28, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":0.94,\"end\":1.28}]"} {"utterance_id": "5543-27761-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0018.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.92, "mother's"]], "response_mode": "offline", "answer": "[{\"word\":\"mother's\",\"start\":0.54,\"end\":0.92}]"} {"utterance_id": "7641-96670-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.25, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"daughter\",\"start\":1.82,\"end\":2.25}]"} {"utterance_id": "1686-142278-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0033.flac", "task_type": "lexical_counting", "query": "thou", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"thou\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.86, 1.06, "one"], [5.96, 6.16, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.86,\"end\":1.06},{\"word\":\"two\",\"start\":5.96,\"end\":6.16}]"} {"utterance_id": "6455-67804-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.08, "octavia"], [5.06, 5.72, "missus burnham"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":0.54,\"end\":1.08},{\"word\":\"missus burnham\",\"start\":5.06,\"end\":5.72}]"} {"utterance_id": "7641-96252-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0002.flac", "task_type": "lexical_detection", "query": "modified", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"modified\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.63, 5.15, "modified"]], "response_mode": "offline", "answer": "[{\"word\":\"modified\",\"start\":4.63,\"end\":5.15}]"} {"utterance_id": "4515-11057-0081", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0081.flac", "task_type": "lexical_detection", "query": "father", "prompt": "Once the audio has finished, identify every instance of the word \"father\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.68, 3.16, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.68,\"end\":3.16}]"} {"utterance_id": "1650-173552-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0005.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Process the full audio offline. Find every instance of an organization name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 2.15, "kremlin"], [3.98, 4.63, "kremlin"]], "response_mode": "offline", "answer": "[{\"word\":\"kremlin\",\"start\":1.48,\"end\":2.15},{\"word\":\"kremlin\",\"start\":3.98,\"end\":4.63}]"} {"utterance_id": "8173-294714-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0048.flac", "task_type": "lexical_detection", "query": "mistaken", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mistaken\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.06, 1.55, "mistaken"]], "response_mode": "offline", "answer": "[{\"word\":\"mistaken\",\"start\":1.06,\"end\":1.55}]"} {"utterance_id": "7697-245712-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0014.flac", "task_type": "lexical_counting", "query": "regards", "prompt": "Process the full audio offline. For every occurrence of the word \"regards\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.77, 7.15, "one"], [12.22, 12.62, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.77,\"end\":7.15},{\"word\":\"two\",\"start\":12.22,\"end\":12.62}]"} {"utterance_id": "3660-172183-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.5, 1.91, "erbin"]], "response_mode": "offline", "answer": "[{\"word\":\"erbin\",\"start\":1.5,\"end\":1.91}]"} {"utterance_id": "7641-96684-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.19, 1.75, "admetus"], [3.04, 3.55, "jason"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":1.19,\"end\":1.75},{\"word\":\"jason\",\"start\":3.04,\"end\":3.55}]"} {"utterance_id": "700-122866-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0012.flac", "task_type": "lexical_detection", "query": "actually", "prompt": "After the audio ends, extract all occurrences of the word \"actually\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 1.02, "actually"]], "response_mode": "offline", "answer": "[{\"word\":\"actually\",\"start\":0.61,\"end\":1.02}]"} {"utterance_id": "1585-131718-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0030.flac", "task_type": "lexical_detection", "query": "worked", "prompt": "After listening to the whole recording, output every instance of the word \"worked\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.55, 2.88, "worked"]], "response_mode": "offline", "answer": "[{\"word\":\"worked\",\"start\":2.55,\"end\":2.88}]"} {"utterance_id": "4515-11057-0069", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0069.flac", "task_type": "lexical_detection", "query": "willingly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"willingly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.93, 2.52, "willingly"]], "response_mode": "offline", "answer": "[{\"word\":\"willingly\",\"start\":1.93,\"end\":2.52}]"} {"utterance_id": "4831-25894-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.61, 1.84, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.61,\"end\":1.84}]"} {"utterance_id": "3663-172528-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.8, "felice"]], "response_mode": "offline", "answer": "[{\"word\":\"felice\",\"start\":0.37,\"end\":0.8}]"} {"utterance_id": "3663-172005-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0004.flac", "task_type": "lexical_detection", "query": "gun", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"gun\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.53, 2.95, "gun"]], "response_mode": "offline", "answer": "[{\"word\":\"gun\",\"start\":2.53,\"end\":2.95}]"} {"utterance_id": "700-122867-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0021.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.61, 3.97, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":3.61,\"end\":3.97}]"} {"utterance_id": "6467-94831-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0002.flac", "task_type": "lexical_detection", "query": "strong", "prompt": "Once the audio has finished, identify every instance of the word \"strong\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.65, 2.11, "strong"]], "response_mode": "offline", "answer": "[{\"word\":\"strong\",\"start\":1.65,\"end\":2.11}]"} {"utterance_id": "1255-90413-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0027.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.52, "lord justice"]], "response_mode": "offline", "answer": "[{\"word\":\"lord justice\",\"start\":2.76,\"end\":3.52}]"} {"utterance_id": "4831-18525-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0017.flac", "task_type": "lexical_counting", "query": "polly", "prompt": "Once the audio has finished, list each occurrence of the word \"polly\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.66, 1.05, "one"], [3.38, 3.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.66,\"end\":1.05},{\"word\":\"two\",\"start\":3.38,\"end\":3.7}]"} {"utterance_id": "7697-105817-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0000.flac", "task_type": "lexical_detection", "query": "hath", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"hath\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.6, "hath"]], "response_mode": "offline", "answer": "[{\"word\":\"hath\",\"start\":1.41,\"end\":1.6}]"} {"utterance_id": "700-122866-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0014.flac", "task_type": "lexical_detection", "query": "nose", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"nose\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.21, 2.86, "nose"]], "response_mode": "offline", "answer": "[{\"word\":\"nose\",\"start\":2.21,\"end\":2.86}]"} {"utterance_id": "3915-98647-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0003.flac", "task_type": "lexical_detection", "query": "temples", "prompt": "When the utterance is complete, report all instances of the word \"temples\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.71, "temples"]], "response_mode": "offline", "answer": "[{\"word\":\"temples\",\"start\":1.37,\"end\":1.71}]"} {"utterance_id": "7601-101619-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[22.2, 22.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":22.2,\"end\":22.7}]"} {"utterance_id": "700-122866-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0005.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "You will hear a complete utterance. When it ends, list every instance of an organization name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 2.9, "school"]], "response_mode": "offline", "answer": "[{\"word\":\"school\",\"start\":2.52,\"end\":2.9}]"} {"utterance_id": "4831-18525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0011.flac", "task_type": "lexical_detection", "query": "i'd", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"i'd\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.2, "i'd"]], "response_mode": "offline", "answer": "[{\"word\":\"i'd\",\"start\":3.05,\"end\":3.2}]"} {"utterance_id": "4153-186223-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0019.flac", "task_type": "lexical_detection", "query": "away", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"away\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.37, "away"]], "response_mode": "offline", "answer": "[{\"word\":\"away\",\"start\":0.96,\"end\":1.37}]"} {"utterance_id": "4831-18525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0015.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[1.56, 1.99, "one"], [6.1, 6.38, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.56,\"end\":1.99},{\"word\":\"two\",\"start\":6.1,\"end\":6.38}]"} {"utterance_id": "6467-97061-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0002.flac", "task_type": "lexical_detection", "query": "stake", "prompt": "When the utterance is complete, report all instances of the word \"stake\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.05, 6.44, "stake"]], "response_mode": "offline", "answer": "[{\"word\":\"stake\",\"start\":6.05,\"end\":6.44}]"} {"utterance_id": "6123-59150-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0021.flac", "task_type": "lexical_detection", "query": "brutal", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"brutal\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.28, 7.58, "brutal"]], "response_mode": "offline", "answer": "[{\"word\":\"brutal\",\"start\":7.28,\"end\":7.58}]"} {"utterance_id": "3660-172183-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0011.flac", "task_type": "lexical_detection", "query": "unless", "prompt": "Once the audio has finished, identify every instance of the word \"unless\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.56, 0.9, "unless"]], "response_mode": "offline", "answer": "[{\"word\":\"unless\",\"start\":0.56,\"end\":0.9}]"} {"utterance_id": "4831-18525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0022.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.54, 4.01, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.54,\"end\":4.01}]"} {"utterance_id": "7641-96684-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0009.flac", "task_type": "semantic_counting", "query": "color", "prompt": "After the audio ends, tally every instance of a color name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[5.89, 6.15, "one"], [7.56, 7.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.89,\"end\":6.15},{\"word\":\"two\",\"start\":7.56,\"end\":7.86}]"} {"utterance_id": "6841-88294-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0008.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Once the audio has finished, identify every instance of a disease name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.65, 1.98, "wound"]], "response_mode": "offline", "answer": "[{\"word\":\"wound\",\"start\":1.65,\"end\":1.98}]"} {"utterance_id": "1650-167613-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0042.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a body part (word, start, end). If none, reply [].", "labels": [[3.89, 4.27, "one"], [6.69, 7.07, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.89,\"end\":4.27},{\"word\":\"two\",\"start\":6.69,\"end\":7.07}]"} {"utterance_id": "8254-115543-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0030.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.13, 3.57, "parao"]], "response_mode": "offline", "answer": "[{\"word\":\"parao\",\"start\":3.13,\"end\":3.57}]"} {"utterance_id": "7641-96684-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0002.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.75, "one"], [3.04, 3.55, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.19,\"end\":1.75},{\"word\":\"two\",\"start\":3.04,\"end\":3.55}]"} {"utterance_id": "3663-172528-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0022.flac", "task_type": "lexical_detection", "query": "prepared", "prompt": "When the utterance is complete, report all instances of the word \"prepared\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.55, 3.12, "prepared"]], "response_mode": "offline", "answer": "[{\"word\":\"prepared\",\"start\":2.55,\"end\":3.12}]"} {"utterance_id": "6467-97061-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0014.flac", "task_type": "lexical_detection", "query": "flying", "prompt": "After listening to the whole recording, output every instance of the word \"flying\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.44, "flying"]], "response_mode": "offline", "answer": "[{\"word\":\"flying\",\"start\":2.01,\"end\":2.44}]"} {"utterance_id": "8288-274150-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0007.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.91, 8.2, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.91,\"end\":8.2}]"} {"utterance_id": "6267-65525-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0028.flac", "task_type": "lexical_detection", "query": "understanding", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"understanding\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.71, 2.56, "understanding"]], "response_mode": "offline", "answer": "[{\"word\":\"understanding\",\"start\":1.71,\"end\":2.56}]"} {"utterance_id": "700-122866-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0036.flac", "task_type": "lexical_detection", "query": "agreed", "prompt": "When the utterance is complete, report all instances of the word \"agreed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.46, "agreed"]], "response_mode": "offline", "answer": "[{\"word\":\"agreed\",\"start\":3.21,\"end\":3.46}]"} {"utterance_id": "8173-294714-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0014.flac", "task_type": "lexical_counting", "query": "pride", "prompt": "Listen to the full audio. Count every occurrence of the word \"pride\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.45, 3.81, "one"], [5.07, 5.62, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.45,\"end\":3.81},{\"word\":\"two\",\"start\":5.07,\"end\":5.62}]"} {"utterance_id": "7641-96670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.19, 3.46, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":3.19,\"end\":3.46}]"} {"utterance_id": "7641-96252-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0021.flac", "task_type": "lexical_detection", "query": "sufferer", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"sufferer\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.62, 3.13, "sufferer"]], "response_mode": "offline", "answer": "[{\"word\":\"sufferer\",\"start\":2.62,\"end\":3.13}]"} {"utterance_id": "4153-185072-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.12, "thomas"], [2.16, 2.54, "allen's"]], "response_mode": "offline", "answer": "[{\"word\":\"thomas\",\"start\":0.71,\"end\":1.12},{\"word\":\"allen's\",\"start\":2.16,\"end\":2.54}]"} {"utterance_id": "1701-141760-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0005.flac", "task_type": "lexical_counting", "query": "clean", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"clean\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[5.23, 5.63, "one"], [8.58, 8.89, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.23,\"end\":5.63},{\"word\":\"two\",\"start\":8.58,\"end\":8.89}]"} {"utterance_id": "6467-94831-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 3.94, "fred garrison larry colby"]], "response_mode": "offline", "answer": "[{\"word\":\"fred garrison larry colby\",\"start\":1.52,\"end\":3.94}]"} {"utterance_id": "1585-157660-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0002.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[3.48, 4.04, "three hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"three hundred\",\"start\":3.48,\"end\":4.04}]"} {"utterance_id": "4153-61735-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0002.flac", "task_type": "lexical_detection", "query": "southern", "prompt": "After the audio ends, extract all occurrences of the word \"southern\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.73, 5.2, "southern"]], "response_mode": "offline", "answer": "[{\"word\":\"southern\",\"start\":4.73,\"end\":5.2}]"} {"utterance_id": "7641-96684-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0028.flac", "task_type": "lexical_detection", "query": "heracles", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"heracles\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.71, 1.3, "heracles"]], "response_mode": "offline", "answer": "[{\"word\":\"heracles\",\"start\":0.71,\"end\":1.3}]"} {"utterance_id": "7601-175351-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0003.flac", "task_type": "lexical_detection", "query": "others", "prompt": "When the utterance is complete, report all instances of the word \"others\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.76, 2.38, "others"]], "response_mode": "offline", "answer": "[{\"word\":\"others\",\"start\":1.76,\"end\":2.38}]"} {"utterance_id": "6599-38590-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0008.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.27, 9.59, "minds"]], "response_mode": "offline", "answer": "[{\"word\":\"minds\",\"start\":9.27,\"end\":9.59}]"} {"utterance_id": "4831-29134-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0016.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Do not answer until the audio ends. Then list each instance of a date as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.62, 1.98, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":1.62,\"end\":1.98}]"} {"utterance_id": "4323-55228-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.62, 3.93, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":3.62,\"end\":3.93}]"} {"utterance_id": "6267-65525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.66, "sidney"], [6.8, 7.62, "aunt jane"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":3.05,\"end\":3.66},{\"word\":\"aunt jane\",\"start\":6.8,\"end\":7.62}]"} {"utterance_id": "8254-115543-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0035.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a profession: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.03, 2.43, "one"], [12.44, 12.82, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.03,\"end\":2.43},{\"word\":\"two\",\"start\":12.44,\"end\":12.82}]"} {"utterance_id": "6599-38591-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0011.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.14, 8.5, "supper"]], "response_mode": "offline", "answer": "[{\"word\":\"supper\",\"start\":8.14,\"end\":8.5}]"} {"utterance_id": "4153-61735-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0006.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.17, 2.75, "copse"]], "response_mode": "offline", "answer": "[{\"word\":\"copse\",\"start\":2.17,\"end\":2.75}]"} {"utterance_id": "6841-88294-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0048.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.56, 3.04, "treasure"]], "response_mode": "offline", "answer": "[{\"word\":\"treasure\",\"start\":2.56,\"end\":3.04}]"} {"utterance_id": "6267-65525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.5, 4.49, "mister lincoln"]], "response_mode": "offline", "answer": "[{\"word\":\"mister lincoln\",\"start\":3.5,\"end\":4.49}]"} {"utterance_id": "6467-94831-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0005.flac", "task_type": "lexical_detection", "query": "hullo", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hullo\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.58, 0.98, "hullo"]], "response_mode": "offline", "answer": "[{\"word\":\"hullo\",\"start\":0.58,\"end\":0.98}]"} {"utterance_id": "8254-115543-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0017.flac", "task_type": "lexical_detection", "query": "chola", "prompt": "After the audio ends, extract all occurrences of the word \"chola\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.38, 5.76, "chola"]], "response_mode": "offline", "answer": "[{\"word\":\"chola\",\"start\":5.38,\"end\":5.76}]"} {"utterance_id": "1585-157660-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.8, "footsteps"], [9.24, 9.62, "lap"]], "response_mode": "offline", "answer": "[{\"word\":\"footsteps\",\"start\":2.24,\"end\":2.8},{\"word\":\"lap\",\"start\":9.24,\"end\":9.62}]"} {"utterance_id": "2506-11267-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0007.flac", "task_type": "lexical_detection", "query": "leaves", "prompt": "Process the full audio offline. Find every instance of the word \"leaves\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.79, "leaves"]], "response_mode": "offline", "answer": "[{\"word\":\"leaves\",\"start\":1.36,\"end\":1.79}]"} {"utterance_id": "1585-131718-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0051.flac", "task_type": "lexical_detection", "query": "protect", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"protect\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.42, "protect"]], "response_mode": "offline", "answer": "[{\"word\":\"protect\",\"start\":1.07,\"end\":1.42}]"} {"utterance_id": "8173-294714-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0008.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.78, 5.2, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.78,\"end\":5.2}]"} {"utterance_id": "700-122867-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0000.flac", "task_type": "lexical_detection", "query": "vexation", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"vexation\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.68, "vexation"]], "response_mode": "offline", "answer": "[{\"word\":\"vexation\",\"start\":1.07,\"end\":1.68}]"} {"utterance_id": "4323-18416-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0006.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.86, "one"], [2.9, 3.23, "two"], [5.05, 5.41, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.39,\"end\":1.86},{\"word\":\"two\",\"start\":2.9,\"end\":3.23},{\"word\":\"three\",\"start\":5.05,\"end\":5.41}]"} {"utterance_id": "1686-142278-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0001.flac", "task_type": "lexical_detection", "query": "blue", "prompt": "When the utterance is complete, report all instances of the word \"blue\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.15, "blue"]], "response_mode": "offline", "answer": "[{\"word\":\"blue\",\"start\":0.95,\"end\":1.15}]"} {"utterance_id": "3915-57461-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0004.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.55, 0.93, "boatmen"], [2.94, 3.28, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"boatmen\",\"start\":0.55,\"end\":0.93},{\"word\":\"priest\",\"start\":2.94,\"end\":3.28}]"} {"utterance_id": "8254-84205-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0057.flac", "task_type": "lexical_detection", "query": "once", "prompt": "After listening to the whole recording, output every instance of the word \"once\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 2.74, "once"]], "response_mode": "offline", "answer": "[{\"word\":\"once\",\"start\":2.43,\"end\":2.74}]"} {"utterance_id": "7641-96670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0024.flac", "task_type": "lexical_detection", "query": "ground", "prompt": "Once the audio has finished, identify every instance of the word \"ground\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.79, 4.24, "ground"]], "response_mode": "offline", "answer": "[{\"word\":\"ground\",\"start\":3.79,\"end\":4.24}]"} {"utterance_id": "7641-96684-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.95, 2.5, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":1.95,\"end\":2.5}]"} {"utterance_id": "1585-131718-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0041.flac", "task_type": "lexical_detection", "query": "remember", "prompt": "Process the full audio offline. Find every instance of the word \"remember\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.45, "remember"]], "response_mode": "offline", "answer": "[{\"word\":\"remember\",\"start\":0.97,\"end\":1.45}]"} {"utterance_id": "4831-18525-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0014.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Process the full audio offline. Find every instance of a money or currency term and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.31, 4.63, "franc"]], "response_mode": "offline", "answer": "[{\"word\":\"franc\",\"start\":4.31,\"end\":4.63}]"} {"utterance_id": "3915-98647-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0005.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.47, 5.0, "mother's"]], "response_mode": "offline", "answer": "[{\"word\":\"mother's\",\"start\":4.47,\"end\":5.0}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "After listening to the whole recording, output timed count words for an organization name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.95, 3.47, "one"], [9.76, 10.4, "two"], [12.03, 12.68, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.95,\"end\":3.47},{\"word\":\"two\",\"start\":9.76,\"end\":10.4},{\"word\":\"three\",\"start\":12.03,\"end\":12.68}]"} {"utterance_id": "8254-84205-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0059.flac", "task_type": "lexical_detection", "query": "water", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"water\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.79, "water"]], "response_mode": "offline", "answer": "[{\"word\":\"water\",\"start\":1.44,\"end\":1.79}]"} {"utterance_id": "6455-67803-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0005.flac", "task_type": "lexical_detection", "query": "talk", "prompt": "After the audio ends, extract all occurrences of the word \"talk\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.66, 2.05, "talk"]], "response_mode": "offline", "answer": "[{\"word\":\"talk\",\"start\":1.66,\"end\":2.05}]"} {"utterance_id": "3660-172183-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.51, "enid"]], "response_mode": "offline", "answer": "[{\"word\":\"enid\",\"start\":1.23,\"end\":1.51}]"} {"utterance_id": "6123-59186-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.39, 2.0, "mister hardy"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy\",\"start\":1.39,\"end\":2.0}]"} {"utterance_id": "6841-88294-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0063.flac", "task_type": "semantic_detection", "query": "money", "prompt": "When the utterance is complete, report all instances of a money or currency term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.75, "money"], [2.07, 2.36, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":0.45,\"end\":0.75},{\"word\":\"money\",\"start\":2.07,\"end\":2.36}]"} {"utterance_id": "700-122868-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.65, 1.09, "elaine"]], "response_mode": "offline", "answer": "[{\"word\":\"elaine\",\"start\":0.65,\"end\":1.09}]"} {"utterance_id": "700-122868-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0022.flac", "task_type": "lexical_detection", "query": "manner", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"manner\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.24, 2.51, "manner"]], "response_mode": "offline", "answer": "[{\"word\":\"manner\",\"start\":2.24,\"end\":2.51}]"} {"utterance_id": "1585-131718-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0025.flac", "task_type": "lexical_detection", "query": "incidents", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"incidents\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.06, 1.73, "incidents"]], "response_mode": "offline", "answer": "[{\"word\":\"incidents\",\"start\":1.06,\"end\":1.73}]"} {"utterance_id": "5543-27761-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0018.flac", "task_type": "lexical_detection", "query": "passionate", "prompt": "Once the audio has finished, identify every instance of the word \"passionate\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.75, 2.27, "passionate"]], "response_mode": "offline", "answer": "[{\"word\":\"passionate\",\"start\":1.75,\"end\":2.27}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "lexical_detection", "query": "imaginable", "prompt": "Process the full audio offline. Find every instance of the word \"imaginable\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[16.84, 17.6, "imaginable"]], "response_mode": "offline", "answer": "[{\"word\":\"imaginable\",\"start\":16.84,\"end\":17.6}]"} {"utterance_id": "4323-55228-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0009.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.5, "one"], [4.44, 4.97, "two"], [6.69, 7.01, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.23,\"end\":0.5},{\"word\":\"two\",\"start\":4.44,\"end\":4.97},{\"word\":\"three\",\"start\":6.69,\"end\":7.01}]"} {"utterance_id": "6267-65525-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0012.flac", "task_type": "lexical_detection", "query": "excitement", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"excitement\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.31, 4.0, "excitement"]], "response_mode": "offline", "answer": "[{\"word\":\"excitement\",\"start\":3.31,\"end\":4.0}]"} {"utterance_id": "1701-141760-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0002.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "Listen to the audio in full, then emit a JSON array of count events for an organization name (word, start, end). If none, reply [].", "labels": [[0.64, 1.14, "one"], [3.22, 3.6, "two"], [11.19, 11.54, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.64,\"end\":1.14},{\"word\":\"two\",\"start\":3.22,\"end\":3.6},{\"word\":\"three\",\"start\":11.19,\"end\":11.54}]"} {"utterance_id": "1650-167613-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0050.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.78, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":1.99,\"end\":2.78}]"} {"utterance_id": "4831-29134-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0004.flac", "task_type": "lexical_detection", "query": "populous", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"populous\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.54, "populous"]], "response_mode": "offline", "answer": "[{\"word\":\"populous\",\"start\":0.99,\"end\":1.54}]"} {"utterance_id": "7697-105815-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.68, 3.87, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":3.68,\"end\":3.87}]"} {"utterance_id": "4323-18416-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0030.flac", "task_type": "lexical_detection", "query": "hang", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"hang\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.36, 0.62, "hang"]], "response_mode": "offline", "answer": "[{\"word\":\"hang\",\"start\":0.36,\"end\":0.62}]"} {"utterance_id": "116-288045-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0031.flac", "task_type": "lexical_detection", "query": "impatiently", "prompt": "Process the full audio offline. Find every instance of the word \"impatiently\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.4, 4.15, "impatiently"]], "response_mode": "offline", "answer": "[{\"word\":\"impatiently\",\"start\":3.4,\"end\":4.15}]"} {"utterance_id": "4515-11057-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.17, 1.57, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":1.17,\"end\":1.57}]"} {"utterance_id": "1585-131718-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0026.flac", "task_type": "lexical_detection", "query": "feeling", "prompt": "Process the full audio offline. Find every instance of the word \"feeling\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.04, "feeling"]], "response_mode": "offline", "answer": "[{\"word\":\"feeling\",\"start\":0.69,\"end\":1.04}]"} {"utterance_id": "1255-90407-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0009.flac", "task_type": "lexical_detection", "query": "bigger", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"bigger\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.68, "bigger"]], "response_mode": "offline", "answer": "[{\"word\":\"bigger\",\"start\":1.45,\"end\":1.68}]"} {"utterance_id": "4515-11057-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.89, 3.25, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.89,\"end\":3.25}]"} {"utterance_id": "4323-13259-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.97, "miss page"]], "response_mode": "offline", "answer": "[{\"word\":\"miss page\",\"start\":1.48,\"end\":1.97}]"} {"utterance_id": "1686-142278-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0052.flac", "task_type": "lexical_counting", "query": "love", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"love\" with word/start/end. If none, reply [].", "labels": [[6.73, 6.96, "one"], [10.46, 10.77, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.73,\"end\":6.96},{\"word\":\"two\",\"start\":10.46,\"end\":10.77}]"} {"utterance_id": "4570-102353-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0007.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Listen to the full audio. Count every occurrence of a family relation. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.17, 9.65, "one"], [13.13, 13.65, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.17,\"end\":9.65},{\"word\":\"two\",\"start\":13.13,\"end\":13.65}]"} {"utterance_id": "6123-59186-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0006.flac", "task_type": "lexical_detection", "query": "choice", "prompt": "Once the audio has finished, identify every instance of the word \"choice\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.69, 2.04, "choice"]], "response_mode": "offline", "answer": "[{\"word\":\"choice\",\"start\":1.69,\"end\":2.04}]"} {"utterance_id": "4323-13259-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.94, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":1.59,\"end\":1.94}]"} {"utterance_id": "8288-274162-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0019.flac", "task_type": "lexical_detection", "query": "wretched", "prompt": "Once the audio has finished, identify every instance of the word \"wretched\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.73, 6.25, "wretched"]], "response_mode": "offline", "answer": "[{\"word\":\"wretched\",\"start\":5.73,\"end\":6.25}]"} {"utterance_id": "3660-172182-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0020.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.16, 9.69, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":9.16,\"end\":9.69}]"} {"utterance_id": "6467-97061-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0001.flac", "task_type": "lexical_detection", "query": "hearing", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hearing\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.1, 9.48, "hearing"]], "response_mode": "offline", "answer": "[{\"word\":\"hearing\",\"start\":9.1,\"end\":9.48}]"} {"utterance_id": "6467-94831-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.84, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":0.54,\"end\":0.84}]"} {"utterance_id": "1701-141759-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0027.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[11.59, 11.9, "both"]], "response_mode": "offline", "answer": "[{\"word\":\"both\",\"start\":11.59,\"end\":11.9}]"} {"utterance_id": "5849-50873-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.83, "charles"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":0.5,\"end\":0.83}]"} {"utterance_id": "1630-73710-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0006.flac", "task_type": "lexical_detection", "query": "bird", "prompt": "After the audio ends, extract all occurrences of the word \"bird\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 2.75, "bird"]], "response_mode": "offline", "answer": "[{\"word\":\"bird\",\"start\":2.43,\"end\":2.75}]"} {"utterance_id": "6455-67803-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0027.flac", "task_type": "lexical_detection", "query": "laughed", "prompt": "Once the audio has finished, identify every instance of the word \"laughed\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.69, 3.08, "laughed"]], "response_mode": "offline", "answer": "[{\"word\":\"laughed\",\"start\":2.69,\"end\":3.08}]"} {"utterance_id": "6467-94831-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.91, 1.29, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":0.91,\"end\":1.29}]"} {"utterance_id": "8254-84205-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.55, 2.95, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":2.55,\"end\":2.95}]"} {"utterance_id": "4831-18525-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0009.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 2.11, "martigny"]], "response_mode": "offline", "answer": "[{\"word\":\"martigny\",\"start\":1.64,\"end\":2.11}]"} {"utterance_id": "1650-167613-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0033.flac", "task_type": "lexical_detection", "query": "destroy", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"destroy\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.1, 4.65, "destroy"]], "response_mode": "offline", "answer": "[{\"word\":\"destroy\",\"start\":4.1,\"end\":4.65}]"} {"utterance_id": "6467-94831-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 2.11, "mister strong"]], "response_mode": "offline", "answer": "[{\"word\":\"mister strong\",\"start\":1.37,\"end\":2.11}]"} {"utterance_id": "4515-11057-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0063.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen to the full audio. After it finishes, find every instance of a money or currency term. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.9, 5.25, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":4.9,\"end\":5.25}]"} {"utterance_id": "1651-136854-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0007.flac", "task_type": "lexical_detection", "query": "pursue", "prompt": "Once the audio has finished, identify every instance of the word \"pursue\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.89, 9.56, "pursue"]], "response_mode": "offline", "answer": "[{\"word\":\"pursue\",\"start\":8.89,\"end\":9.56}]"} {"utterance_id": "4515-11057-0100", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0100.flac", "task_type": "semantic_counting", "query": "weather", "prompt": "Once the audio has finished, list each occurrence of a weather term as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.67, 0.99, "one"], [5.41, 5.93, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.67,\"end\":0.99},{\"word\":\"two\",\"start\":5.41,\"end\":5.93}]"} {"utterance_id": "7697-245712-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.55, 4.13, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":3.55,\"end\":4.13}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.29, 2.95, "thirty seven"], [5.12, 5.47, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"thirty seven\",\"start\":2.29,\"end\":2.95},{\"word\":\"one\",\"start\":5.12,\"end\":5.47}]"} {"utterance_id": "5849-50873-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0030.flac", "task_type": "lexical_detection", "query": "charter", "prompt": "When the utterance is complete, report all instances of the word \"charter\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.6, "charter"]], "response_mode": "offline", "answer": "[{\"word\":\"charter\",\"start\":1.06,\"end\":1.6}]"} {"utterance_id": "116-288045-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0025.flac", "task_type": "lexical_detection", "query": "search", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"search\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.13, "search"]], "response_mode": "offline", "answer": "[{\"word\":\"search\",\"start\":0.73,\"end\":1.13}]"} {"utterance_id": "7641-96670-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.68, 3.07, "jason"], [4.68, 5.42, "phrixus"]], "response_mode": "offline", "answer": "[{\"word\":\"jason\",\"start\":2.68,\"end\":3.07},{\"word\":\"phrixus\",\"start\":4.68,\"end\":5.42}]"} {"utterance_id": "6267-53049-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0024.flac", "task_type": "lexical_detection", "query": "fairweather", "prompt": "Once the audio has finished, identify every instance of the word \"fairweather\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.31, 3.41, "fairweather"]], "response_mode": "offline", "answer": "[{\"word\":\"fairweather\",\"start\":2.31,\"end\":3.41}]"} {"utterance_id": "700-122866-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0013.flac", "task_type": "lexical_detection", "query": "possibly", "prompt": "Once the audio has finished, identify every instance of the word \"possibly\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.81, 3.45, "possibly"]], "response_mode": "offline", "answer": "[{\"word\":\"possibly\",\"start\":2.81,\"end\":3.45}]"} {"utterance_id": "6467-94831-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 1.87, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":1.49,\"end\":1.87}]"} {"utterance_id": "1686-142278-0073", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0073.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.5, 1.02, "one"], [8.26, 8.93, "two"], [10.47, 11.49, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":1.02},{\"word\":\"two\",\"start\":8.26,\"end\":8.93},{\"word\":\"three\",\"start\":10.47,\"end\":11.49}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a time of day with fields word, start, end (seconds). If none, reply [].", "labels": [[12.14, 12.5, "morning"], [15.01, 15.42, "nine"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":12.14,\"end\":12.5},{\"word\":\"nine\",\"start\":15.01,\"end\":15.42}]"} {"utterance_id": "4153-186222-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0012.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "When the utterance is complete, report how a family relation accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.85, "one"], [2.33, 2.61, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.51,\"end\":0.85},{\"word\":\"two\",\"start\":2.33,\"end\":2.61}]"} {"utterance_id": "1255-74899-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.7, 1.01, "tailor"]], "response_mode": "offline", "answer": "[{\"word\":\"tailor\",\"start\":0.7,\"end\":1.01}]"} {"utterance_id": "6599-38591-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0003.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.83, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":0.49,\"end\":0.83}]"} {"utterance_id": "4323-18416-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.44, 1.73, "polly"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":1.44,\"end\":1.73}]"} {"utterance_id": "4572-64670-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.46, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":1.23,\"end\":1.46}]"} {"utterance_id": "116-288048-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0004.flac", "task_type": "lexical_detection", "query": "difficulties", "prompt": "After listening to the whole recording, output every instance of the word \"difficulties\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.68, 4.65, "difficulties"]], "response_mode": "offline", "answer": "[{\"word\":\"difficulties\",\"start\":3.68,\"end\":4.65}]"} {"utterance_id": "1585-131718-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0043.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.52, 2.95, "edison"], [6.75, 7.39, "remenyi"]], "response_mode": "offline", "answer": "[{\"word\":\"edison\",\"start\":2.52,\"end\":2.95},{\"word\":\"remenyi\",\"start\":6.75,\"end\":7.39}]"} {"utterance_id": "7601-175351-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0020.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.3, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":0.75,\"end\":1.3}]"} {"utterance_id": "8254-84205-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0063.flac", "task_type": "lexical_detection", "query": "opportunity", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"opportunity\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.68, "opportunity"]], "response_mode": "offline", "answer": "[{\"word\":\"opportunity\",\"start\":3.06,\"end\":3.68}]"} {"utterance_id": "3915-57461-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0019.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a state or province name with fields word, start, end (seconds). If none, reply [].", "labels": [[7.98, 8.78, "provinces"]], "response_mode": "offline", "answer": "[{\"word\":\"provinces\",\"start\":7.98,\"end\":8.78}]"} {"utterance_id": "8173-294714-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0004.flac", "task_type": "lexical_detection", "query": "bitterly", "prompt": "Process the full audio offline. Find every instance of the word \"bitterly\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.37, 6.89, "bitterly"]], "response_mode": "offline", "answer": "[{\"word\":\"bitterly\",\"start\":6.37,\"end\":6.89}]"} {"utterance_id": "6267-65525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.64, 6.12, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":5.64,\"end\":6.12}]"} {"utterance_id": "116-288046-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.72, 8.39, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":7.72,\"end\":8.39}]"} {"utterance_id": "1255-90413-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0007.flac", "task_type": "lexical_detection", "query": "thither", "prompt": "Process the full audio offline. Find every instance of the word \"thither\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.45, 5.75, "thither"]], "response_mode": "offline", "answer": "[{\"word\":\"thither\",\"start\":5.45,\"end\":5.75}]"} {"utterance_id": "1585-131718-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0048.flac", "task_type": "lexical_detection", "query": "fire", "prompt": "After listening to the whole recording, output every instance of the word \"fire\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.73, "fire"]], "response_mode": "offline", "answer": "[{\"word\":\"fire\",\"start\":1.36,\"end\":1.73}]"} {"utterance_id": "700-122868-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0002.flac", "task_type": "lexical_detection", "query": "mourned", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"mourned\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.24, 3.57, "mourned"]], "response_mode": "offline", "answer": "[{\"word\":\"mourned\",\"start\":3.24,\"end\":3.57}]"} {"utterance_id": "8173-294714-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0057.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.78, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.49,\"end\":0.78}]"} {"utterance_id": "3915-57461-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0004.flac", "task_type": "lexical_detection", "query": "boatmen", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"boatmen\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.55, 0.93, "boatmen"]], "response_mode": "offline", "answer": "[{\"word\":\"boatmen\",\"start\":0.55,\"end\":0.93}]"} {"utterance_id": "7697-105815-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0008.flac", "task_type": "lexical_detection", "query": "sin", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"sin\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.79, 4.06, "sin"]], "response_mode": "offline", "answer": "[{\"word\":\"sin\",\"start\":3.79,\"end\":4.06}]"} {"utterance_id": "1630-102884-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0012.flac", "task_type": "lexical_detection", "query": "system", "prompt": "Process the full audio offline. Find every instance of the word \"system\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.38, "system"]], "response_mode": "offline", "answer": "[{\"word\":\"system\",\"start\":3.82,\"end\":4.38}]"} {"utterance_id": "6455-66379-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0005.flac", "task_type": "lexical_counting", "query": "learn", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"learn\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.91, 1.25, "one"], [2.36, 2.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.91,\"end\":1.25},{\"word\":\"two\",\"start\":2.36,\"end\":2.6}]"} {"utterance_id": "1585-157660-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0014.flac", "task_type": "lexical_detection", "query": "clouds", "prompt": "After listening to the whole recording, output every instance of the word \"clouds\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.68, "clouds"]], "response_mode": "offline", "answer": "[{\"word\":\"clouds\",\"start\":2.16,\"end\":2.68}]"} {"utterance_id": "1701-141760-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0049.flac", "task_type": "lexical_detection", "query": "rising", "prompt": "After listening to the whole recording, output every instance of the word \"rising\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.85, "rising"]], "response_mode": "offline", "answer": "[{\"word\":\"rising\",\"start\":1.41,\"end\":1.85}]"} {"utterance_id": "6267-65525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0011.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.42, 0.87, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.42,\"end\":0.87}]"} {"utterance_id": "4831-25894-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0005.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.35, 3.64, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":3.35,\"end\":3.64}]"} {"utterance_id": "7601-101622-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0004.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.89, 9.2, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":8.89,\"end\":9.2}]"} {"utterance_id": "7697-105815-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.31, 0.88, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.31,\"end\":0.88}]"} {"utterance_id": "1630-102884-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0007.flac", "task_type": "lexical_detection", "query": "vision", "prompt": "Once the audio has finished, identify every instance of the word \"vision\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.49, 7.05, "vision"]], "response_mode": "offline", "answer": "[{\"word\":\"vision\",\"start\":6.49,\"end\":7.05}]"} {"utterance_id": "6455-67804-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0014.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After listening to the whole recording, output every instance of a country name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.37, "russia"]], "response_mode": "offline", "answer": "[{\"word\":\"russia\",\"start\":1.0,\"end\":1.37}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a body part with word/start/end. If none, reply [].", "labels": [[7.18, 7.56, "one"], [8.79, 9.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.18,\"end\":7.56},{\"word\":\"two\",\"start\":8.79,\"end\":9.13}]"} {"utterance_id": "5849-50873-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0011.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.74, "captain's"], [6.73, 7.28, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain's\",\"start\":3.21,\"end\":3.74},{\"word\":\"captain\",\"start\":6.73,\"end\":7.28}]"} {"utterance_id": "116-288047-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0019.flac", "task_type": "lexical_detection", "query": "god's", "prompt": "After the audio ends, extract all occurrences of the word \"god's\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.91, "god's"]], "response_mode": "offline", "answer": "[{\"word\":\"god's\",\"start\":0.62,\"end\":0.91}]"} {"utterance_id": "1701-141760-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0042.flac", "task_type": "lexical_detection", "query": "staff", "prompt": "Once the audio has finished, identify every instance of the word \"staff\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.64, 3.14, "staff"]], "response_mode": "offline", "answer": "[{\"word\":\"staff\",\"start\":2.64,\"end\":3.14}]"} {"utterance_id": "7697-105815-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0033.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.68, "breast"]], "response_mode": "offline", "answer": "[{\"word\":\"breast\",\"start\":1.34,\"end\":1.68}]"} {"utterance_id": "6841-88291-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0006.flac", "task_type": "lexical_detection", "query": "upright", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"upright\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.74, 1.15, "upright"]], "response_mode": "offline", "answer": "[{\"word\":\"upright\",\"start\":0.74,\"end\":1.15}]"} {"utterance_id": "3660-172182-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.58, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.34,\"end\":0.58}]"} {"utterance_id": "5849-50873-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0042.flac", "task_type": "lexical_detection", "query": "england", "prompt": "When the utterance is complete, report all instances of the word \"england\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.46, 4.91, "england"]], "response_mode": "offline", "answer": "[{\"word\":\"england\",\"start\":4.46,\"end\":4.91}]"} {"utterance_id": "6467-97061-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0010.flac", "task_type": "lexical_counting", "query": "elephant", "prompt": "After the audio ends, tally every instance of the word \"elephant\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[2.4, 2.86, "one"], [13.65, 14.08, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.4,\"end\":2.86},{\"word\":\"two\",\"start\":13.65,\"end\":14.08}]"} {"utterance_id": "8288-274162-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[5.79, 6.27, "legs"], [8.65, 9.35, "mustache"]], "response_mode": "offline", "answer": "[{\"word\":\"legs\",\"start\":5.79,\"end\":6.27},{\"word\":\"mustache\",\"start\":8.65,\"end\":9.35}]"} {"utterance_id": "6267-53049-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0011.flac", "task_type": "lexical_detection", "query": "heirloom", "prompt": "After the audio ends, extract all occurrences of the word \"heirloom\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.48, 10.1, "heirloom"]], "response_mode": "offline", "answer": "[{\"word\":\"heirloom\",\"start\":9.48,\"end\":10.1}]"} {"utterance_id": "1630-141772-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0016.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Process the full audio offline. Find every instance of a color name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.99, 8.51, "gold"]], "response_mode": "offline", "answer": "[{\"word\":\"gold\",\"start\":7.99,\"end\":8.51}]"} {"utterance_id": "3660-172183-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.38, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":0.83,\"end\":1.38}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "lexical_detection", "query": "flew", "prompt": "After listening to the whole recording, output every instance of the word \"flew\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.54, "flew"]], "response_mode": "offline", "answer": "[{\"word\":\"flew\",\"start\":2.24,\"end\":2.54}]"} {"utterance_id": "2506-11267-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0003.flac", "task_type": "lexical_detection", "query": "continent", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"continent\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.33, 4.94, "continent"]], "response_mode": "offline", "answer": "[{\"word\":\"continent\",\"start\":4.33,\"end\":4.94}]"} {"utterance_id": "3660-172182-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0012.flac", "task_type": "lexical_detection", "query": "geraint", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"geraint\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.51, 3.88, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":3.51,\"end\":3.88}]"} {"utterance_id": "4153-186222-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0034.flac", "task_type": "lexical_detection", "query": "refuses", "prompt": "After the audio ends, extract all occurrences of the word \"refuses\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.41, "refuses"]], "response_mode": "offline", "answer": "[{\"word\":\"refuses\",\"start\":2.93,\"end\":3.41}]"} {"utterance_id": "8173-294714-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0028.flac", "task_type": "lexical_detection", "query": "forgiven", "prompt": "After the audio ends, extract all occurrences of the word \"forgiven\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.05, 5.48, "forgiven"]], "response_mode": "offline", "answer": "[{\"word\":\"forgiven\",\"start\":5.05,\"end\":5.48}]"} {"utterance_id": "1630-102884-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0014.flac", "task_type": "lexical_counting", "query": "accident", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"accident\" (word, start, end). If none, reply [].", "labels": [[11.84, 12.48, "one"], [16.17, 16.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":11.84,\"end\":12.48},{\"word\":\"two\",\"start\":16.17,\"end\":16.76}]"} {"utterance_id": "8254-84205-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0008.flac", "task_type": "lexical_counting", "query": "canter", "prompt": "Process the full audio offline. For every occurrence of the word \"canter\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.16, "one"], [2.7, 3.43, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.78,\"end\":1.16},{\"word\":\"two\",\"start\":2.7,\"end\":3.43}]"} {"utterance_id": "1255-90413-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0001.flac", "task_type": "lexical_detection", "query": "remembered", "prompt": "When the utterance is complete, report all instances of the word \"remembered\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.7, 4.15, "remembered"]], "response_mode": "offline", "answer": "[{\"word\":\"remembered\",\"start\":3.7,\"end\":4.15}]"} {"utterance_id": "4153-185072-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.42, 6.02, "little beard's"]], "response_mode": "offline", "answer": "[{\"word\":\"little beard's\",\"start\":5.42,\"end\":6.02}]"} {"utterance_id": "8254-84205-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0015.flac", "task_type": "lexical_detection", "query": "indians", "prompt": "Once the audio has finished, identify every instance of the word \"indians\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.06, 1.56, "indians"]], "response_mode": "offline", "answer": "[{\"word\":\"indians\",\"start\":1.06,\"end\":1.56}]"} {"utterance_id": "1585-131718-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0053.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.84, 6.27, "edison's"]], "response_mode": "offline", "answer": "[{\"word\":\"edison's\",\"start\":5.84,\"end\":6.27}]"} {"utterance_id": "700-122867-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.7, 2.23, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":1.7,\"end\":2.23}]"} {"utterance_id": "4831-25894-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.58, "rose"]], "response_mode": "offline", "answer": "[{\"word\":\"rose\",\"start\":1.25,\"end\":1.58}]"} {"utterance_id": "4323-55228-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0031.flac", "task_type": "lexical_detection", "query": "finger", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"finger\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.36, "finger"]], "response_mode": "offline", "answer": "[{\"word\":\"finger\",\"start\":1.07,\"end\":1.36}]"} {"utterance_id": "4831-29134-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0000.flac", "task_type": "lexical_detection", "query": "superintendent", "prompt": "Once the audio has finished, identify every instance of the word \"superintendent\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.05, 4.77, "superintendent"]], "response_mode": "offline", "answer": "[{\"word\":\"superintendent\",\"start\":4.05,\"end\":4.77}]"} {"utterance_id": "4323-55228-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.66, 2.54, "pete miss billy"]], "response_mode": "offline", "answer": "[{\"word\":\"pete miss billy\",\"start\":0.66,\"end\":2.54}]"} {"utterance_id": "4323-18416-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0022.flac", "task_type": "lexical_detection", "query": "wouldn't", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wouldn't\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.76, "wouldn't"]], "response_mode": "offline", "answer": "[{\"word\":\"wouldn't\",\"start\":0.51,\"end\":0.76}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.91, 2.24, "tessa"], [7.56, 7.95, "tommo s"], [11.81, 12.28, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":1.91,\"end\":2.24},{\"word\":\"tommo s\",\"start\":7.56,\"end\":7.95},{\"word\":\"tessa\",\"start\":11.81,\"end\":12.28}]"} {"utterance_id": "4570-102353-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0007.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen carefully to the entire clip, then report every instance of an organization name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.2, 11.56, "whig"], [14.65, 15.36, "democrat"]], "response_mode": "offline", "answer": "[{\"word\":\"whig\",\"start\":11.2,\"end\":11.56},{\"word\":\"democrat\",\"start\":14.65,\"end\":15.36}]"} {"utterance_id": "6455-66379-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.3, 1.82, "thompson"]], "response_mode": "offline", "answer": "[{\"word\":\"thompson\",\"start\":1.3,\"end\":1.82}]"} {"utterance_id": "6267-65525-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0042.flac", "task_type": "lexical_detection", "query": "blankly", "prompt": "Once the audio has finished, identify every instance of the word \"blankly\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.51, 3.18, "blankly"]], "response_mode": "offline", "answer": "[{\"word\":\"blankly\",\"start\":2.51,\"end\":3.18}]"} {"utterance_id": "1686-142278-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.93, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.51,\"end\":0.93}]"} {"utterance_id": "4323-13259-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0011.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Do not answer until the audio ends. Then list each instance of a date as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.02, 1.34, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":1.02,\"end\":1.34}]"} {"utterance_id": "1630-96099-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0010.flac", "task_type": "lexical_detection", "query": "played", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"played\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.25, 3.66, "played"]], "response_mode": "offline", "answer": "[{\"word\":\"played\",\"start\":3.25,\"end\":3.66}]"} {"utterance_id": "6455-67803-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.47, 0.84, "lucia"], [4.07, 4.96, "lady theobald"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia\",\"start\":0.47,\"end\":0.84},{\"word\":\"lady theobald\",\"start\":4.07,\"end\":4.96}]"} {"utterance_id": "1585-131718-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0031.flac", "task_type": "semantic_detection", "query": "date", "prompt": "When the utterance is complete, report all instances of a date including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.5, "eighteen eighty one"]], "response_mode": "offline", "answer": "[{\"word\":\"eighteen eighty one\",\"start\":1.55,\"end\":2.5}]"} {"utterance_id": "8254-115543-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0026.flac", "task_type": "lexical_detection", "query": "ride", "prompt": "After listening to the whole recording, output every instance of the word \"ride\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.35, 1.53, "ride"]], "response_mode": "offline", "answer": "[{\"word\":\"ride\",\"start\":1.35,\"end\":1.53}]"} {"utterance_id": "2506-169427-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.44, 2.69, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":2.44,\"end\":2.69}]"} {"utterance_id": "6467-97061-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0010.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Once the audio has finished, identify every instance of an organization name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.53, 7.96, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"army\",\"start\":7.53,\"end\":7.96}]"} {"utterance_id": "6467-94831-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0013.flac", "task_type": "lexical_detection", "query": "tongue", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"tongue\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.11, 4.48, "tongue"]], "response_mode": "offline", "answer": "[{\"word\":\"tongue\",\"start\":4.11,\"end\":4.48}]"} {"utterance_id": "5849-50963-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0006.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.44, 3.71, "storm"]], "response_mode": "offline", "answer": "[{\"word\":\"storm\",\"start\":3.44,\"end\":3.71}]"} {"utterance_id": "6841-88291-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0004.flac", "task_type": "lexical_detection", "query": "short", "prompt": "Once the audio has finished, identify every instance of the word \"short\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.09, 1.39, "short"]], "response_mode": "offline", "answer": "[{\"word\":\"short\",\"start\":1.09,\"end\":1.39}]"} {"utterance_id": "1701-141760-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[13.62, 14.05, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":13.62,\"end\":14.05}]"} {"utterance_id": "4323-18416-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0013.flac", "task_type": "lexical_detection", "query": "presently", "prompt": "When the utterance is complete, report all instances of the word \"presently\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.41, "presently"]], "response_mode": "offline", "answer": "[{\"word\":\"presently\",\"start\":0.93,\"end\":1.41}]"} {"utterance_id": "6267-65525-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0044.flac", "task_type": "lexical_detection", "query": "thing", "prompt": "After listening to the whole recording, output every instance of the word \"thing\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.14, 4.58, "thing"]], "response_mode": "offline", "answer": "[{\"word\":\"thing\",\"start\":4.14,\"end\":4.58}]"} {"utterance_id": "116-288045-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.78, "eyes"], [2.81, 3.2, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":1.4,\"end\":1.78},{\"word\":\"voice\",\"start\":2.81,\"end\":3.2}]"} {"utterance_id": "6123-59186-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0002.flac", "task_type": "lexical_counting", "query": "sufficiently", "prompt": "Once the audio has finished, list each occurrence of the word \"sufficiently\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.39, 3.09, "one"], [4.45, 5.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.39,\"end\":3.09},{\"word\":\"two\",\"start\":4.45,\"end\":5.02}]"} {"utterance_id": "3660-172182-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.67, 7.79, "modred arthur's"]], "response_mode": "offline", "answer": "[{\"word\":\"modred arthur's\",\"start\":6.67,\"end\":7.79}]"} {"utterance_id": "6841-88291-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0046.flac", "task_type": "lexical_detection", "query": "attended", "prompt": "After the audio ends, extract all occurrences of the word \"attended\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.99, "attended"]], "response_mode": "offline", "answer": "[{\"word\":\"attended\",\"start\":0.64,\"end\":0.99}]"} {"utterance_id": "4323-55228-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0004.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.52, 0.91, "one"], [5.53, 5.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.52,\"end\":0.91},{\"word\":\"two\",\"start\":5.53,\"end\":5.94}]"} {"utterance_id": "1585-131718-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0022.flac", "task_type": "lexical_detection", "query": "balance", "prompt": "After listening to the whole recording, output every instance of the word \"balance\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.13, 5.54, "balance"]], "response_mode": "offline", "answer": "[{\"word\":\"balance\",\"start\":5.13,\"end\":5.54}]"} {"utterance_id": "3660-172183-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0000.flac", "task_type": "lexical_detection", "query": "court", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"court\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.3, 3.66, "court"]], "response_mode": "offline", "answer": "[{\"word\":\"court\",\"start\":3.3,\"end\":3.66}]"} {"utterance_id": "4831-25894-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0005.flac", "task_type": "lexical_detection", "query": "tommo", "prompt": "After the audio ends, extract all occurrences of the word \"tommo\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.24, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":0.87,\"end\":1.24}]"} {"utterance_id": "4831-25894-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0016.flac", "task_type": "lexical_detection", "query": "mittens", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"mittens\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.78, 3.28, "mittens"]], "response_mode": "offline", "answer": "[{\"word\":\"mittens\",\"start\":2.78,\"end\":3.28}]"} {"utterance_id": "4831-18525-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0031.flac", "task_type": "lexical_detection", "query": "that's", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"that's\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.1, 4.38, "that's"]], "response_mode": "offline", "answer": "[{\"word\":\"that's\",\"start\":4.1,\"end\":4.38}]"} {"utterance_id": "1701-141759-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0016.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.34, "one"], [4.29, 4.65, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.9,\"end\":1.34},{\"word\":\"two\",\"start\":4.29,\"end\":4.65}]"} {"utterance_id": "700-122867-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0026.flac", "task_type": "lexical_detection", "query": "weep", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"weep\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.01, 1.39, "weep"]], "response_mode": "offline", "answer": "[{\"word\":\"weep\",\"start\":1.01,\"end\":1.39}]"} {"utterance_id": "4831-18525-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.88, "mister king"], [2.11, 2.66, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"mister king\",\"start\":0.33,\"end\":0.88},{\"word\":\"phronsie\",\"start\":2.11,\"end\":2.66}]"} {"utterance_id": "8254-115543-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0036.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.87, 3.03, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.87,\"end\":3.03}]"} {"utterance_id": "7641-96670-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0026.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[1.76, 2.09, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":1.76,\"end\":2.09}]"} {"utterance_id": "8288-274162-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0051.flac", "task_type": "lexical_detection", "query": "painter", "prompt": "Once the audio has finished, identify every instance of the word \"painter\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.74, 1.12, "painter"]], "response_mode": "offline", "answer": "[{\"word\":\"painter\",\"start\":0.74,\"end\":1.12}]"} {"utterance_id": "1650-157641-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0009.flac", "task_type": "lexical_detection", "query": "jealous", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"jealous\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.76, "jealous"]], "response_mode": "offline", "answer": "[{\"word\":\"jealous\",\"start\":2.25,\"end\":2.76}]"} {"utterance_id": "5849-50873-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0027.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.12, 4.43, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":4.12,\"end\":4.43}]"} {"utterance_id": "7641-96252-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.31, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.01,\"end\":2.31}]"} {"utterance_id": "1585-131718-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0018.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.31, 4.7, "new york"]], "response_mode": "offline", "answer": "[{\"word\":\"new york\",\"start\":4.31,\"end\":4.7}]"} {"utterance_id": "6267-65525-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0022.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.55, "plainfield"]], "response_mode": "offline", "answer": "[{\"word\":\"plainfield\",\"start\":1.97,\"end\":2.55}]"} {"utterance_id": "1650-167613-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0031.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.23, 2.99, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":2.23,\"end\":2.99}]"} {"utterance_id": "6123-59150-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[6.39, 6.57, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":6.39,\"end\":6.57}]"} {"utterance_id": "5849-50962-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 1.81, "bert"]], "response_mode": "offline", "answer": "[{\"word\":\"bert\",\"start\":1.49,\"end\":1.81}]"} {"utterance_id": "6467-94831-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0036.flac", "task_type": "lexical_detection", "query": "boys", "prompt": "After listening to the whole recording, output every instance of the word \"boys\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.36, 0.74, "boys"]], "response_mode": "offline", "answer": "[{\"word\":\"boys\",\"start\":0.36,\"end\":0.74}]"} {"utterance_id": "700-122866-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0018.flac", "task_type": "semantic_detection", "query": "date", "prompt": "You will hear a complete utterance. When it ends, list every instance of a date you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.45, "monday"]], "response_mode": "offline", "answer": "[{\"word\":\"monday\",\"start\":1.14,\"end\":1.45}]"} {"utterance_id": "1255-90413-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0002.flac", "task_type": "lexical_detection", "query": "eventually", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"eventually\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.4, 4.87, "eventually"]], "response_mode": "offline", "answer": "[{\"word\":\"eventually\",\"start\":4.4,\"end\":4.87}]"} {"utterance_id": "4572-112375-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.78, 4.31, "steps"]], "response_mode": "offline", "answer": "[{\"word\":\"steps\",\"start\":3.78,\"end\":4.31}]"} {"utterance_id": "4323-18416-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 1.97, "pickering"]], "response_mode": "offline", "answer": "[{\"word\":\"pickering\",\"start\":1.53,\"end\":1.97}]"} {"utterance_id": "6267-65525-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0024.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.33, 4.93, "novelist"]], "response_mode": "offline", "answer": "[{\"word\":\"novelist\",\"start\":4.33,\"end\":4.93}]"} {"utterance_id": "6267-53049-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0017.flac", "task_type": "lexical_detection", "query": "coming", "prompt": "Process the full audio offline. Find every instance of the word \"coming\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.3, "coming"]], "response_mode": "offline", "answer": "[{\"word\":\"coming\",\"start\":2.81,\"end\":3.3}]"} {"utterance_id": "1255-138279-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 2.21, "four five"]], "response_mode": "offline", "answer": "[{\"word\":\"four five\",\"start\":0.37,\"end\":2.21}]"} {"utterance_id": "3660-172182-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0028.flac", "task_type": "lexical_counting", "query": "greet", "prompt": "When the utterance is complete, report how the word \"greet\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 2.69, "one"], [5.27, 5.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.46,\"end\":2.69},{\"word\":\"two\",\"start\":5.27,\"end\":5.46}]"} {"utterance_id": "8254-84205-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0002.flac", "task_type": "lexical_detection", "query": "injured", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"injured\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.4, 2.72, "injured"]], "response_mode": "offline", "answer": "[{\"word\":\"injured\",\"start\":2.4,\"end\":2.72}]"} {"utterance_id": "3663-172528-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0050.flac", "task_type": "lexical_detection", "query": "design", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"design\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 2.17, "design"]], "response_mode": "offline", "answer": "[{\"word\":\"design\",\"start\":1.48,\"end\":2.17}]"} {"utterance_id": "3660-172182-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0039.flac", "task_type": "lexical_detection", "query": "homage", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"homage\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.54, "homage"]], "response_mode": "offline", "answer": "[{\"word\":\"homage\",\"start\":1.23,\"end\":1.54}]"} {"utterance_id": "1701-141759-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0019.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[1.47, 2.06, "officer"]], "response_mode": "offline", "answer": "[{\"word\":\"officer\",\"start\":1.47,\"end\":2.06}]"} {"utterance_id": "1255-138279-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0021.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.0,\"end\":1.18}]"} {"utterance_id": "1650-157641-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0002.flac", "task_type": "lexical_detection", "query": "please", "prompt": "Once the audio has finished, identify every instance of the word \"please\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.27, 2.91, "please"]], "response_mode": "offline", "answer": "[{\"word\":\"please\",\"start\":2.27,\"end\":2.91}]"} {"utterance_id": "6267-53049-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0001.flac", "task_type": "lexical_detection", "query": "older", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"older\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.24, 7.62, "older"]], "response_mode": "offline", "answer": "[{\"word\":\"older\",\"start\":7.24,\"end\":7.62}]"} {"utterance_id": "7641-96670-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0021.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.98, 4.34, "teeth"]], "response_mode": "offline", "answer": "[{\"word\":\"teeth\",\"start\":3.98,\"end\":4.34}]"} {"utterance_id": "6841-88294-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0067.flac", "task_type": "lexical_detection", "query": "intention", "prompt": "After listening to the whole recording, output every instance of the word \"intention\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.84, 4.25, "intention"]], "response_mode": "offline", "answer": "[{\"word\":\"intention\",\"start\":3.84,\"end\":4.25}]"} {"utterance_id": "4153-185072-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0009.flac", "task_type": "lexical_detection", "query": "indian", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"indian\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.96, 1.32, "indian"]], "response_mode": "offline", "answer": "[{\"word\":\"indian\",\"start\":0.96,\"end\":1.32}]"} {"utterance_id": "7641-96684-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0022.flac", "task_type": "lexical_detection", "query": "soon", "prompt": "Once the audio has finished, identify every instance of the word \"soon\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.99, 1.35, "soon"]], "response_mode": "offline", "answer": "[{\"word\":\"soon\",\"start\":0.99,\"end\":1.35}]"} {"utterance_id": "1630-73710-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.72, "messenger"]], "response_mode": "offline", "answer": "[{\"word\":\"messenger\",\"start\":1.14,\"end\":1.72}]"} {"utterance_id": "1255-74899-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0007.flac", "task_type": "lexical_counting", "query": "put", "prompt": "Listen to the full audio. Count every occurrence of the word \"put\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.11, "one"], [3.43, 3.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.93,\"end\":1.11},{\"word\":\"two\",\"start\":3.43,\"end\":3.6}]"} {"utterance_id": "4515-11057-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0048.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.63, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":3.21,\"end\":3.63}]"} {"utterance_id": "1650-157641-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0001.flac", "task_type": "lexical_detection", "query": "quilp", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"quilp\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.22, 1.65, "quilp"]], "response_mode": "offline", "answer": "[{\"word\":\"quilp\",\"start\":1.22,\"end\":1.65}]"} {"utterance_id": "8254-84205-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0028.flac", "task_type": "lexical_detection", "query": "doctor", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"doctor\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.26, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":0.93,\"end\":1.26}]"} {"utterance_id": "116-288046-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0004.flac", "task_type": "lexical_detection", "query": "tears", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"tears\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.02, 1.35, "tears"]], "response_mode": "offline", "answer": "[{\"word\":\"tears\",\"start\":1.02,\"end\":1.35}]"} {"utterance_id": "6455-67803-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0005.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.67, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":0.23,\"end\":0.67}]"} {"utterance_id": "5543-27761-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0070.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.75, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.44,\"end\":0.75}]"} {"utterance_id": "8254-115543-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.38, 5.76, "chola"]], "response_mode": "offline", "answer": "[{\"word\":\"chola\",\"start\":5.38,\"end\":5.76}]"} {"utterance_id": "8254-84205-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0056.flac", "task_type": "lexical_detection", "query": "chris's", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"chris's\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.78, "chris's"]], "response_mode": "offline", "answer": "[{\"word\":\"chris's\",\"start\":1.39,\"end\":1.78}]"} {"utterance_id": "8254-84205-0071", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0071.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.87, "scalps"]], "response_mode": "offline", "answer": "[{\"word\":\"scalps\",\"start\":1.41,\"end\":1.87}]"} {"utterance_id": "4831-25894-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0013.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.39, 0.62, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.39,\"end\":0.62}]"} {"utterance_id": "6841-88291-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0056.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.46, 4.1, "cowboy's"], [6.26, 6.87, "oarsman's"]], "response_mode": "offline", "answer": "[{\"word\":\"cowboy's\",\"start\":3.46,\"end\":4.1},{\"word\":\"oarsman's\",\"start\":6.26,\"end\":6.87}]"} {"utterance_id": "4831-25894-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0001.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Do not answer until the audio ends. Then list each instance of a date as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.75, 3.37, "christmas"]], "response_mode": "offline", "answer": "[{\"word\":\"christmas\",\"start\":2.75,\"end\":3.37}]"} {"utterance_id": "4570-24733-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.81, 10.31, "god's"]], "response_mode": "offline", "answer": "[{\"word\":\"god's\",\"start\":9.81,\"end\":10.31}]"} {"utterance_id": "1686-142278-0081", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0081.flac", "task_type": "lexical_detection", "query": "helstone", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"helstone\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.94, 10.46, "helstone"]], "response_mode": "offline", "answer": "[{\"word\":\"helstone\",\"start\":9.94,\"end\":10.46}]"} {"utterance_id": "3915-57461-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0030.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Listen to the full audio. After it finishes, find every instance of a weather term. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.21, "wind"]], "response_mode": "offline", "answer": "[{\"word\":\"wind\",\"start\":4.82,\"end\":5.21}]"} {"utterance_id": "4515-11057-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0005.flac", "task_type": "lexical_detection", "query": "marian", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"marian\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.84, 5.41, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":4.84,\"end\":5.41}]"} {"utterance_id": "4153-186222-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0035.flac", "task_type": "lexical_detection", "query": "bitterness", "prompt": "After listening to the whole recording, output every instance of the word \"bitterness\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.98, 7.55, "bitterness"]], "response_mode": "offline", "answer": "[{\"word\":\"bitterness\",\"start\":6.98,\"end\":7.55}]"} {"utterance_id": "4515-11057-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0047.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":0.81,\"end\":1.03}]"} {"utterance_id": "700-122866-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.8, "alice bell's"], [3.09, 3.27, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"alice bell's\",\"start\":1.08,\"end\":1.8},{\"word\":\"anne\",\"start\":3.09,\"end\":3.27}]"} {"utterance_id": "5849-50873-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0012.flac", "task_type": "lexical_detection", "query": "bowed", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"bowed\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.61, "bowed"]], "response_mode": "offline", "answer": "[{\"word\":\"bowed\",\"start\":1.22,\"end\":1.61}]"} {"utterance_id": "4515-11057-0092", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0092.flac", "task_type": "lexical_counting", "query": "believe", "prompt": "Once the audio has finished, list each occurrence of the word \"believe\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.73, 1.19, "one"], [2.32, 2.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.73,\"end\":1.19},{\"word\":\"two\",\"start\":2.32,\"end\":2.76}]"} {"utterance_id": "1255-90407-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0005.flac", "task_type": "lexical_detection", "query": "think", "prompt": "Once the audio has finished, identify every instance of the word \"think\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.81, 1.07, "think"]], "response_mode": "offline", "answer": "[{\"word\":\"think\",\"start\":0.81,\"end\":1.07}]"} {"utterance_id": "4153-61735-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.62, "daughter"], [2.76, 3.42, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"daughter\",\"start\":1.16,\"end\":1.62},{\"word\":\"daughter\",\"start\":2.76,\"end\":3.42}]"} {"utterance_id": "8254-84205-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0060.flac", "task_type": "lexical_detection", "query": "packed", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"packed\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.43, "packed"]], "response_mode": "offline", "answer": "[{\"word\":\"packed\",\"start\":1.14,\"end\":1.43}]"} {"utterance_id": "5849-50962-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0002.flac", "task_type": "lexical_detection", "query": "cord", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"cord\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.59, 0.98, "cord"]], "response_mode": "offline", "answer": "[{\"word\":\"cord\",\"start\":0.59,\"end\":0.98}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_counting", "query": "clock_time", "prompt": "Listen to the full audio. Count every occurrence of a time of day. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.14, 12.5, "one"], [15.01, 15.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":12.14,\"end\":12.5},{\"word\":\"two\",\"start\":15.01,\"end\":15.42}]"} {"utterance_id": "7641-96670-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0012.flac", "task_type": "lexical_detection", "query": "subdue", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"subdue\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.83, 2.2, "subdue"]], "response_mode": "offline", "answer": "[{\"word\":\"subdue\",\"start\":1.83,\"end\":2.2}]"} {"utterance_id": "1686-142278-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0058.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.54, "farmer"]], "response_mode": "offline", "answer": "[{\"word\":\"farmer\",\"start\":2.22,\"end\":2.54}]"} {"utterance_id": "8288-274162-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0014.flac", "task_type": "lexical_detection", "query": "capable", "prompt": "After listening to the whole recording, output every instance of the word \"capable\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.58, 5.25, "capable"]], "response_mode": "offline", "answer": "[{\"word\":\"capable\",\"start\":4.58,\"end\":5.25}]"} {"utterance_id": "4831-25894-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.24, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":0.87,\"end\":1.24}]"} {"utterance_id": "4153-186223-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0020.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.15, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":1.8,\"end\":2.15}]"} {"utterance_id": "6599-38590-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.0, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":0.72,\"end\":1.0}]"} {"utterance_id": "3663-172528-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0044.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.88, 4.39, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":3.88,\"end\":4.39}]"} {"utterance_id": "6841-88294-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0012.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Once the audio has finished, identify every instance of a state or province name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.39, 3.98, "colorado"]], "response_mode": "offline", "answer": "[{\"word\":\"colorado\",\"start\":3.39,\"end\":3.98}]"} {"utterance_id": "6267-65525-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0017.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.54, 0.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":0.54,\"end\":0.72}]"} {"utterance_id": "6841-88291-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0011.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.22, 0.65, "one"], [2.01, 2.45, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.22,\"end\":0.65},{\"word\":\"two\",\"start\":2.01,\"end\":2.45}]"} {"utterance_id": "1585-157660-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0013.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.69, 4.94, "couple"]], "response_mode": "offline", "answer": "[{\"word\":\"couple\",\"start\":4.69,\"end\":4.94}]"} {"utterance_id": "4153-186222-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[3.27, 3.51, "beth"]], "response_mode": "offline", "answer": "[{\"word\":\"beth\",\"start\":3.27,\"end\":3.51}]"} {"utterance_id": "4570-56594-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0017.flac", "task_type": "lexical_detection", "query": "salt", "prompt": "When the utterance is complete, report all instances of the word \"salt\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.36, "salt"]], "response_mode": "offline", "answer": "[{\"word\":\"salt\",\"start\":1.92,\"end\":2.36}]"} {"utterance_id": "4572-112375-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0007.flac", "task_type": "lexical_detection", "query": "nor", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"nor\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.35, 3.62, "nor"]], "response_mode": "offline", "answer": "[{\"word\":\"nor\",\"start\":3.35,\"end\":3.62}]"} {"utterance_id": "7641-96684-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0017.flac", "task_type": "lexical_detection", "query": "lamented", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"lamented\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.31, 6.95, "lamented"], [12.45, 12.94, "lamented"]], "response_mode": "offline", "answer": "[{\"word\":\"lamented\",\"start\":6.31,\"end\":6.95},{\"word\":\"lamented\",\"start\":12.45,\"end\":12.94}]"} {"utterance_id": "1686-142278-0094", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0094.flac", "task_type": "lexical_detection", "query": "moments", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"moments\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.69, 4.08, "moments"]], "response_mode": "offline", "answer": "[{\"word\":\"moments\",\"start\":3.69,\"end\":4.08}]"} {"utterance_id": "1630-96099-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0009.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.38, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.98,\"end\":1.38}]"} {"utterance_id": "116-288045-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0005.flac", "task_type": "lexical_detection", "query": "worship", "prompt": "Once the audio has finished, identify every instance of the word \"worship\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.79, 3.26, "worship"]], "response_mode": "offline", "answer": "[{\"word\":\"worship\",\"start\":2.79,\"end\":3.26}]"} {"utterance_id": "8288-274162-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.31, 0.88, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":0.31,\"end\":0.88}]"} {"utterance_id": "116-288047-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0015.flac", "task_type": "lexical_detection", "query": "mankind", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"mankind\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.02, 3.68, "mankind"]], "response_mode": "offline", "answer": "[{\"word\":\"mankind\",\"start\":3.02,\"end\":3.68}]"} {"utterance_id": "7601-175351-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0027.flac", "task_type": "lexical_detection", "query": "mountain", "prompt": "After the audio ends, extract all occurrences of the word \"mountain\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 4.19, "mountain"]], "response_mode": "offline", "answer": "[{\"word\":\"mountain\",\"start\":3.6,\"end\":4.19}]"} {"utterance_id": "7641-96684-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0010.flac", "task_type": "lexical_detection", "query": "underworld", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"underworld\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.87, "underworld"], [7.83, 8.33, "underworld"]], "response_mode": "offline", "answer": "[{\"word\":\"underworld\",\"start\":2.24,\"end\":2.87},{\"word\":\"underworld\",\"start\":7.83,\"end\":8.33}]"} {"utterance_id": "5849-50873-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0001.flac", "task_type": "lexical_detection", "query": "andros", "prompt": "Process the full audio offline. Find every instance of the word \"andros\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.11, 7.61, "andros"]], "response_mode": "offline", "answer": "[{\"word\":\"andros\",\"start\":7.11,\"end\":7.61}]"} {"utterance_id": "4323-55228-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0044.flac", "task_type": "lexical_detection", "query": "sweetheart", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sweetheart\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 1.02, "sweetheart"]], "response_mode": "offline", "answer": "[{\"word\":\"sweetheart\",\"start\":0.39,\"end\":1.02}]"} {"utterance_id": "4153-61735-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0002.flac", "task_type": "lexical_counting", "query": "daughter", "prompt": "Once the audio has finished, list each occurrence of the word \"daughter\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.16, 1.62, "one"], [2.76, 3.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.16,\"end\":1.62},{\"word\":\"two\",\"start\":2.76,\"end\":3.42}]"} {"utterance_id": "7697-105815-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0040.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a time of day with fields word, start, end (seconds). If none, reply [].", "labels": [[3.91, 4.34, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":3.91,\"end\":4.34}]"} {"utterance_id": "4515-11057-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0058.flac", "task_type": "lexical_detection", "query": "can't", "prompt": "After listening to the whole recording, output every instance of the word \"can't\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.16, "can't"]], "response_mode": "offline", "answer": "[{\"word\":\"can't\",\"start\":1.83,\"end\":2.16}]"} {"utterance_id": "1701-141760-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0040.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.41, 1.08, "one"], [7.16, 7.67, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.41,\"end\":1.08},{\"word\":\"two\",\"start\":7.16,\"end\":7.67}]"} {"utterance_id": "116-288048-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0025.flac", "task_type": "semantic_counting", "query": "disease", "prompt": "After listening to the whole recording, output timed count words for a disease name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[12.35, 13.14, "one"], [15.2, 15.81, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":12.35,\"end\":13.14},{\"word\":\"two\",\"start\":15.2,\"end\":15.81}]"} {"utterance_id": "4153-186223-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0017.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.75, 6.92, "three"], [7.94, 8.39, "thirty"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":6.75,\"end\":6.92},{\"word\":\"thirty\",\"start\":7.94,\"end\":8.39}]"} {"utterance_id": "4323-55228-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0035.flac", "task_type": "lexical_detection", "query": "announced", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"announced\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 2.89, "announced"]], "response_mode": "offline", "answer": "[{\"word\":\"announced\",\"start\":2.57,\"end\":2.89}]"} {"utterance_id": "1701-141760-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0018.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.47, 4.84, "wine"]], "response_mode": "offline", "answer": "[{\"word\":\"wine\",\"start\":4.47,\"end\":4.84}]"} {"utterance_id": "3660-6517-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0012.flac", "task_type": "lexical_detection", "query": "forage", "prompt": "Once the audio has finished, identify every instance of the word \"forage\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.52, 1.11, "forage"]], "response_mode": "offline", "answer": "[{\"word\":\"forage\",\"start\":0.52,\"end\":1.11}]"} {"utterance_id": "7601-101619-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.68, 14.2, "tail"]], "response_mode": "offline", "answer": "[{\"word\":\"tail\",\"start\":13.68,\"end\":14.2}]"} {"utterance_id": "1650-173552-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0003.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Once the audio has finished, identify every instance of a weather term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.3, 8.13, "sunshine"]], "response_mode": "offline", "answer": "[{\"word\":\"sunshine\",\"start\":7.3,\"end\":8.13}]"} {"utterance_id": "8288-274162-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0046.flac", "task_type": "lexical_detection", "query": "carpenter", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"carpenter\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.6, 1.29, "carpenter"]], "response_mode": "offline", "answer": "[{\"word\":\"carpenter\",\"start\":0.6,\"end\":1.29}]"} {"utterance_id": "5849-50873-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0031.flac", "task_type": "lexical_detection", "query": "connecticut", "prompt": "After the audio ends, extract all occurrences of the word \"connecticut\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.45, "connecticut"]], "response_mode": "offline", "answer": "[{\"word\":\"connecticut\",\"start\":3.94,\"end\":4.45}]"} {"utterance_id": "7697-245715-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0002.flac", "task_type": "lexical_detection", "query": "imperfect", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"imperfect\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.55, "imperfect"]], "response_mode": "offline", "answer": "[{\"word\":\"imperfect\",\"start\":2.77,\"end\":3.55}]"} {"utterance_id": "4831-25894-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0034.flac", "task_type": "lexical_detection", "query": "crowd", "prompt": "Once the audio has finished, identify every instance of the word \"crowd\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.92, 1.2, "crowd"]], "response_mode": "offline", "answer": "[{\"word\":\"crowd\",\"start\":0.92,\"end\":1.2}]"} {"utterance_id": "8254-84205-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.22, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":1.95,\"end\":2.22}]"} {"utterance_id": "700-122866-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0020.flac", "task_type": "lexical_detection", "query": "nonsense", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"nonsense\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.85, 3.6, "nonsense"]], "response_mode": "offline", "answer": "[{\"word\":\"nonsense\",\"start\":2.85,\"end\":3.6}]"} {"utterance_id": "1650-167613-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0011.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Process the full audio offline. Find every instance of a city name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 1.42, "smolensk"]], "response_mode": "offline", "answer": "[{\"word\":\"smolensk\",\"start\":0.53,\"end\":1.42}]"} {"utterance_id": "3663-172528-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0036.flac", "task_type": "lexical_detection", "query": "halted", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"halted\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.87, 10.39, "halted"]], "response_mode": "offline", "answer": "[{\"word\":\"halted\",\"start\":9.87,\"end\":10.39}]"} {"utterance_id": "8288-274162-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0045.flac", "task_type": "lexical_detection", "query": "fancies", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"fancies\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.91, 1.47, "fancies"]], "response_mode": "offline", "answer": "[{\"word\":\"fancies\",\"start\":0.91,\"end\":1.47}]"} {"utterance_id": "4572-112375-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0004.flac", "task_type": "lexical_detection", "query": "trod", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"trod\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.27, 6.01, "trod"]], "response_mode": "offline", "answer": "[{\"word\":\"trod\",\"start\":5.27,\"end\":6.01}]"} {"utterance_id": "3663-172528-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0026.flac", "task_type": "lexical_detection", "query": "meanwhile", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"meanwhile\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.8, "meanwhile"]], "response_mode": "offline", "answer": "[{\"word\":\"meanwhile\",\"start\":0.23,\"end\":0.8}]"} {"utterance_id": "6841-88291-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0054.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.23, "hip"]], "response_mode": "offline", "answer": "[{\"word\":\"hip\",\"start\":2.03,\"end\":2.23}]"} {"utterance_id": "1585-131718-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0005.flac", "task_type": "lexical_detection", "query": "travel", "prompt": "Once the audio has finished, identify every instance of the word \"travel\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.12, 2.48, "travel"]], "response_mode": "offline", "answer": "[{\"word\":\"travel\",\"start\":2.12,\"end\":2.48}]"} {"utterance_id": "7697-105815-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.79, "christian sir"]], "response_mode": "offline", "answer": "[{\"word\":\"christian sir\",\"start\":0.4,\"end\":1.79}]"} {"utterance_id": "5849-50873-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0034.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.12, 2.81, "heads"]], "response_mode": "offline", "answer": "[{\"word\":\"heads\",\"start\":2.12,\"end\":2.81}]"} {"utterance_id": "4515-11057-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0070.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.89, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":1.4,\"end\":1.89}]"} {"utterance_id": "1255-90407-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0002.flac", "task_type": "lexical_detection", "query": "why", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"why\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.79, "why"]], "response_mode": "offline", "answer": "[{\"word\":\"why\",\"start\":0.51,\"end\":0.79}]"} {"utterance_id": "2506-13150-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0007.flac", "task_type": "lexical_detection", "query": "heaven's", "prompt": "After the audio ends, extract all occurrences of the word \"heaven's\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.6, 3.01, "heaven's"]], "response_mode": "offline", "answer": "[{\"word\":\"heaven's\",\"start\":2.6,\"end\":3.01}]"} {"utterance_id": "6841-88294-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0068.flac", "task_type": "semantic_counting", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of count events for an ordinal number (word, start, end). If none, reply [].", "labels": [[0.36, 0.57, "one"], [2.32, 2.58, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.36,\"end\":0.57},{\"word\":\"two\",\"start\":2.32,\"end\":2.58}]"} {"utterance_id": "4153-186222-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0025.flac", "task_type": "lexical_detection", "query": "suspicions", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"suspicions\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.72, 5.33, "suspicions"]], "response_mode": "offline", "answer": "[{\"word\":\"suspicions\",\"start\":4.72,\"end\":5.33}]"} {"utterance_id": "7697-105817-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.28, 0.71, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.28,\"end\":0.71}]"} {"utterance_id": "8254-115543-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.62, "back"]], "response_mode": "offline", "answer": "[{\"word\":\"back\",\"start\":1.25,\"end\":1.62}]"} {"utterance_id": "1630-73710-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0002.flac", "task_type": "lexical_detection", "query": "imagination", "prompt": "After listening to the whole recording, output every instance of the word \"imagination\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.65, 5.64, "imagination"]], "response_mode": "offline", "answer": "[{\"word\":\"imagination\",\"start\":4.65,\"end\":5.64}]"} {"utterance_id": "1255-90413-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0028.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.99, 2.13, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.99,\"end\":2.13}]"} {"utterance_id": "4515-11057-0089", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0089.flac", "task_type": "lexical_detection", "query": "fate", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fate\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 3.01, "fate"]], "response_mode": "offline", "answer": "[{\"word\":\"fate\",\"start\":2.35,\"end\":3.01}]"} {"utterance_id": "4515-11057-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0063.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a family relation with word/start/end. If none, reply [].", "labels": [[3.78, 4.2, "one"], [6.43, 6.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.78,\"end\":4.2},{\"word\":\"two\",\"start\":6.43,\"end\":6.94}]"} {"utterance_id": "6267-53049-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.69, 1.37, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.69,\"end\":1.37}]"} {"utterance_id": "3663-172528-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0035.flac", "task_type": "lexical_detection", "query": "accused", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"accused\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.87, 5.51, "accused"]], "response_mode": "offline", "answer": "[{\"word\":\"accused\",\"start\":4.87,\"end\":5.51}]"} {"utterance_id": "6841-88291-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0029.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.87, 4.3, "cattle"]], "response_mode": "offline", "answer": "[{\"word\":\"cattle\",\"start\":3.87,\"end\":4.3}]"} {"utterance_id": "3915-57461-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0004.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.65, 13.03, "hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"hundred\",\"start\":12.65,\"end\":13.03}]"} {"utterance_id": "6123-59186-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0014.flac", "task_type": "lexical_detection", "query": "ride", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"ride\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.29, 10.59, "ride"]], "response_mode": "offline", "answer": "[{\"word\":\"ride\",\"start\":10.29,\"end\":10.59}]"} {"utterance_id": "6123-59150-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0033.flac", "task_type": "lexical_detection", "query": "because", "prompt": "After the audio ends, extract all occurrences of the word \"because\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.73, "because"]], "response_mode": "offline", "answer": "[{\"word\":\"because\",\"start\":0.4,\"end\":0.73}]"} {"utterance_id": "1650-157641-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0004.flac", "task_type": "lexical_detection", "query": "brass", "prompt": "When the utterance is complete, report all instances of the word \"brass\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.5, "brass"]], "response_mode": "offline", "answer": "[{\"word\":\"brass\",\"start\":1.14,\"end\":1.5}]"} {"utterance_id": "1701-141760-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0017.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.54, "guards"]], "response_mode": "offline", "answer": "[{\"word\":\"guards\",\"start\":1.12,\"end\":1.54}]"} {"utterance_id": "3915-98647-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0012.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[20.94, 21.18, "cakes"]], "response_mode": "offline", "answer": "[{\"word\":\"cakes\",\"start\":20.94,\"end\":21.18}]"} {"utterance_id": "6123-59150-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.92, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":1.27,\"end\":1.92}]"} {"utterance_id": "3663-172528-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0041.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.4, 11.8, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":11.4,\"end\":11.8}]"} {"utterance_id": "5849-50873-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0004.flac", "task_type": "lexical_detection", "query": "subdue", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"subdue\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.91, 4.37, "subdue"]], "response_mode": "offline", "answer": "[{\"word\":\"subdue\",\"start\":3.91,\"end\":4.37}]"} {"utterance_id": "8173-294714-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0054.flac", "task_type": "lexical_detection", "query": "alarm", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"alarm\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.57, 5.2, "alarm"]], "response_mode": "offline", "answer": "[{\"word\":\"alarm\",\"start\":4.57,\"end\":5.2}]"} {"utterance_id": "5543-27761-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0026.flac", "task_type": "lexical_detection", "query": "doors", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"doors\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.54, "doors"]], "response_mode": "offline", "answer": "[{\"word\":\"doors\",\"start\":1.25,\"end\":1.54}]"} {"utterance_id": "7697-245712-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0018.flac", "task_type": "lexical_detection", "query": "innocence", "prompt": "After the audio ends, extract all occurrences of the word \"innocence\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.88, "innocence"]], "response_mode": "offline", "answer": "[{\"word\":\"innocence\",\"start\":1.37,\"end\":1.88}]"} {"utterance_id": "8254-84205-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0029.flac", "task_type": "lexical_detection", "query": "risk", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"risk\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.21, 4.54, "risk"]], "response_mode": "offline", "answer": "[{\"word\":\"risk\",\"start\":4.21,\"end\":4.54}]"} {"utterance_id": "7601-291468-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0006.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen to the full audio. After it finishes, find every instance of a food name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[21.42, 21.78, "meat"]], "response_mode": "offline", "answer": "[{\"word\":\"meat\",\"start\":21.42,\"end\":21.78}]"} {"utterance_id": "3915-98647-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0016.flac", "task_type": "lexical_detection", "query": "corpse", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"corpse\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.92, 6.45, "corpse"]], "response_mode": "offline", "answer": "[{\"word\":\"corpse\",\"start\":5.92,\"end\":6.45}]"} {"utterance_id": "6267-53049-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0004.flac", "task_type": "lexical_detection", "query": "penelope", "prompt": "After the audio ends, extract all occurrences of the word \"penelope\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.77, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.97,\"end\":1.77}]"} {"utterance_id": "3660-172182-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0019.flac", "task_type": "lexical_detection", "query": "household", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"household\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.89, 1.55, "household"]], "response_mode": "offline", "answer": "[{\"word\":\"household\",\"start\":0.89,\"end\":1.55}]"} {"utterance_id": "3660-172182-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0005.flac", "task_type": "lexical_detection", "query": "knight", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"knight\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.99, 5.28, "knight"]], "response_mode": "offline", "answer": "[{\"word\":\"knight\",\"start\":4.99,\"end\":5.28}]"} {"utterance_id": "8254-115543-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0018.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After the audio ends, extract all occurrences of a plant name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.49, 3.92, "cane"]], "response_mode": "offline", "answer": "[{\"word\":\"cane\",\"start\":3.49,\"end\":3.92}]"} {"utterance_id": "116-288045-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0029.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the full audio. Count every occurrence of a body part. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.88, 3.38, "one"], [4.69, 5.22, "two"], [10.48, 11.0, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.88,\"end\":3.38},{\"word\":\"two\",\"start\":4.69,\"end\":5.22},{\"word\":\"three\",\"start\":10.48,\"end\":11.0}]"} {"utterance_id": "6267-65525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 3.66, "ellen douglas's"]], "response_mode": "offline", "answer": "[{\"word\":\"ellen douglas's\",\"start\":2.38,\"end\":3.66}]"} {"utterance_id": "6841-88294-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0042.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.54, "eye"], [2.75, 3.24, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"eye\",\"start\":1.22,\"end\":1.54},{\"word\":\"lips\",\"start\":2.75,\"end\":3.24}]"} {"utterance_id": "8288-274162-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0054.flac", "task_type": "lexical_detection", "query": "step", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"step\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.15, 2.48, "step"]], "response_mode": "offline", "answer": "[{\"word\":\"step\",\"start\":2.15,\"end\":2.48}]"} {"utterance_id": "3915-57461-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0013.flac", "task_type": "lexical_detection", "query": "thief", "prompt": "After listening to the whole recording, output every instance of the word \"thief\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.11, 6.55, "thief"]], "response_mode": "offline", "answer": "[{\"word\":\"thief\",\"start\":6.11,\"end\":6.55}]"} {"utterance_id": "8254-115543-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0043.flac", "task_type": "lexical_detection", "query": "river", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"river\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.41, "river"]], "response_mode": "offline", "answer": "[{\"word\":\"river\",\"start\":1.97,\"end\":2.41}]"} {"utterance_id": "6267-53049-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0027.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.95, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":1.59,\"end\":1.95}]"} {"utterance_id": "3663-172528-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.77, "felice"]], "response_mode": "offline", "answer": "[{\"word\":\"felice\",\"start\":0.34,\"end\":0.77}]"} {"utterance_id": "3660-172182-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0038.flac", "task_type": "lexical_detection", "query": "fulness", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fulness\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.59, 2.06, "fulness"]], "response_mode": "offline", "answer": "[{\"word\":\"fulness\",\"start\":1.59,\"end\":2.06}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[6.96, 7.35, "calves"]], "response_mode": "offline", "answer": "[{\"word\":\"calves\",\"start\":6.96,\"end\":7.35}]"} {"utterance_id": "7697-105815-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0014.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.62, 7.4, "mount zion"]], "response_mode": "offline", "answer": "[{\"word\":\"mount zion\",\"start\":6.62,\"end\":7.4}]"} {"utterance_id": "4323-18416-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0002.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.2, 0.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":0.2,\"end\":0.42}]"} {"utterance_id": "4515-11057-0106", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0106.flac", "task_type": "lexical_detection", "query": "shelter", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"shelter\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.66, 4.1, "shelter"]], "response_mode": "offline", "answer": "[{\"word\":\"shelter\",\"start\":3.66,\"end\":4.1}]"} {"utterance_id": "4515-11057-0103", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0103.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.19, 2.72, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":2.19,\"end\":2.72}]"} {"utterance_id": "2506-11267-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0015.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.65, "bull"]], "response_mode": "offline", "answer": "[{\"word\":\"bull\",\"start\":0.32,\"end\":0.65}]"} {"utterance_id": "8254-115543-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0005.flac", "task_type": "lexical_detection", "query": "wagon", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"wagon\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.68, 6.17, "wagon"]], "response_mode": "offline", "answer": "[{\"word\":\"wagon\",\"start\":5.68,\"end\":6.17}]"} {"utterance_id": "7601-101619-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.31, "socrates"], [15.76, 16.75, "metrodorus"]], "response_mode": "offline", "answer": "[{\"word\":\"socrates\",\"start\":0.54,\"end\":1.31},{\"word\":\"metrodorus\",\"start\":15.76,\"end\":16.75}]"} {"utterance_id": "8254-115543-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0028.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.44, 2.83, "bullock"]], "response_mode": "offline", "answer": "[{\"word\":\"bullock\",\"start\":2.44,\"end\":2.83}]"} {"utterance_id": "8254-84205-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0020.flac", "task_type": "lexical_detection", "query": "trap", "prompt": "After the audio ends, extract all occurrences of the word \"trap\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.24, 1.57, "trap"]], "response_mode": "offline", "answer": "[{\"word\":\"trap\",\"start\":1.24,\"end\":1.57}]"} {"utterance_id": "8288-274150-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0005.flac", "task_type": "lexical_detection", "query": "day", "prompt": "After listening to the whole recording, output every instance of the word \"day\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.78, 2.09, "day"]], "response_mode": "offline", "answer": "[{\"word\":\"day\",\"start\":1.78,\"end\":2.09}]"} {"utterance_id": "4831-18525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 1.81, "polly"], [5.21, 5.65, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":1.53,\"end\":1.81},{\"word\":\"jasper\",\"start\":5.21,\"end\":5.65}]"} {"utterance_id": "4831-18525-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0018.flac", "task_type": "lexical_detection", "query": "responsibility", "prompt": "After listening to the whole recording, output every instance of the word \"responsibility\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.89, 4.74, "responsibility"]], "response_mode": "offline", "answer": "[{\"word\":\"responsibility\",\"start\":3.89,\"end\":4.74}]"} {"utterance_id": "3915-98647-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0015.flac", "task_type": "lexical_counting", "query": "old", "prompt": "Once the audio has finished, list each occurrence of the word \"old\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[8.72, 8.98, "one"], [10.71, 10.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.72,\"end\":8.98},{\"word\":\"two\",\"start\":10.71,\"end\":10.94}]"} {"utterance_id": "700-122867-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0006.flac", "task_type": "lexical_detection", "query": "deeper", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"deeper\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.48, "deeper"]], "response_mode": "offline", "answer": "[{\"word\":\"deeper\",\"start\":1.11,\"end\":1.48}]"} {"utterance_id": "7697-105815-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0017.flac", "task_type": "lexical_detection", "query": "doubt", "prompt": "Process the full audio offline. Find every instance of the word \"doubt\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.15, "doubt"]], "response_mode": "offline", "answer": "[{\"word\":\"doubt\",\"start\":1.88,\"end\":2.15}]"} {"utterance_id": "6267-65525-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.32, 0.77, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":0.32,\"end\":0.77}]"} {"utterance_id": "1650-173551-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0003.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.98, 3.5, "second"]], "response_mode": "offline", "answer": "[{\"word\":\"second\",\"start\":2.98,\"end\":3.5}]"} {"utterance_id": "700-122868-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.25, "complexion"], [4.76, 4.99, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"complexion\",\"start\":0.7,\"end\":1.25},{\"word\":\"hair\",\"start\":4.76,\"end\":4.99}]"} {"utterance_id": "7601-101622-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0001.flac", "task_type": "lexical_detection", "query": "luxury", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"luxury\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.89, 4.46, "luxury"]], "response_mode": "offline", "answer": "[{\"word\":\"luxury\",\"start\":3.89,\"end\":4.46}]"} {"utterance_id": "5849-50963-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0009.flac", "task_type": "lexical_detection", "query": "laughlin", "prompt": "Process the full audio offline. Find every instance of the word \"laughlin\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.59, 4.1, "laughlin"]], "response_mode": "offline", "answer": "[{\"word\":\"laughlin\",\"start\":3.59,\"end\":4.1}]"} {"utterance_id": "1686-142278-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0015.flac", "task_type": "lexical_detection", "query": "papa", "prompt": "After the audio ends, extract all occurrences of the word \"papa\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.61, "papa"]], "response_mode": "offline", "answer": "[{\"word\":\"papa\",\"start\":1.16,\"end\":1.61}]"} {"utterance_id": "4323-18416-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0024.flac", "task_type": "lexical_detection", "query": "there's", "prompt": "Once the audio has finished, identify every instance of the word \"there's\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.59, 2.77, "there's"]], "response_mode": "offline", "answer": "[{\"word\":\"there's\",\"start\":2.59,\"end\":2.77}]"} {"utterance_id": "4515-11057-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0053.flac", "task_type": "lexical_detection", "query": "yes", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"yes\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.21, "yes"]], "response_mode": "offline", "answer": "[{\"word\":\"yes\",\"start\":0.54,\"end\":1.21}]"} {"utterance_id": "700-122868-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0008.flac", "task_type": "lexical_detection", "query": "nothing", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"nothing\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.49, 2.73, "nothing"]], "response_mode": "offline", "answer": "[{\"word\":\"nothing\",\"start\":2.49,\"end\":2.73}]"} {"utterance_id": "8288-274162-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0039.flac", "task_type": "lexical_detection", "query": "saying", "prompt": "When the utterance is complete, report all instances of the word \"saying\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.37, "saying"]], "response_mode": "offline", "answer": "[{\"word\":\"saying\",\"start\":2.0,\"end\":2.37}]"} {"utterance_id": "700-122866-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0029.flac", "task_type": "lexical_detection", "query": "take", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"take\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.13, 1.34, "take"]], "response_mode": "offline", "answer": "[{\"word\":\"take\",\"start\":1.13,\"end\":1.34}]"} {"utterance_id": "1255-74899-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0000.flac", "task_type": "lexical_detection", "query": "arrange", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"arrange\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.21, "arrange"]], "response_mode": "offline", "answer": "[{\"word\":\"arrange\",\"start\":0.73,\"end\":1.21}]"} {"utterance_id": "6455-67804-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0031.flac", "task_type": "lexical_detection", "query": "bad", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"bad\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.3, "bad"]], "response_mode": "offline", "answer": "[{\"word\":\"bad\",\"start\":1.05,\"end\":1.3}]"} {"utterance_id": "1585-131718-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0012.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen carefully to the entire clip, then report every instance of an organization name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.2, 5.38, "national electric light association"], [10.63, 12.07, "edison lighting companies"]], "response_mode": "offline", "answer": "[{\"word\":\"national electric light association\",\"start\":3.2,\"end\":5.38},{\"word\":\"edison lighting companies\",\"start\":10.63,\"end\":12.07}]"} {"utterance_id": "8254-84205-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0028.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.26, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":0.93,\"end\":1.26}]"} {"utterance_id": "3915-98647-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0017.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.59, 7.06, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":6.59,\"end\":7.06}]"} {"utterance_id": "1686-142278-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.52, 1.06, "mister hale"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale\",\"start\":0.52,\"end\":1.06}]"} {"utterance_id": "4570-14911-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0004.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.22, 5.64, "hound"]], "response_mode": "offline", "answer": "[{\"word\":\"hound\",\"start\":5.22,\"end\":5.64}]"} {"utterance_id": "3663-172528-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0009.flac", "task_type": "lexical_detection", "query": "bernardino", "prompt": "Process the full audio offline. Find every instance of the word \"bernardino\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 3.56, "bernardino"]], "response_mode": "offline", "answer": "[{\"word\":\"bernardino\",\"start\":2.57,\"end\":3.56}]"} {"utterance_id": "116-288047-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0010.flac", "task_type": "lexical_detection", "query": "fairies", "prompt": "Once the audio has finished, identify every instance of the word \"fairies\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.17, 3.84, "fairies"]], "response_mode": "offline", "answer": "[{\"word\":\"fairies\",\"start\":3.17,\"end\":3.84}]"} {"utterance_id": "1701-141760-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0050.flac", "task_type": "lexical_detection", "query": "bow", "prompt": "Once the audio has finished, identify every instance of the word \"bow\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.3, 2.51, "bow"]], "response_mode": "offline", "answer": "[{\"word\":\"bow\",\"start\":2.3,\"end\":2.51}]"} {"utterance_id": "6267-65525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.24, "foot"], [4.36, 4.79, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":1.84,\"end\":2.24},{\"word\":\"hand\",\"start\":4.36,\"end\":4.79}]"} {"utterance_id": "4323-55228-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.27, 0.53, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.27,\"end\":0.53}]"} {"utterance_id": "6841-88291-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0017.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.53, "side"]], "response_mode": "offline", "answer": "[{\"word\":\"side\",\"start\":2.12,\"end\":2.53}]"} {"utterance_id": "1630-141772-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0017.flac", "task_type": "lexical_detection", "query": "seek", "prompt": "When the utterance is complete, report all instances of the word \"seek\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.53, "seek"]], "response_mode": "offline", "answer": "[{\"word\":\"seek\",\"start\":2.18,\"end\":2.53}]"} {"utterance_id": "6841-88294-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0031.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.3, 1.58, "meal"]], "response_mode": "offline", "answer": "[{\"word\":\"meal\",\"start\":1.3,\"end\":1.58}]"} {"utterance_id": "4153-186222-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0033.flac", "task_type": "lexical_detection", "query": "belonging", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"belonging\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.76, 2.16, "belonging"]], "response_mode": "offline", "answer": "[{\"word\":\"belonging\",\"start\":1.76,\"end\":2.16}]"} {"utterance_id": "5543-27761-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0058.flac", "task_type": "lexical_counting", "query": "child's", "prompt": "After listening to the whole recording, output timed count words for the word \"child's\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[3.81, 4.27, "one"], [7.03, 7.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.81,\"end\":4.27},{\"word\":\"two\",\"start\":7.03,\"end\":7.53}]"} {"utterance_id": "1585-131718-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0049.flac", "task_type": "lexical_counting", "query": "telling", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"telling\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.85, 1.16, "one"], [2.58, 2.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.85,\"end\":1.16},{\"word\":\"two\",\"start\":2.58,\"end\":2.94}]"} {"utterance_id": "8173-294714-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0024.flac", "task_type": "lexical_detection", "query": "speak", "prompt": "When the utterance is complete, report all instances of the word \"speak\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.2, 2.5, "speak"]], "response_mode": "offline", "answer": "[{\"word\":\"speak\",\"start\":2.2,\"end\":2.5}]"} {"utterance_id": "2506-169427-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0002.flac", "task_type": "lexical_detection", "query": "debts", "prompt": "After the audio ends, extract all occurrences of the word \"debts\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.53, 6.94, "debts"]], "response_mode": "offline", "answer": "[{\"word\":\"debts\",\"start\":6.53,\"end\":6.94}]"} {"utterance_id": "3915-98647-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0011.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.67, 4.08, "builder"]], "response_mode": "offline", "answer": "[{\"word\":\"builder\",\"start\":3.67,\"end\":4.08}]"} {"utterance_id": "5543-27761-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0035.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.26, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":1.92,\"end\":2.26}]"} {"utterance_id": "4515-11057-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.84, 5.41, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":4.84,\"end\":5.41}]"} {"utterance_id": "8254-84205-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0039.flac", "task_type": "lexical_detection", "query": "father", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"father\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.67, 2.08, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.67,\"end\":2.08}]"} {"utterance_id": "1630-73710-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.61, 4.48, "henrietta"]], "response_mode": "offline", "answer": "[{\"word\":\"henrietta\",\"start\":3.61,\"end\":4.48}]"} {"utterance_id": "7697-245712-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0003.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an ordinal number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.92, "secondly"]], "response_mode": "offline", "answer": "[{\"word\":\"secondly\",\"start\":0.21,\"end\":0.92}]"} {"utterance_id": "7641-96670-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0018.flac", "task_type": "lexical_detection", "query": "softly", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"softly\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.5, "softly"]], "response_mode": "offline", "answer": "[{\"word\":\"softly\",\"start\":0.87,\"end\":1.5}]"} {"utterance_id": "5849-50873-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0029.flac", "task_type": "lexical_counting", "query": "soldiers", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"soldiers\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.33, 0.85, "one"], [8.81, 9.52, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.33,\"end\":0.85},{\"word\":\"two\",\"start\":8.81,\"end\":9.52}]"} {"utterance_id": "1701-141760-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0019.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen to the full audio. After it finishes, find every instance of a money or currency term. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.45, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":1.22,\"end\":1.45}]"} {"utterance_id": "6467-94831-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0041.flac", "task_type": "lexical_detection", "query": "donkey's", "prompt": "After the audio ends, extract all occurrences of the word \"donkey's\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.13, "donkey's"]], "response_mode": "offline", "answer": "[{\"word\":\"donkey's\",\"start\":0.59,\"end\":1.13}]"} {"utterance_id": "1701-141760-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.62, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":2.13,\"end\":2.62}]"} {"utterance_id": "4831-25894-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.19, 4.57, "ranza"], [9.33, 9.73, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"ranza\",\"start\":4.19,\"end\":4.57},{\"word\":\"tessa\",\"start\":9.33,\"end\":9.73}]"} {"utterance_id": "8254-84205-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0053.flac", "task_type": "lexical_detection", "query": "i've", "prompt": "Process the full audio offline. Find every instance of the word \"i've\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.02, "i've"]], "response_mode": "offline", "answer": "[{\"word\":\"i've\",\"start\":1.77,\"end\":2.02}]"} {"utterance_id": "2506-11278-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0032.flac", "task_type": "lexical_detection", "query": "what's", "prompt": "After the audio ends, extract all occurrences of the word \"what's\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.43, "what's"]], "response_mode": "offline", "answer": "[{\"word\":\"what's\",\"start\":2.1,\"end\":2.43}]"} {"utterance_id": "6467-94831-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0040.flac", "task_type": "lexical_detection", "query": "peleg", "prompt": "After the audio ends, extract all occurrences of the word \"peleg\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.82, "peleg"]], "response_mode": "offline", "answer": "[{\"word\":\"peleg\",\"start\":1.4,\"end\":1.82}]"} {"utterance_id": "4323-55228-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.19, 1.44, "kate"], [2.79, 3.18, "bertram's"]], "response_mode": "offline", "answer": "[{\"word\":\"kate\",\"start\":1.19,\"end\":1.44},{\"word\":\"bertram's\",\"start\":2.79,\"end\":3.18}]"} {"utterance_id": "700-122867-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0034.flac", "task_type": "lexical_detection", "query": "makes", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"makes\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.59, 0.84, "makes"]], "response_mode": "offline", "answer": "[{\"word\":\"makes\",\"start\":0.59,\"end\":0.84}]"} {"utterance_id": "116-288045-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0011.flac", "task_type": "lexical_detection", "query": "athens", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"athens\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.27, "athens"]], "response_mode": "offline", "answer": "[{\"word\":\"athens\",\"start\":0.87,\"end\":1.27}]"} {"utterance_id": "8254-84205-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0037.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.0, 3.42, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":3.0,\"end\":3.42}]"} {"utterance_id": "2506-11278-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0004.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.7, 5.86, "parasites"]], "response_mode": "offline", "answer": "[{\"word\":\"parasites\",\"start\":4.7,\"end\":5.86}]"} {"utterance_id": "4153-185072-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0005.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.28, 8.86, "niagara"]], "response_mode": "offline", "answer": "[{\"word\":\"niagara\",\"start\":8.28,\"end\":8.86}]"} {"utterance_id": "1630-73710-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.43, 2.75, "bird"]], "response_mode": "offline", "answer": "[{\"word\":\"bird\",\"start\":2.43,\"end\":2.75}]"} {"utterance_id": "5849-50962-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0011.flac", "task_type": "lexical_detection", "query": "showed", "prompt": "Once the audio has finished, identify every instance of the word \"showed\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.2, 3.54, "showed"]], "response_mode": "offline", "answer": "[{\"word\":\"showed\",\"start\":3.2,\"end\":3.54}]"} {"utterance_id": "4323-13259-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0003.flac", "task_type": "lexical_detection", "query": "condemned", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"condemned\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 1.96, "condemned"]], "response_mode": "offline", "answer": "[{\"word\":\"condemned\",\"start\":1.5,\"end\":1.96}]"} {"utterance_id": "7697-105815-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0051.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.5, "ass"]], "response_mode": "offline", "answer": "[{\"word\":\"ass\",\"start\":2.18,\"end\":2.5}]"} {"utterance_id": "116-288047-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0011.flac", "task_type": "lexical_detection", "query": "covet", "prompt": "Process the full audio offline. Find every instance of the word \"covet\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.83, "covet"]], "response_mode": "offline", "answer": "[{\"word\":\"covet\",\"start\":0.5,\"end\":0.83}]"} {"utterance_id": "8254-84205-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0012.flac", "task_type": "lexical_detection", "query": "we're", "prompt": "Process the full audio offline. Find every instance of the word \"we're\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 0.89, "we're"]], "response_mode": "offline", "answer": "[{\"word\":\"we're\",\"start\":0.79,\"end\":0.89}]"} {"utterance_id": "1701-141760-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.41, 1.08, "prince andrew"], [7.16, 7.67, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":0.41,\"end\":1.08},{\"word\":\"boris\",\"start\":7.16,\"end\":7.67}]"} {"utterance_id": "4323-55228-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.42, 0.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.42,\"end\":0.63}]"} {"utterance_id": "1650-157641-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0000.flac", "task_type": "lexical_detection", "query": "victorian", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"victorian\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.01, 3.67, "victorian"]], "response_mode": "offline", "answer": "[{\"word\":\"victorian\",\"start\":3.01,\"end\":3.67}]"} {"utterance_id": "4323-55228-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 2.49, "knees"], [4.78, 5.0, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"knees\",\"start\":2.3,\"end\":2.49},{\"word\":\"eyes\",\"start\":4.78,\"end\":5.0}]"} {"utterance_id": "3915-98647-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0023.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "When the utterance is complete, report all instances of an ordinal number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.73, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":0.42,\"end\":0.73}]"} {"utterance_id": "6455-66379-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0000.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.81, 2.14, "wood"]], "response_mode": "offline", "answer": "[{\"word\":\"wood\",\"start\":1.81,\"end\":2.14}]"} {"utterance_id": "1630-73710-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0011.flac", "task_type": "lexical_detection", "query": "ferdinand", "prompt": "Once the audio has finished, identify every instance of the word \"ferdinand\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.53, 1.31, "ferdinand"]], "response_mode": "offline", "answer": "[{\"word\":\"ferdinand\",\"start\":0.53,\"end\":1.31}]"} {"utterance_id": "3660-172182-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0004.flac", "task_type": "lexical_detection", "query": "beheld", "prompt": "Process the full audio offline. Find every instance of the word \"beheld\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.74, "beheld"]], "response_mode": "offline", "answer": "[{\"word\":\"beheld\",\"start\":1.37,\"end\":1.74}]"} {"utterance_id": "2506-11267-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0013.flac", "task_type": "lexical_detection", "query": "splendid", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"splendid\" with fields word, start, end (seconds). If none, reply [].", "labels": [[8.73, 9.57, "splendid"]], "response_mode": "offline", "answer": "[{\"word\":\"splendid\",\"start\":8.73,\"end\":9.57}]"} {"utterance_id": "4831-18525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0002.flac", "task_type": "lexical_detection", "query": "sober", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"sober\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.69, 2.05, "sober"]], "response_mode": "offline", "answer": "[{\"word\":\"sober\",\"start\":1.69,\"end\":2.05}]"} {"utterance_id": "116-288048-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.21, 6.82, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":6.21,\"end\":6.82}]"} {"utterance_id": "7601-291468-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0006.flac", "task_type": "lexical_counting", "query": "dutch", "prompt": "After listening to the whole recording, output timed count words for the word \"dutch\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[9.91, 10.42, "one"], [19.84, 20.15, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.91,\"end\":10.42},{\"word\":\"two\",\"start\":19.84,\"end\":20.15}]"} {"utterance_id": "4323-55228-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0050.flac", "task_type": "lexical_detection", "query": "nodded", "prompt": "Once the audio has finished, identify every instance of the word \"nodded\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.88, 1.22, "nodded"]], "response_mode": "offline", "answer": "[{\"word\":\"nodded\",\"start\":0.88,\"end\":1.22}]"} {"utterance_id": "1651-136854-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0006.flac", "task_type": "lexical_detection", "query": "instance", "prompt": "Once the audio has finished, identify every instance of the word \"instance\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.72, 1.42, "instance"]], "response_mode": "offline", "answer": "[{\"word\":\"instance\",\"start\":0.72,\"end\":1.42}]"} {"utterance_id": "1650-167613-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0033.flac", "task_type": "semantic_detection", "query": "country", "prompt": "You will hear a complete utterance. When it ends, list every instance of a country name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.85, "french"]], "response_mode": "offline", "answer": "[{\"word\":\"french\",\"start\":0.42,\"end\":0.85}]"} {"utterance_id": "7697-105817-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0002.flac", "task_type": "lexical_detection", "query": "sum", "prompt": "Once the audio has finished, identify every instance of the word \"sum\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.6, 1.85, "sum"]], "response_mode": "offline", "answer": "[{\"word\":\"sum\",\"start\":1.6,\"end\":1.85}]"} {"utterance_id": "1686-142278-0087", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0087.flac", "task_type": "lexical_detection", "query": "almost", "prompt": "Process the full audio offline. Find every instance of the word \"almost\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.54, "almost"]], "response_mode": "offline", "answer": "[{\"word\":\"almost\",\"start\":1.21,\"end\":1.54}]"} {"utterance_id": "3915-98647-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.18, 6.81, "fortune teller"]], "response_mode": "offline", "answer": "[{\"word\":\"fortune teller\",\"start\":6.18,\"end\":6.81}]"} {"utterance_id": "8254-84205-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0043.flac", "task_type": "lexical_detection", "query": "sharply", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sharply\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.94, 2.37, "sharply"]], "response_mode": "offline", "answer": "[{\"word\":\"sharply\",\"start\":1.94,\"end\":2.37}]"} {"utterance_id": "4515-11057-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0028.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.97, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":0.68,\"end\":0.97}]"} {"utterance_id": "2506-11267-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0011.flac", "task_type": "lexical_detection", "query": "houndsditch", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"houndsditch\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.09, 1.7, "houndsditch"]], "response_mode": "offline", "answer": "[{\"word\":\"houndsditch\",\"start\":1.09,\"end\":1.7}]"} {"utterance_id": "116-288048-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.42, 1.71, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":1.42,\"end\":1.71}]"} {"utterance_id": "1585-157660-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0004.flac", "task_type": "lexical_detection", "query": "bat", "prompt": "After listening to the whole recording, output every instance of the word \"bat\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.69, 11.05, "bat"]], "response_mode": "offline", "answer": "[{\"word\":\"bat\",\"start\":10.69,\"end\":11.05}]"} {"utterance_id": "1686-142278-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0043.flac", "task_type": "lexical_detection", "query": "myself", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"myself\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.06, 2.75, "myself"]], "response_mode": "offline", "answer": "[{\"word\":\"myself\",\"start\":2.06,\"end\":2.75}]"} {"utterance_id": "1630-96099-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0022.flac", "task_type": "lexical_detection", "query": "hurry", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"hurry\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.66, 3.23, "hurry"]], "response_mode": "offline", "answer": "[{\"word\":\"hurry\",\"start\":2.66,\"end\":3.23}]"} {"utterance_id": "1630-96099-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0007.flac", "task_type": "lexical_detection", "query": "pleasant", "prompt": "After the audio ends, extract all occurrences of the word \"pleasant\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.72, "pleasant"]], "response_mode": "offline", "answer": "[{\"word\":\"pleasant\",\"start\":2.12,\"end\":2.72}]"} {"utterance_id": "6455-67803-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0004.flac", "task_type": "lexical_detection", "query": "octavia", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"octavia\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.89, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":2.25,\"end\":2.89}]"} {"utterance_id": "4831-18525-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0003.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.58, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.1,\"end\":1.58}]"} {"utterance_id": "4831-25894-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0011.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the full audio. Count every occurrence of a body part. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.52, "one"], [4.94, 5.23, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.25,\"end\":1.52},{\"word\":\"two\",\"start\":4.94,\"end\":5.23}]"} {"utterance_id": "4831-25894-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.51, 3.84, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":3.51,\"end\":3.84}]"} {"utterance_id": "7641-96684-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0015.flac", "task_type": "lexical_counting", "query": "death", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"death\" (word, start, end). If none, reply [].", "labels": [[1.65, 2.01, "one"], [6.16, 6.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.65,\"end\":2.01},{\"word\":\"two\",\"start\":6.16,\"end\":6.46}]"} {"utterance_id": "7641-96670-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0015.flac", "task_type": "lexical_counting", "query": "whether", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"whether\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.68, 1.93, "one"], [5.66, 5.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.68,\"end\":1.93},{\"word\":\"two\",\"start\":5.66,\"end\":5.91}]"} {"utterance_id": "1650-167613-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0026.flac", "task_type": "lexical_detection", "query": "tomorrow", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"tomorrow\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.45, 0.94, "tomorrow"]], "response_mode": "offline", "answer": "[{\"word\":\"tomorrow\",\"start\":0.45,\"end\":0.94}]"} {"utterance_id": "1701-141760-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.55, 1.99, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":1.55,\"end\":1.99}]"} {"utterance_id": "4323-55228-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.47, "pete"], [3.08, 3.53, "aunt hannah"]], "response_mode": "offline", "answer": "[{\"word\":\"pete\",\"start\":0.26,\"end\":0.47},{\"word\":\"aunt hannah\",\"start\":3.08,\"end\":3.53}]"} {"utterance_id": "4831-29134-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.92, "senator"], [2.38, 2.81, "minister"], [4.05, 4.77, "superintendent"]], "response_mode": "offline", "answer": "[{\"word\":\"senator\",\"start\":0.54,\"end\":0.92},{\"word\":\"minister\",\"start\":2.38,\"end\":2.81},{\"word\":\"superintendent\",\"start\":4.05,\"end\":4.77}]"} {"utterance_id": "1630-96099-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0021.flac", "task_type": "lexical_detection", "query": "francs", "prompt": "After the audio ends, extract all occurrences of the word \"francs\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.17, 4.99, "francs"], [7.83, 8.48, "francs"]], "response_mode": "offline", "answer": "[{\"word\":\"francs\",\"start\":4.17,\"end\":4.99},{\"word\":\"francs\",\"start\":7.83,\"end\":8.48}]"} {"utterance_id": "4570-14911-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0005.flac", "task_type": "lexical_detection", "query": "beagle", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"beagle\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.41, 7.1, "beagle"]], "response_mode": "offline", "answer": "[{\"word\":\"beagle\",\"start\":6.41,\"end\":7.1}]"} {"utterance_id": "1650-157641-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0007.flac", "task_type": "lexical_detection", "query": "odour", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"odour\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.91, "odour"]], "response_mode": "offline", "answer": "[{\"word\":\"odour\",\"start\":3.39,\"end\":3.91}]"} {"utterance_id": "4153-185072-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0002.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.39, 4.14, "victuals"]], "response_mode": "offline", "answer": "[{\"word\":\"victuals\",\"start\":3.39,\"end\":4.14}]"} {"utterance_id": "1701-141760-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0008.flac", "task_type": "lexical_detection", "query": "changed", "prompt": "After the audio ends, extract all occurrences of the word \"changed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.89, 2.58, "changed"]], "response_mode": "offline", "answer": "[{\"word\":\"changed\",\"start\":1.89,\"end\":2.58}]"} {"utterance_id": "8254-84205-0073", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0073.flac", "task_type": "lexical_detection", "query": "starve", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"starve\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.3, 0.8, "starve"]], "response_mode": "offline", "answer": "[{\"word\":\"starve\",\"start\":0.3,\"end\":0.8}]"} {"utterance_id": "1630-73710-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0012.flac", "task_type": "lexical_counting", "query": "henrietta", "prompt": "Once the audio has finished, list each occurrence of the word \"henrietta\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.37, 2.1, "one"], [4.36, 5.07, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.37,\"end\":2.1},{\"word\":\"two\",\"start\":4.36,\"end\":5.07}]"} {"utterance_id": "1255-138279-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0006.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Process the full audio offline. Find every instance of a weather term and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.41, 4.9, "snow"]], "response_mode": "offline", "answer": "[{\"word\":\"snow\",\"start\":4.41,\"end\":4.9}]"} {"utterance_id": "6267-65525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0000.flac", "task_type": "lexical_detection", "query": "below", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"below\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.65, 6.06, "below"]], "response_mode": "offline", "answer": "[{\"word\":\"below\",\"start\":5.65,\"end\":6.06}]"} {"utterance_id": "8288-274162-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0009.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After the audio ends, extract all occurrences of a country name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.51, "england"]], "response_mode": "offline", "answer": "[{\"word\":\"england\",\"start\":0.98,\"end\":1.51}]"} {"utterance_id": "1701-141760-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0016.flac", "task_type": "lexical_detection", "query": "dinners", "prompt": "When the utterance is complete, report all instances of the word \"dinners\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.82, "dinners"]], "response_mode": "offline", "answer": "[{\"word\":\"dinners\",\"start\":0.44,\"end\":0.82}]"} {"utterance_id": "8288-274162-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0035.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.22, 7.55, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":7.22,\"end\":7.55}]"} {"utterance_id": "4153-186223-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0003.flac", "task_type": "lexical_detection", "query": "discover", "prompt": "Process the full audio offline. Find every instance of the word \"discover\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.67, 3.22, "discover"]], "response_mode": "offline", "answer": "[{\"word\":\"discover\",\"start\":2.67,\"end\":3.22}]"} {"utterance_id": "1686-142278-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0024.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "When the utterance is complete, report all instances of a disease name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.65, "injury"]], "response_mode": "offline", "answer": "[{\"word\":\"injury\",\"start\":3.38,\"end\":3.65}]"} {"utterance_id": "700-122867-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0012.flac", "task_type": "lexical_detection", "query": "scrape", "prompt": "After the audio ends, extract all occurrences of the word \"scrape\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.31, 1.59, "scrape"]], "response_mode": "offline", "answer": "[{\"word\":\"scrape\",\"start\":1.31,\"end\":1.59}]"} {"utterance_id": "5849-50964-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.34, "bert"]], "response_mode": "offline", "answer": "[{\"word\":\"bert\",\"start\":1.99,\"end\":2.34}]"} {"utterance_id": "1650-167613-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.49, 1.12, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":0.49,\"end\":1.12}]"} {"utterance_id": "1701-141760-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0045.flac", "task_type": "lexical_detection", "query": "camp", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"camp\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.46, 4.94, "camp"]], "response_mode": "offline", "answer": "[{\"word\":\"camp\",\"start\":4.46,\"end\":4.94}]"} {"utterance_id": "6467-56885-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.56, 0.87, "body"], [4.57, 4.92, "sight"]], "response_mode": "offline", "answer": "[{\"word\":\"body\",\"start\":0.56,\"end\":0.87},{\"word\":\"sight\",\"start\":4.57,\"end\":4.92}]"} {"utterance_id": "6267-65525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0030.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.73, 0.99, "one"], [5.09, 5.59, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.73,\"end\":0.99},{\"word\":\"two\",\"start\":5.09,\"end\":5.59}]"} {"utterance_id": "4515-11057-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0010.flac", "task_type": "lexical_detection", "query": "news", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"news\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.42, "news"]], "response_mode": "offline", "answer": "[{\"word\":\"news\",\"start\":1.88,\"end\":2.42}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "lexical_counting", "query": "spite", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"spite\" (word, start, end). If none, reply [].", "labels": [[0.62, 0.96, "one"], [4.37, 4.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.62,\"end\":0.96},{\"word\":\"two\",\"start\":4.37,\"end\":4.7}]"} {"utterance_id": "8254-115543-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 2.69, "chola"]], "response_mode": "offline", "answer": "[{\"word\":\"chola\",\"start\":2.34,\"end\":2.69}]"} {"utterance_id": "7697-105815-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0024.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.01, 5.5, "beasts"]], "response_mode": "offline", "answer": "[{\"word\":\"beasts\",\"start\":5.01,\"end\":5.5}]"} {"utterance_id": "4323-55228-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0025.flac", "task_type": "lexical_detection", "query": "minute", "prompt": "Once the audio has finished, identify every instance of the word \"minute\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.65, 2.0, "minute"]], "response_mode": "offline", "answer": "[{\"word\":\"minute\",\"start\":1.65,\"end\":2.0}]"} {"utterance_id": "700-122866-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 3.27, "rosamond montmorency"]], "response_mode": "offline", "answer": "[{\"word\":\"rosamond montmorency\",\"start\":1.1,\"end\":3.27}]"} {"utterance_id": "6467-94831-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0043.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.77, "donkey's"]], "response_mode": "offline", "answer": "[{\"word\":\"donkey's\",\"start\":1.34,\"end\":1.77}]"} {"utterance_id": "4831-25894-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.48, 0.77, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":0.48,\"end\":0.77}]"} {"utterance_id": "4831-18525-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.25, "tom"], [5.12, 5.56, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":1.9,\"end\":2.25},{\"word\":\"tom\",\"start\":5.12,\"end\":5.56}]"} {"utterance_id": "4515-11057-0078", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0078.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.06, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"voice\",\"start\":2.7,\"end\":3.06}]"} {"utterance_id": "6267-65525-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0058.flac", "task_type": "lexical_detection", "query": "certainly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"certainly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.29, 1.76, "certainly"]], "response_mode": "offline", "answer": "[{\"word\":\"certainly\",\"start\":1.29,\"end\":1.76}]"} {"utterance_id": "2506-11278-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0018.flac", "task_type": "lexical_detection", "query": "name", "prompt": "When the utterance is complete, report all instances of the word \"name\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.36, "name"]], "response_mode": "offline", "answer": "[{\"word\":\"name\",\"start\":3.06,\"end\":3.36}]"} {"utterance_id": "7697-245712-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.18, 8.5, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":8.18,\"end\":8.5}]"} {"utterance_id": "8288-274162-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0003.flac", "task_type": "lexical_detection", "query": "ears", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"ears\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.83, "ears"]], "response_mode": "offline", "answer": "[{\"word\":\"ears\",\"start\":3.38,\"end\":3.83}]"} {"utterance_id": "4572-64670-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0001.flac", "task_type": "lexical_detection", "query": "eldest", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"eldest\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.79, 1.21, "eldest"]], "response_mode": "offline", "answer": "[{\"word\":\"eldest\",\"start\":0.79,\"end\":1.21}]"} {"utterance_id": "3660-172182-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[11.19, 11.39, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":11.19,\"end\":11.39}]"} {"utterance_id": "4515-11057-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0041.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After the audio ends, extract all occurrences of a color name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.53, "white"]], "response_mode": "offline", "answer": "[{\"word\":\"white\",\"start\":2.06,\"end\":2.53}]"} {"utterance_id": "6455-67804-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 1.07, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":0.42,\"end\":1.07}]"} {"utterance_id": "6455-67803-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0001.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.87, 3.11, "fourth"]], "response_mode": "offline", "answer": "[{\"word\":\"fourth\",\"start\":2.87,\"end\":3.11}]"} {"utterance_id": "6267-65525-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0025.flac", "task_type": "lexical_detection", "query": "henceforth", "prompt": "Once the audio has finished, identify every instance of the word \"henceforth\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.94, 9.82, "henceforth"]], "response_mode": "offline", "answer": "[{\"word\":\"henceforth\",\"start\":8.94,\"end\":9.82}]"} {"utterance_id": "4323-18416-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0014.flac", "task_type": "lexical_counting", "query": "grandpapa", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"grandpapa\" (word, start, end). If none, reply [].", "labels": [[0.64, 1.19, "one"], [4.58, 5.13, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.64,\"end\":1.19},{\"word\":\"two\",\"start\":4.58,\"end\":5.13}]"} {"utterance_id": "6267-53049-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0014.flac", "task_type": "lexical_detection", "query": "stuff", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"stuff\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.2, 2.78, "stuff"]], "response_mode": "offline", "answer": "[{\"word\":\"stuff\",\"start\":2.2,\"end\":2.78}]"} {"utterance_id": "1686-142278-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.37, 2.63, "god's"]], "response_mode": "offline", "answer": "[{\"word\":\"god's\",\"start\":2.37,\"end\":2.63}]"} {"utterance_id": "6123-59150-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0022.flac", "task_type": "lexical_detection", "query": "shouted", "prompt": "When the utterance is complete, report all instances of the word \"shouted\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.67, "shouted"]], "response_mode": "offline", "answer": "[{\"word\":\"shouted\",\"start\":1.39,\"end\":1.67}]"} {"utterance_id": "116-288047-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0004.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.14, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.88,\"end\":1.14}]"} {"utterance_id": "1255-74899-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0016.flac", "task_type": "lexical_detection", "query": "future", "prompt": "After the audio ends, extract all occurrences of the word \"future\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.37, 4.67, "future"]], "response_mode": "offline", "answer": "[{\"word\":\"future\",\"start\":4.37,\"end\":4.67}]"} {"utterance_id": "1650-167613-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0042.flac", "task_type": "lexical_detection", "query": "lips", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"lips\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.69, 7.07, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"lips\",\"start\":6.69,\"end\":7.07}]"} {"utterance_id": "8254-115543-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0032.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.54, 0.89, "servants"]], "response_mode": "offline", "answer": "[{\"word\":\"servants\",\"start\":0.54,\"end\":0.89}]"} {"utterance_id": "116-288047-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0001.flac", "task_type": "lexical_detection", "query": "questions", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"questions\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.43, "questions"]], "response_mode": "offline", "answer": "[{\"word\":\"questions\",\"start\":2.95,\"end\":3.43}]"} {"utterance_id": "4572-112381-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0002.flac", "task_type": "lexical_detection", "query": "thy", "prompt": "When the utterance is complete, report all instances of the word \"thy\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.58, 3.8, "thy"]], "response_mode": "offline", "answer": "[{\"word\":\"thy\",\"start\":3.58,\"end\":3.8}]"} {"utterance_id": "6467-94831-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0003.flac", "task_type": "lexical_detection", "query": "teacher", "prompt": "Process the full audio offline. Find every instance of the word \"teacher\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 2.05, "teacher"]], "response_mode": "offline", "answer": "[{\"word\":\"teacher\",\"start\":1.63,\"end\":2.05}]"} {"utterance_id": "3915-98647-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0013.flac", "task_type": "lexical_counting", "query": "come", "prompt": "Process the full audio offline. For every occurrence of the word \"come\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 0.85, "one"], [2.31, 2.54, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.57,\"end\":0.85},{\"word\":\"two\",\"start\":2.31,\"end\":2.54}]"} {"utterance_id": "3663-172528-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0052.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.46, 8.78, "two"], [9.9, 10.35, "single"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":8.46,\"end\":8.78},{\"word\":\"single\",\"start\":9.9,\"end\":10.35}]"} {"utterance_id": "700-122867-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0027.flac", "task_type": "lexical_detection", "query": "tragic", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"tragic\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.11, 1.55, "tragic"]], "response_mode": "offline", "answer": "[{\"word\":\"tragic\",\"start\":1.11,\"end\":1.55}]"} {"utterance_id": "4153-186222-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0022.flac", "task_type": "lexical_detection", "query": "cousin", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"cousin\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 4.0, "cousin"]], "response_mode": "offline", "answer": "[{\"word\":\"cousin\",\"start\":3.54,\"end\":4.0}]"} {"utterance_id": "1630-141772-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.49, 6.87, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":6.49,\"end\":6.87}]"} {"utterance_id": "4323-55228-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0014.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.84, 7.32, "hillside"]], "response_mode": "offline", "answer": "[{\"word\":\"hillside\",\"start\":6.84,\"end\":7.32}]"} {"utterance_id": "4831-18525-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.76, "adela"], [5.51, 5.86, "polly's"]], "response_mode": "offline", "answer": "[{\"word\":\"adela\",\"start\":1.46,\"end\":1.76},{\"word\":\"polly's\",\"start\":5.51,\"end\":5.86}]"} {"utterance_id": "7601-291468-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0007.flac", "task_type": "lexical_detection", "query": "bottom", "prompt": "Process the full audio offline. Find every instance of the word \"bottom\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.36, 14.86, "bottom"]], "response_mode": "offline", "answer": "[{\"word\":\"bottom\",\"start\":14.36,\"end\":14.86}]"} {"utterance_id": "1701-141760-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0023.flac", "task_type": "lexical_detection", "query": "rostov", "prompt": "When the utterance is complete, report all instances of the word \"rostov\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.6, 3.0, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"rostov\",\"start\":2.6,\"end\":3.0}]"} {"utterance_id": "7641-96670-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.88, "medea"]], "response_mode": "offline", "answer": "[{\"word\":\"medea\",\"start\":0.45,\"end\":0.88}]"} {"utterance_id": "6841-88294-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0044.flac", "task_type": "lexical_detection", "query": "odd", "prompt": "After listening to the whole recording, output every instance of the word \"odd\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.9, 4.12, "odd"]], "response_mode": "offline", "answer": "[{\"word\":\"odd\",\"start\":3.9,\"end\":4.12}]"} {"utterance_id": "4323-18416-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.44, 1.82, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.44,\"end\":1.82}]"} {"utterance_id": "1651-136854-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.32, 0.92, "addison's"]], "response_mode": "offline", "answer": "[{\"word\":\"addison's\",\"start\":0.32,\"end\":0.92}]"} {"utterance_id": "1630-73710-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0018.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.56, 3.01, "bath"]], "response_mode": "offline", "answer": "[{\"word\":\"bath\",\"start\":2.56,\"end\":3.01}]"} {"utterance_id": "4153-61735-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0008.flac", "task_type": "lexical_detection", "query": "likely", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"likely\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.65, 1.11, "likely"]], "response_mode": "offline", "answer": "[{\"word\":\"likely\",\"start\":0.65,\"end\":1.11}]"} {"utterance_id": "6841-88294-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0025.flac", "task_type": "lexical_detection", "query": "pleaded", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pleaded\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.1, 1.58, "pleaded"]], "response_mode": "offline", "answer": "[{\"word\":\"pleaded\",\"start\":1.1,\"end\":1.58}]"} {"utterance_id": "700-122868-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.2, 1.59, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":1.2,\"end\":1.59}]"} {"utterance_id": "7641-96252-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0004.flac", "task_type": "lexical_detection", "query": "rewards", "prompt": "Process the full audio offline. Find every instance of the word \"rewards\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.27, 6.74, "rewards"]], "response_mode": "offline", "answer": "[{\"word\":\"rewards\",\"start\":6.27,\"end\":6.74}]"} {"utterance_id": "1701-141759-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0025.flac", "task_type": "lexical_detection", "query": "armchair", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"armchair\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.83, 3.46, "armchair"]], "response_mode": "offline", "answer": "[{\"word\":\"armchair\",\"start\":2.83,\"end\":3.46}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.45, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.16,\"end\":0.45}]"} {"utterance_id": "1585-131718-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0045.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After listening to the whole recording, output every instance of a time of day with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.99, "night"], [3.14, 3.91, "twelve o'clock"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":0.62,\"end\":0.99},{\"word\":\"twelve o'clock\",\"start\":3.14,\"end\":3.91}]"} {"utterance_id": "7697-105815-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0012.flac", "task_type": "lexical_detection", "query": "vale", "prompt": "After the audio ends, extract all occurrences of the word \"vale\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 2.88, "vale"]], "response_mode": "offline", "answer": "[{\"word\":\"vale\",\"start\":2.4,\"end\":2.88}]"} {"utterance_id": "1701-141760-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.4, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"rostov\",\"start\":1.79,\"end\":2.4}]"} {"utterance_id": "1650-167613-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0016.flac", "task_type": "lexical_detection", "query": "meant", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"meant\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.58, "meant"]], "response_mode": "offline", "answer": "[{\"word\":\"meant\",\"start\":2.14,\"end\":2.58}]"} {"utterance_id": "3660-6517-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0034.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 1.97, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.8,\"end\":1.97}]"} {"utterance_id": "1255-90413-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0013.flac", "task_type": "lexical_detection", "query": "lonely", "prompt": "When the utterance is complete, report all instances of the word \"lonely\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.53, "lonely"]], "response_mode": "offline", "answer": "[{\"word\":\"lonely\",\"start\":1.19,\"end\":1.53}]"} {"utterance_id": "1585-131718-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0001.flac", "task_type": "lexical_detection", "query": "hard", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hard\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.93, "hard"]], "response_mode": "offline", "answer": "[{\"word\":\"hard\",\"start\":1.44,\"end\":1.93}]"} {"utterance_id": "4153-185072-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0004.flac", "task_type": "lexical_detection", "query": "thomas", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"thomas\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.71, 1.12, "thomas"]], "response_mode": "offline", "answer": "[{\"word\":\"thomas\",\"start\":0.71,\"end\":1.12}]"} {"utterance_id": "1650-167613-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0041.flac", "task_type": "lexical_detection", "query": "unexpected", "prompt": "After the audio ends, extract all occurrences of the word \"unexpected\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.56, 12.07, "unexpected"]], "response_mode": "offline", "answer": "[{\"word\":\"unexpected\",\"start\":11.56,\"end\":12.07}]"} {"utterance_id": "1255-90407-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0001.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.15, "rain"]], "response_mode": "offline", "answer": "[{\"word\":\"rain\",\"start\":0.86,\"end\":1.15}]"} {"utterance_id": "6123-59150-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0038.flac", "task_type": "lexical_detection", "query": "compassion", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"compassion\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.5, 7.17, "compassion"]], "response_mode": "offline", "answer": "[{\"word\":\"compassion\",\"start\":6.5,\"end\":7.17}]"} {"utterance_id": "6455-67803-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0032.flac", "task_type": "lexical_detection", "query": "answered", "prompt": "When the utterance is complete, report all instances of the word \"answered\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.19, "answered"]], "response_mode": "offline", "answer": "[{\"word\":\"answered\",\"start\":0.85,\"end\":1.19}]"} {"utterance_id": "4831-18525-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.79, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":0.54,\"end\":0.79}]"} {"utterance_id": "6123-59150-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0019.flac", "task_type": "lexical_detection", "query": "invents", "prompt": "After listening to the whole recording, output every instance of the word \"invents\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 1.07, "invents"], [2.99, 3.37, "invents"]], "response_mode": "offline", "answer": "[{\"word\":\"invents\",\"start\":0.64,\"end\":1.07},{\"word\":\"invents\",\"start\":2.99,\"end\":3.37}]"} {"utterance_id": "7697-105815-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.55, 3.14, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":2.55,\"end\":3.14}]"} {"utterance_id": "4570-14911-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0011.flac", "task_type": "lexical_detection", "query": "hanging", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"hanging\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.12, 6.52, "hanging"]], "response_mode": "offline", "answer": "[{\"word\":\"hanging\",\"start\":6.12,\"end\":6.52}]"} {"utterance_id": "3915-98647-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0032.flac", "task_type": "lexical_detection", "query": "require", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"require\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.87, 6.32, "require"]], "response_mode": "offline", "answer": "[{\"word\":\"require\",\"start\":5.87,\"end\":6.32}]"} {"utterance_id": "1701-141760-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0022.flac", "task_type": "lexical_counting", "query": "way", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"way\" (word, start, end). If none, reply [].", "labels": [[9.62, 9.88, "one"], [14.03, 14.26, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.62,\"end\":9.88},{\"word\":\"two\",\"start\":14.03,\"end\":14.26}]"} {"utterance_id": "6841-88294-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0002.flac", "task_type": "lexical_detection", "query": "steel", "prompt": "Process the full audio offline. Find every instance of the word \"steel\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.32, "steel"]], "response_mode": "offline", "answer": "[{\"word\":\"steel\",\"start\":1.95,\"end\":2.32}]"} {"utterance_id": "6123-59186-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.52, 1.15, "mister hardy"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy\",\"start\":0.52,\"end\":1.15}]"} {"utterance_id": "6599-38591-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0008.flac", "task_type": "lexical_detection", "query": "chatting", "prompt": "After the audio ends, extract all occurrences of the word \"chatting\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.6, 3.9, "chatting"]], "response_mode": "offline", "answer": "[{\"word\":\"chatting\",\"start\":3.6,\"end\":3.9}]"} {"utterance_id": "4323-55228-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0000.flac", "task_type": "lexical_detection", "query": "starry", "prompt": "When the utterance is complete, report all instances of the word \"starry\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.85, "starry"]], "response_mode": "offline", "answer": "[{\"word\":\"starry\",\"start\":3.39,\"end\":3.85}]"} {"utterance_id": "4831-25894-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0033.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[0.81, 1.17, "one"], [4.51, 4.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.81,\"end\":1.17},{\"word\":\"two\",\"start\":4.51,\"end\":4.87}]"} {"utterance_id": "4153-186222-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.52, 4.01, "back"]], "response_mode": "offline", "answer": "[{\"word\":\"back\",\"start\":3.52,\"end\":4.01}]"} {"utterance_id": "1585-157660-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0010.flac", "task_type": "lexical_detection", "query": "amazement", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"amazement\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.09, "amazement"]], "response_mode": "offline", "answer": "[{\"word\":\"amazement\",\"start\":1.55,\"end\":2.09}]"} {"utterance_id": "6267-53049-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.21, 0.97, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.21,\"end\":0.97}]"} {"utterance_id": "8254-84205-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0045.flac", "task_type": "lexical_counting", "query": "talking", "prompt": "Listen to the full audio. Count every occurrence of the word \"talking\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.88, 3.33, "one"], [10.13, 10.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.88,\"end\":3.33},{\"word\":\"two\",\"start\":10.13,\"end\":10.46}]"} {"utterance_id": "8254-84205-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0061.flac", "task_type": "lexical_detection", "query": "driven", "prompt": "After the audio ends, extract all occurrences of the word \"driven\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.54, 2.85, "driven"]], "response_mode": "offline", "answer": "[{\"word\":\"driven\",\"start\":2.54,\"end\":2.85}]"} {"utterance_id": "8173-294714-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0013.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "When the utterance is complete, report how a family relation accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 2.01, "one"], [5.12, 5.58, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.64,\"end\":2.01},{\"word\":\"two\",\"start\":5.12,\"end\":5.58}]"} {"utterance_id": "7697-105815-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0011.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.56, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":1.37,\"end\":1.56}]"} {"utterance_id": "3660-172182-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.46, 0.78, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":0.46,\"end\":0.78}]"} {"utterance_id": "7697-245712-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0011.flac", "task_type": "lexical_detection", "query": "innocence", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"innocence\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.1, 3.62, "innocence"]], "response_mode": "offline", "answer": "[{\"word\":\"innocence\",\"start\":3.1,\"end\":3.62}]"} {"utterance_id": "7601-291468-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0002.flac", "task_type": "lexical_detection", "query": "desolation", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"desolation\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.08, 2.8, "desolation"]], "response_mode": "offline", "answer": "[{\"word\":\"desolation\",\"start\":2.08,\"end\":2.8}]"} {"utterance_id": "700-122866-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0025.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.12, 7.54, "sisters"]], "response_mode": "offline", "answer": "[{\"word\":\"sisters\",\"start\":7.12,\"end\":7.54}]"} {"utterance_id": "7697-105815-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0052.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.62, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":2.16,\"end\":2.62}]"} {"utterance_id": "3915-57461-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0001.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Once the audio has finished, identify every instance of a money or currency term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.38, 4.68, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":4.38,\"end\":4.68}]"} {"utterance_id": "7601-175351-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.51, 3.11, "johnson"]], "response_mode": "offline", "answer": "[{\"word\":\"johnson\",\"start\":2.51,\"end\":3.11}]"} {"utterance_id": "4323-13259-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0011.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen carefully to the entire clip, then report every instance of an organization name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.19, "company"]], "response_mode": "offline", "answer": "[{\"word\":\"company\",\"start\":1.83,\"end\":2.19}]"} {"utterance_id": "1701-141759-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.55, 2.0, "sonya"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":1.55,\"end\":2.0}]"} {"utterance_id": "3660-172182-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0021.flac", "task_type": "lexical_detection", "query": "place", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"place\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.29, 1.62, "place"]], "response_mode": "offline", "answer": "[{\"word\":\"place\",\"start\":1.29,\"end\":1.62}]"} {"utterance_id": "2506-11278-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0004.flac", "task_type": "lexical_detection", "query": "confessed", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"confessed\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.49, "confessed"]], "response_mode": "offline", "answer": "[{\"word\":\"confessed\",\"start\":2.7,\"end\":3.49}]"} {"utterance_id": "1686-142278-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.38, 0.88, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.38,\"end\":0.88}]"} {"utterance_id": "7697-105815-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0016.flac", "task_type": "lexical_detection", "query": "score", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"score\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.5, "score"]], "response_mode": "offline", "answer": "[{\"word\":\"score\",\"start\":3.1,\"end\":3.5}]"} {"utterance_id": "5849-50873-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0011.flac", "task_type": "lexical_detection", "query": "determination", "prompt": "After listening to the whole recording, output every instance of the word \"determination\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.47, 6.39, "determination"]], "response_mode": "offline", "answer": "[{\"word\":\"determination\",\"start\":5.47,\"end\":6.39}]"} {"utterance_id": "6455-67803-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0034.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.93, 2.43, "cheeks"], [3.97, 4.41, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"cheeks\",\"start\":1.93,\"end\":2.43},{\"word\":\"eyes\",\"start\":3.97,\"end\":4.41}]"} {"utterance_id": "1651-136854-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0022.flac", "task_type": "lexical_counting", "query": "against", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"against\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.8, 3.22, "one"], [4.72, 5.1, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.8,\"end\":3.22},{\"word\":\"two\",\"start\":4.72,\"end\":5.1}]"} {"utterance_id": "4831-25894-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0024.flac", "task_type": "lexical_detection", "query": "rose", "prompt": "After listening to the whole recording, output every instance of the word \"rose\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.58, "rose"]], "response_mode": "offline", "answer": "[{\"word\":\"rose\",\"start\":1.25,\"end\":1.58}]"} {"utterance_id": "6267-65525-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0050.flac", "task_type": "lexical_detection", "query": "please", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"please\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.72, 0.99, "please"]], "response_mode": "offline", "answer": "[{\"word\":\"please\",\"start\":0.72,\"end\":0.99}]"} {"utterance_id": "4515-11057-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0044.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.39, 9.81, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":9.39,\"end\":9.81}]"} {"utterance_id": "5849-50873-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0026.flac", "task_type": "lexical_detection", "query": "followed", "prompt": "After listening to the whole recording, output every instance of the word \"followed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.04, "followed"]], "response_mode": "offline", "answer": "[{\"word\":\"followed\",\"start\":0.7,\"end\":1.04}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 1.23, "captain bull"], [10.14, 10.46, "bull"]], "response_mode": "offline", "answer": "[{\"word\":\"captain bull\",\"start\":0.34,\"end\":1.23},{\"word\":\"bull\",\"start\":10.14,\"end\":10.46}]"} {"utterance_id": "4153-186222-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.65, 3.2, "silas"]], "response_mode": "offline", "answer": "[{\"word\":\"silas\",\"start\":2.65,\"end\":3.2}]"} {"utterance_id": "8288-274150-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0004.flac", "task_type": "lexical_detection", "query": "quickly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"quickly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.25, 4.6, "quickly"]], "response_mode": "offline", "answer": "[{\"word\":\"quickly\",\"start\":4.25,\"end\":4.6}]"} {"utterance_id": "6841-88294-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0039.flac", "task_type": "lexical_detection", "query": "knelt", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"knelt\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.38, "knelt"]], "response_mode": "offline", "answer": "[{\"word\":\"knelt\",\"start\":1.2,\"end\":1.38}]"} {"utterance_id": "700-122866-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0027.flac", "task_type": "lexical_detection", "query": "club", "prompt": "After listening to the whole recording, output every instance of the word \"club\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.77, "club"]], "response_mode": "offline", "answer": "[{\"word\":\"club\",\"start\":1.48,\"end\":1.77}]"} {"utterance_id": "1686-142278-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0046.flac", "task_type": "lexical_detection", "query": "sudden", "prompt": "Once the audio has finished, identify every instance of the word \"sudden\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.26, 1.69, "sudden"]], "response_mode": "offline", "answer": "[{\"word\":\"sudden\",\"start\":1.26,\"end\":1.69}]"} {"utterance_id": "8288-274150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0003.flac", "task_type": "lexical_detection", "query": "saint", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"saint\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.47, 2.84, "saint"]], "response_mode": "offline", "answer": "[{\"word\":\"saint\",\"start\":2.47,\"end\":2.84}]"} {"utterance_id": "1630-141772-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0015.flac", "task_type": "lexical_detection", "query": "icon", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"icon\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.11, 6.76, "icon"]], "response_mode": "offline", "answer": "[{\"word\":\"icon\",\"start\":6.11,\"end\":6.76}]"} {"utterance_id": "8173-294714-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0005.flac", "task_type": "lexical_detection", "query": "wrong", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wrong\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.12, "wrong"]], "response_mode": "offline", "answer": "[{\"word\":\"wrong\",\"start\":3.82,\"end\":4.12}]"} {"utterance_id": "5849-50962-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0021.flac", "task_type": "lexical_detection", "query": "comes", "prompt": "Process the full audio offline. Find every instance of the word \"comes\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.97, "comes"]], "response_mode": "offline", "answer": "[{\"word\":\"comes\",\"start\":0.68,\"end\":0.97}]"} {"utterance_id": "4831-25894-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0032.flac", "task_type": "lexical_counting", "query": "full", "prompt": "After listening to the whole recording, output timed count words for the word \"full\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[4.45, 4.67, "one"], [9.99, 10.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.45,\"end\":4.67},{\"word\":\"two\",\"start\":9.99,\"end\":10.28}]"} {"utterance_id": "4323-55228-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0047.flac", "task_type": "lexical_detection", "query": "wants", "prompt": "Process the full audio offline. Find every instance of the word \"wants\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.53, 2.77, "wants"]], "response_mode": "offline", "answer": "[{\"word\":\"wants\",\"start\":2.53,\"end\":2.77}]"} {"utterance_id": "700-122867-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0005.flac", "task_type": "lexical_detection", "query": "anxiously", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"anxiously\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.93, "anxiously"]], "response_mode": "offline", "answer": "[{\"word\":\"anxiously\",\"start\":1.3,\"end\":1.93}]"} {"utterance_id": "4831-25894-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0027.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.37, "brothers"]], "response_mode": "offline", "answer": "[{\"word\":\"brothers\",\"start\":0.98,\"end\":1.37}]"} {"utterance_id": "4323-55228-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0043.flac", "task_type": "lexical_detection", "query": "glow", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"glow\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.96, 1.18, "glow"]], "response_mode": "offline", "answer": "[{\"word\":\"glow\",\"start\":0.96,\"end\":1.18}]"} {"utterance_id": "7641-96684-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0029.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.85, 3.25, "hound"]], "response_mode": "offline", "answer": "[{\"word\":\"hound\",\"start\":2.85,\"end\":3.25}]"} {"utterance_id": "1650-167613-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0011.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of a country name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.92, 4.38, "french"]], "response_mode": "offline", "answer": "[{\"word\":\"french\",\"start\":3.92,\"end\":4.38}]"} {"utterance_id": "3660-172183-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0009.flac", "task_type": "lexical_detection", "query": "person", "prompt": "Process the full audio offline. Find every instance of the word \"person\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.31, "person"]], "response_mode": "offline", "answer": "[{\"word\":\"person\",\"start\":0.93,\"end\":1.31}]"} {"utterance_id": "8173-294714-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0019.flac", "task_type": "lexical_detection", "query": "forced", "prompt": "After the audio ends, extract all occurrences of the word \"forced\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.63, 3.97, "forced"]], "response_mode": "offline", "answer": "[{\"word\":\"forced\",\"start\":3.63,\"end\":3.97}]"} {"utterance_id": "116-288048-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0024.flac", "task_type": "lexical_detection", "query": "mythical", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mythical\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.69, 1.21, "mythical"]], "response_mode": "offline", "answer": "[{\"word\":\"mythical\",\"start\":0.69,\"end\":1.21}]"} {"utterance_id": "6467-94831-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0043.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.83, "master rover"]], "response_mode": "offline", "answer": "[{\"word\":\"master rover\",\"start\":3.21,\"end\":3.83}]"} {"utterance_id": "6467-56885-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0010.flac", "task_type": "lexical_detection", "query": "wedding", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wedding\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.3, 3.64, "wedding"]], "response_mode": "offline", "answer": "[{\"word\":\"wedding\",\"start\":3.3,\"end\":3.64}]"} {"utterance_id": "3660-172183-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0007.flac", "task_type": "lexical_detection", "query": "quest", "prompt": "Process the full audio offline. Find every instance of the word \"quest\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.43, 2.82, "quest"]], "response_mode": "offline", "answer": "[{\"word\":\"quest\",\"start\":2.43,\"end\":2.82}]"} {"utterance_id": "3660-6517-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0016.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 2.07, "terra nova"]], "response_mode": "offline", "answer": "[{\"word\":\"terra nova\",\"start\":1.54,\"end\":2.07}]"} {"utterance_id": "1585-131718-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0047.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.58, 5.35, "captain shaw"]], "response_mode": "offline", "answer": "[{\"word\":\"captain shaw\",\"start\":4.58,\"end\":5.35}]"} {"utterance_id": "1255-138279-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0017.flac", "task_type": "lexical_detection", "query": "forgive", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"forgive\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.96, 2.38, "forgive"]], "response_mode": "offline", "answer": "[{\"word\":\"forgive\",\"start\":1.96,\"end\":2.38}]"} {"utterance_id": "3660-172183-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0002.flac", "task_type": "lexical_detection", "query": "erbin", "prompt": "When the utterance is complete, report all instances of the word \"erbin\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 1.91, "erbin"]], "response_mode": "offline", "answer": "[{\"word\":\"erbin\",\"start\":1.5,\"end\":1.91}]"} {"utterance_id": "4831-18525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0015.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Once the audio has finished, identify every instance of a color name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.84, 8.17, "black"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":7.84,\"end\":8.17}]"} {"utterance_id": "6841-88294-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0027.flac", "task_type": "lexical_detection", "query": "robbed", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"robbed\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.38, "robbed"]], "response_mode": "offline", "answer": "[{\"word\":\"robbed\",\"start\":0.88,\"end\":1.38}]"} {"utterance_id": "1255-138279-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0024.flac", "task_type": "lexical_counting", "query": "good", "prompt": "Listen to the full audio. Count every occurrence of the word \"good\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.66, "one"], [1.9, 2.07, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":0.66},{\"word\":\"two\",\"start\":1.9,\"end\":2.07}]"} {"utterance_id": "5543-27761-0087", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0087.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.02, 1.62, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":1.02,\"end\":1.62}]"} {"utterance_id": "700-122868-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0018.flac", "task_type": "lexical_detection", "query": "bridge", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"bridge\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 1.86, "bridge"]], "response_mode": "offline", "answer": "[{\"word\":\"bridge\",\"start\":1.42,\"end\":1.86}]"} {"utterance_id": "6455-66379-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0012.flac", "task_type": "lexical_detection", "query": "deal", "prompt": "When the utterance is complete, report all instances of the word \"deal\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.17, 0.41, "deal"]], "response_mode": "offline", "answer": "[{\"word\":\"deal\",\"start\":0.17,\"end\":0.41}]"} {"utterance_id": "4570-14911-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.97, "ears"], [8.54, 8.88, "cheek"]], "response_mode": "offline", "answer": "[{\"word\":\"ears\",\"start\":0.54,\"end\":0.97},{\"word\":\"cheek\",\"start\":8.54,\"end\":8.88}]"} {"utterance_id": "1255-90413-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0016.flac", "task_type": "lexical_detection", "query": "punishment", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"punishment\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.38, "punishment"]], "response_mode": "offline", "answer": "[{\"word\":\"punishment\",\"start\":0.77,\"end\":1.38}]"} {"utterance_id": "4831-18525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0030.flac", "task_type": "lexical_detection", "query": "venture", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"venture\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.02, 3.38, "venture"]], "response_mode": "offline", "answer": "[{\"word\":\"venture\",\"start\":3.02,\"end\":3.38}]"} {"utterance_id": "6841-88291-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0024.flac", "task_type": "lexical_detection", "query": "brand", "prompt": "Once the audio has finished, identify every instance of the word \"brand\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.26, 0.58, "brand"]], "response_mode": "offline", "answer": "[{\"word\":\"brand\",\"start\":0.26,\"end\":0.58}]"} {"utterance_id": "1651-136854-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.42, 0.94, "scholar's"]], "response_mode": "offline", "answer": "[{\"word\":\"scholar's\",\"start\":0.42,\"end\":0.94}]"} {"utterance_id": "6455-67804-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0040.flac", "task_type": "lexical_detection", "query": "brought", "prompt": "Process the full audio offline. Find every instance of the word \"brought\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.05, "brought"]], "response_mode": "offline", "answer": "[{\"word\":\"brought\",\"start\":0.87,\"end\":1.05}]"} {"utterance_id": "8173-294714-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0012.flac", "task_type": "lexical_detection", "query": "knows", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"knows\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.47, "knows"]], "response_mode": "offline", "answer": "[{\"word\":\"knows\",\"start\":2.03,\"end\":2.47}]"} {"utterance_id": "4323-55228-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0015.flac", "task_type": "lexical_detection", "query": "dawned", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"dawned\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.75, 2.05, "dawned"]], "response_mode": "offline", "answer": "[{\"word\":\"dawned\",\"start\":1.75,\"end\":2.05}]"} {"utterance_id": "8173-294714-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0042.flac", "task_type": "lexical_detection", "query": "returned", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"returned\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.6, 2.08, "returned"]], "response_mode": "offline", "answer": "[{\"word\":\"returned\",\"start\":1.6,\"end\":2.08}]"} {"utterance_id": "7641-96684-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.51, "admetus"], [2.59, 3.44, "alcestis"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":0.94,\"end\":1.51},{\"word\":\"alcestis\",\"start\":2.59,\"end\":3.44}]"} {"utterance_id": "1701-141759-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.48, 0.96, "sonya"], [2.9, 3.5, "nicholas"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":0.48,\"end\":0.96},{\"word\":\"nicholas\",\"start\":2.9,\"end\":3.5}]"} {"utterance_id": "5849-50873-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0027.flac", "task_type": "lexical_detection", "query": "youth", "prompt": "After the audio ends, extract all occurrences of the word \"youth\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.92, "youth"]], "response_mode": "offline", "answer": "[{\"word\":\"youth\",\"start\":0.54,\"end\":0.92}]"} {"utterance_id": "3915-57461-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0007.flac", "task_type": "lexical_detection", "query": "passengers", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"passengers\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.01, 5.57, "passengers"]], "response_mode": "offline", "answer": "[{\"word\":\"passengers\",\"start\":5.01,\"end\":5.57}]"} {"utterance_id": "7641-96684-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0003.flac", "task_type": "lexical_detection", "query": "alcestis", "prompt": "Once the audio has finished, identify every instance of the word \"alcestis\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.59, 3.44, "alcestis"]], "response_mode": "offline", "answer": "[{\"word\":\"alcestis\",\"start\":2.59,\"end\":3.44}]"} {"utterance_id": "8254-115543-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0025.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.72, 2.17, "elephant"], [3.26, 3.65, "beast"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant\",\"start\":1.72,\"end\":2.17},{\"word\":\"beast\",\"start\":3.26,\"end\":3.65}]"} {"utterance_id": "4323-18416-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0033.flac", "task_type": "lexical_detection", "query": "chair", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"chair\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.89, 3.13, "chair"]], "response_mode": "offline", "answer": "[{\"word\":\"chair\",\"start\":2.89,\"end\":3.13}]"} {"utterance_id": "7697-105815-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0002.flac", "task_type": "lexical_detection", "query": "zion", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"zion\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.8, 7.45, "zion"]], "response_mode": "offline", "answer": "[{\"word\":\"zion\",\"start\":6.8,\"end\":7.45}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.02, 10.35, "dinner"]], "response_mode": "offline", "answer": "[{\"word\":\"dinner\",\"start\":10.02,\"end\":10.35}]"} {"utterance_id": "3660-6517-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0020.flac", "task_type": "lexical_detection", "query": "perforce", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"perforce\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.84, 1.32, "perforce"]], "response_mode": "offline", "answer": "[{\"word\":\"perforce\",\"start\":0.84,\"end\":1.32}]"} {"utterance_id": "700-122866-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0030.flac", "task_type": "lexical_detection", "query": "existence", "prompt": "After the audio ends, extract all occurrences of the word \"existence\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.04, 2.85, "existence"]], "response_mode": "offline", "answer": "[{\"word\":\"existence\",\"start\":2.04,\"end\":2.85}]"} {"utterance_id": "4570-24733-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0004.flac", "task_type": "lexical_counting", "query": "constituting", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"constituting\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[4.47, 5.17, "one"], [20.05, 21.03, "two"], [23.09, 23.99, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.47,\"end\":5.17},{\"word\":\"two\",\"start\":20.05,\"end\":21.03},{\"word\":\"three\",\"start\":23.09,\"end\":23.99}]"} {"utterance_id": "6267-65525-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0023.flac", "task_type": "lexical_counting", "query": "love", "prompt": "Once the audio has finished, list each occurrence of the word \"love\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.37, 0.63, "one"], [4.04, 4.47, "two"], [5.99, 6.22, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.37,\"end\":0.63},{\"word\":\"two\",\"start\":4.04,\"end\":4.47},{\"word\":\"three\",\"start\":5.99,\"end\":6.22}]"} {"utterance_id": "8254-84205-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0044.flac", "task_type": "lexical_detection", "query": "game", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"game\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 3.93, "game"]], "response_mode": "offline", "answer": "[{\"word\":\"game\",\"start\":3.56,\"end\":3.93}]"} {"utterance_id": "8254-84205-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0066.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.55, 4.8, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":4.55,\"end\":4.8}]"} {"utterance_id": "4831-25894-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0026.flac", "task_type": "lexical_detection", "query": "gladness", "prompt": "When the utterance is complete, report all instances of the word \"gladness\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.52, "gladness"]], "response_mode": "offline", "answer": "[{\"word\":\"gladness\",\"start\":2.03,\"end\":2.52}]"} {"utterance_id": "8288-274162-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.69, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":0.31,\"end\":0.69}]"} {"utterance_id": "4570-24733-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0004.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[5.94, 6.51, "one"], [9.56, 10.12, "two"], [24.85, 25.23, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.94,\"end\":6.51},{\"word\":\"two\",\"start\":9.56,\"end\":10.12},{\"word\":\"three\",\"start\":24.85,\"end\":25.23}]"} {"utterance_id": "3660-172183-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0017.flac", "task_type": "lexical_counting", "query": "sleep", "prompt": "Once the audio has finished, list each occurrence of the word \"sleep\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.26, 2.64, "one"], [5.93, 6.26, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.26,\"end\":2.64},{\"word\":\"two\",\"start\":5.93,\"end\":6.26}]"} {"utterance_id": "6267-65525-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0051.flac", "task_type": "lexical_detection", "query": "thought", "prompt": "When the utterance is complete, report all instances of the word \"thought\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.25, 2.73, "thought"]], "response_mode": "offline", "answer": "[{\"word\":\"thought\",\"start\":2.25,\"end\":2.73}]"} {"utterance_id": "5849-50964-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0010.flac", "task_type": "lexical_detection", "query": "vacation", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"vacation\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 2.17, "vacation"]], "response_mode": "offline", "answer": "[{\"word\":\"vacation\",\"start\":1.67,\"end\":2.17}]"} {"utterance_id": "1650-167613-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0034.flac", "task_type": "lexical_detection", "query": "opinion", "prompt": "After listening to the whole recording, output every instance of the word \"opinion\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.27, "opinion"]], "response_mode": "offline", "answer": "[{\"word\":\"opinion\",\"start\":0.69,\"end\":1.27}]"} {"utterance_id": "6841-88291-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0023.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 0.85, "calf"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":0.59,\"end\":0.85}]"} {"utterance_id": "6455-67803-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0024.flac", "task_type": "lexical_detection", "query": "seen", "prompt": "Once the audio has finished, identify every instance of the word \"seen\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.5, 2.72, "seen"]], "response_mode": "offline", "answer": "[{\"word\":\"seen\",\"start\":2.5,\"end\":2.72}]"} {"utterance_id": "6841-88291-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0015.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.18, 5.48, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":5.18,\"end\":5.48}]"} {"utterance_id": "1630-141772-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 1.39, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":0.49,\"end\":1.39}]"} {"utterance_id": "5849-50962-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0017.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.3, 6.68, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"daughter\",\"start\":6.3,\"end\":6.68}]"} {"utterance_id": "4153-186222-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0002.flac", "task_type": "lexical_detection", "query": "hesitated", "prompt": "After the audio ends, extract all occurrences of the word \"hesitated\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.5, "hesitated"]], "response_mode": "offline", "answer": "[{\"word\":\"hesitated\",\"start\":0.78,\"end\":1.5}]"} {"utterance_id": "6467-94831-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0016.flac", "task_type": "lexical_detection", "query": "sit", "prompt": "Process the full audio offline. Find every instance of the word \"sit\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 0.83, "sit"]], "response_mode": "offline", "answer": "[{\"word\":\"sit\",\"start\":0.65,\"end\":0.83}]"} {"utterance_id": "116-288047-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0009.flac", "task_type": "lexical_detection", "query": "reflection", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"reflection\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.33, "reflection"]], "response_mode": "offline", "answer": "[{\"word\":\"reflection\",\"start\":0.72,\"end\":1.33}]"} {"utterance_id": "700-122866-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0025.flac", "task_type": "lexical_detection", "query": "ruby", "prompt": "Process the full audio offline. Find every instance of the word \"ruby\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.06, "ruby"]], "response_mode": "offline", "answer": "[{\"word\":\"ruby\",\"start\":0.87,\"end\":1.06}]"} {"utterance_id": "6467-94831-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 3.98, "several"]], "response_mode": "offline", "answer": "[{\"word\":\"several\",\"start\":3.56,\"end\":3.98}]"} {"utterance_id": "4515-11057-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0019.flac", "task_type": "lexical_detection", "query": "converse", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"converse\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 2.2, "converse"]], "response_mode": "offline", "answer": "[{\"word\":\"converse\",\"start\":1.67,\"end\":2.2}]"} {"utterance_id": "6267-65525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.23, 1.02, "john lincoln"], [11.84, 12.32, "sidney"], [19.09, 19.67, "sidney's"]], "response_mode": "offline", "answer": "[{\"word\":\"john lincoln\",\"start\":0.23,\"end\":1.02},{\"word\":\"sidney\",\"start\":11.84,\"end\":12.32},{\"word\":\"sidney's\",\"start\":19.09,\"end\":19.67}]"} {"utterance_id": "4153-185072-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0006.flac", "task_type": "lexical_detection", "query": "allen", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"allen\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.9, 4.39, "allen"]], "response_mode": "offline", "answer": "[{\"word\":\"allen\",\"start\":3.9,\"end\":4.39}]"} {"utterance_id": "4831-25894-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0025.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After the audio ends, extract all occurrences of a color name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.47, 7.81, "rosy"]], "response_mode": "offline", "answer": "[{\"word\":\"rosy\",\"start\":7.47,\"end\":7.81}]"} {"utterance_id": "1630-96099-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0023.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[3.78, 4.19, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":3.78,\"end\":4.19}]"} {"utterance_id": "1701-141760-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0044.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.76, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":0.5,\"end\":0.76}]"} {"utterance_id": "4831-18525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.33, 5.79, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":5.33,\"end\":5.79}]"} {"utterance_id": "4153-186222-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0035.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.44, "dozen"]], "response_mode": "offline", "answer": "[{\"word\":\"dozen\",\"start\":1.08,\"end\":1.44}]"} {"utterance_id": "116-288046-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.13, 5.62, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":5.13,\"end\":5.62}]"} {"utterance_id": "1585-131718-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.33, "twenty"], [8.92, 9.63, "sixty seven"]], "response_mode": "offline", "answer": "[{\"word\":\"twenty\",\"start\":1.06,\"end\":1.33},{\"word\":\"sixty seven\",\"start\":8.92,\"end\":9.63}]"} {"utterance_id": "4572-112381-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0012.flac", "task_type": "lexical_detection", "query": "weal", "prompt": "After the audio ends, extract all occurrences of the word \"weal\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.75, 4.2, "weal"]], "response_mode": "offline", "answer": "[{\"word\":\"weal\",\"start\":3.75,\"end\":4.2}]"} {"utterance_id": "6467-97061-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0010.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "After listening to the whole recording, output timed count words for an animal name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.4, 2.86, "one"], [13.65, 14.08, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.4,\"end\":2.86},{\"word\":\"two\",\"start\":13.65,\"end\":14.08}]"} {"utterance_id": "1701-141760-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0026.flac", "task_type": "lexical_detection", "query": "thrown", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"thrown\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.22, "thrown"]], "response_mode": "offline", "answer": "[{\"word\":\"thrown\",\"start\":0.88,\"end\":1.22}]"} {"utterance_id": "8254-84205-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0055.flac", "task_type": "lexical_detection", "query": "wait", "prompt": "After the audio ends, extract all occurrences of the word \"wait\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.51, "wait"]], "response_mode": "offline", "answer": "[{\"word\":\"wait\",\"start\":1.21,\"end\":1.51}]"} {"utterance_id": "1630-141772-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0002.flac", "task_type": "lexical_detection", "query": "first", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"first\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.19, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.65,\"end\":1.19}]"} {"utterance_id": "8254-115543-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0022.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.15, "elephant"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant\",\"start\":1.55,\"end\":2.15}]"} {"utterance_id": "3663-172528-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0020.flac", "task_type": "lexical_detection", "query": "permit", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"permit\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.98, 1.41, "permit"]], "response_mode": "offline", "answer": "[{\"word\":\"permit\",\"start\":0.98,\"end\":1.41}]"} {"utterance_id": "5543-27761-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0061.flac", "task_type": "lexical_detection", "query": "glance", "prompt": "Process the full audio offline. Find every instance of the word \"glance\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.91, 2.41, "glance"]], "response_mode": "offline", "answer": "[{\"word\":\"glance\",\"start\":1.91,\"end\":2.41}]"} {"utterance_id": "4831-25894-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0011.flac", "task_type": "lexical_detection", "query": "clung", "prompt": "After the audio ends, extract all occurrences of the word \"clung\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 2.95, "clung"]], "response_mode": "offline", "answer": "[{\"word\":\"clung\",\"start\":2.69,\"end\":2.95}]"} {"utterance_id": "7641-96252-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0018.flac", "task_type": "lexical_counting", "query": "sympathy", "prompt": "When the utterance is complete, report how the word \"sympathy\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 2.19, "one"], [3.39, 3.89, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.51,\"end\":2.19},{\"word\":\"two\",\"start\":3.39,\"end\":3.89}]"} {"utterance_id": "5849-50873-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0028.flac", "task_type": "lexical_detection", "query": "care", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"care\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.03, "care"]], "response_mode": "offline", "answer": "[{\"word\":\"care\",\"start\":0.77,\"end\":1.03}]"} {"utterance_id": "4323-55228-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0043.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "After the audio ends, tally every instance of a body part. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.64, 1.86, "one"], [3.81, 3.99, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.64,\"end\":1.86},{\"word\":\"two\",\"start\":3.81,\"end\":3.99}]"} {"utterance_id": "1255-90413-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0022.flac", "task_type": "lexical_counting", "query": "room", "prompt": "Process the full audio offline. For every occurrence of the word \"room\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 1.95, "one"], [4.07, 4.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.81,\"end\":1.95},{\"word\":\"two\",\"start\":4.07,\"end\":4.28}]"} {"utterance_id": "1686-142278-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0051.flac", "task_type": "lexical_detection", "query": "shocked", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"shocked\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.71, 3.23, "shocked"]], "response_mode": "offline", "answer": "[{\"word\":\"shocked\",\"start\":2.71,\"end\":3.23}]"} {"utterance_id": "4153-61735-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0005.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Do not answer until the audio ends. Then list each instance of an animal name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.47, 3.87, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":3.47,\"end\":3.87}]"} {"utterance_id": "7697-105815-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0027.flac", "task_type": "lexical_detection", "query": "scroll", "prompt": "When the utterance is complete, report all instances of the word \"scroll\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.1, 7.48, "scroll"]], "response_mode": "offline", "answer": "[{\"word\":\"scroll\",\"start\":7.1,\"end\":7.48}]"} {"utterance_id": "6467-94831-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0004.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.86, 1.38, "one"], [4.93, 5.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.86,\"end\":1.38},{\"word\":\"two\",\"start\":4.93,\"end\":5.63}]"} {"utterance_id": "6123-59186-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0022.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.28, 6.46, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.28,\"end\":6.46}]"} {"utterance_id": "8173-294714-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0026.flac", "task_type": "lexical_detection", "query": "uttered", "prompt": "After listening to the whole recording, output every instance of the word \"uttered\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.51, "uttered"]], "response_mode": "offline", "answer": "[{\"word\":\"uttered\",\"start\":1.15,\"end\":1.51}]"} {"utterance_id": "6123-59150-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0015.flac", "task_type": "lexical_detection", "query": "belongs", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"belongs\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.33, "belongs"]], "response_mode": "offline", "answer": "[{\"word\":\"belongs\",\"start\":1.98,\"end\":2.33}]"} {"utterance_id": "7697-105815-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0018.flac", "task_type": "lexical_detection", "query": "pray", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"pray\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.68, 3.13, "pray"]], "response_mode": "offline", "answer": "[{\"word\":\"pray\",\"start\":2.68,\"end\":3.13}]"} {"utterance_id": "4153-186222-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0034.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.02, 4.37, "second"]], "response_mode": "offline", "answer": "[{\"word\":\"second\",\"start\":4.02,\"end\":4.37}]"} {"utterance_id": "3663-172005-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.25, 3.75, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":3.25,\"end\":3.75}]"} {"utterance_id": "4323-18416-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0034.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Do not answer until the audio ends. Then list each instance of an animal name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.99, 1.27, "dog"]], "response_mode": "offline", "answer": "[{\"word\":\"dog\",\"start\":0.99,\"end\":1.27}]"} {"utterance_id": "3915-57461-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0006.flac", "task_type": "lexical_detection", "query": "misfortune", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"misfortune\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.75, 2.43, "misfortune"]], "response_mode": "offline", "answer": "[{\"word\":\"misfortune\",\"start\":1.75,\"end\":2.43}]"} {"utterance_id": "116-288048-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0006.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Process the full audio offline. Find every instance of a date and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 1.07, "good friday"]], "response_mode": "offline", "answer": "[{\"word\":\"good friday\",\"start\":0.31,\"end\":1.07}]"} {"utterance_id": "4323-13259-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0012.flac", "task_type": "lexical_detection", "query": "deliverance", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"deliverance\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.73, 3.32, "deliverance"]], "response_mode": "offline", "answer": "[{\"word\":\"deliverance\",\"start\":2.73,\"end\":3.32}]"} {"utterance_id": "5849-50962-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[2.67, 2.91, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":2.67,\"end\":2.91}]"} {"utterance_id": "5543-27761-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0070.flac", "task_type": "lexical_detection", "query": "blanket", "prompt": "After the audio ends, extract all occurrences of the word \"blanket\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.3, "blanket"]], "response_mode": "offline", "answer": "[{\"word\":\"blanket\",\"start\":1.77,\"end\":2.3}]"} {"utterance_id": "700-122867-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0031.flac", "task_type": "lexical_detection", "query": "expect", "prompt": "After the audio ends, extract all occurrences of the word \"expect\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.93, "expect"]], "response_mode": "offline", "answer": "[{\"word\":\"expect\",\"start\":0.5,\"end\":0.93}]"} {"utterance_id": "116-288048-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.17, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":2.69,\"end\":3.17}]"} {"utterance_id": "8288-274162-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0035.flac", "task_type": "lexical_detection", "query": "information", "prompt": "After listening to the whole recording, output every instance of the word \"information\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.7, 9.35, "information"]], "response_mode": "offline", "answer": "[{\"word\":\"information\",\"start\":8.7,\"end\":9.35}]"} {"utterance_id": "3663-172528-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0045.flac", "task_type": "lexical_detection", "query": "medal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"medal\" with fields word, start, end (seconds). If none, reply [].", "labels": [[10.93, 11.34, "medal"]], "response_mode": "offline", "answer": "[{\"word\":\"medal\",\"start\":10.93,\"end\":11.34}]"} {"utterance_id": "116-288045-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0013.flac", "task_type": "lexical_detection", "query": "inquired", "prompt": "After the audio ends, extract all occurrences of the word \"inquired\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.88, "inquired"]], "response_mode": "offline", "answer": "[{\"word\":\"inquired\",\"start\":0.34,\"end\":0.88}]"} {"utterance_id": "7601-101619-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0002.flac", "task_type": "lexical_detection", "query": "confound", "prompt": "Once the audio has finished, identify every instance of the word \"confound\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.14, 6.07, "confound"]], "response_mode": "offline", "answer": "[{\"word\":\"confound\",\"start\":5.14,\"end\":6.07}]"} {"utterance_id": "8288-274162-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0006.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After the audio ends, extract all occurrences of a color name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.58, "pale"]], "response_mode": "offline", "answer": "[{\"word\":\"pale\",\"start\":1.23,\"end\":1.58}]"} {"utterance_id": "6841-88291-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0014.flac", "task_type": "lexical_detection", "query": "paid", "prompt": "Process the full audio offline. Find every instance of the word \"paid\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.93, "paid"]], "response_mode": "offline", "answer": "[{\"word\":\"paid\",\"start\":0.66,\"end\":0.93}]"} {"utterance_id": "6455-67804-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0025.flac", "task_type": "lexical_detection", "query": "belinda", "prompt": "After listening to the whole recording, output every instance of the word \"belinda\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.65, 3.11, "belinda"]], "response_mode": "offline", "answer": "[{\"word\":\"belinda\",\"start\":2.65,\"end\":3.11}]"} {"utterance_id": "6841-88294-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.49, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.22,\"end\":1.49}]"} {"utterance_id": "1686-142278-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0016.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.3, "minister"]], "response_mode": "offline", "answer": "[{\"word\":\"minister\",\"start\":1.75,\"end\":2.3}]"} {"utterance_id": "6841-88291-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0013.flac", "task_type": "lexical_detection", "query": "wrapped", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"wrapped\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.63, 0.87, "wrapped"]], "response_mode": "offline", "answer": "[{\"word\":\"wrapped\",\"start\":0.63,\"end\":0.87}]"} {"utterance_id": "3915-57461-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0027.flac", "task_type": "lexical_detection", "query": "overtake", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"overtake\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[10.27, 10.78, "overtake"]], "response_mode": "offline", "answer": "[{\"word\":\"overtake\",\"start\":10.27,\"end\":10.78}]"} {"utterance_id": "2506-11278-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0022.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "When the utterance is complete, report all instances of a plant name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.69, 4.12, "bamboo"]], "response_mode": "offline", "answer": "[{\"word\":\"bamboo\",\"start\":3.69,\"end\":4.12}]"} {"utterance_id": "1686-142278-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0063.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.89, 4.38, "papa"]], "response_mode": "offline", "answer": "[{\"word\":\"papa\",\"start\":3.89,\"end\":4.38}]"} {"utterance_id": "1686-142278-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0061.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 1.79, "head"], [4.01, 4.38, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.63,\"end\":1.79},{\"word\":\"hand\",\"start\":4.01,\"end\":4.38}]"} {"utterance_id": "6267-65525-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0054.flac", "task_type": "lexical_detection", "query": "unconsciously", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"unconsciously\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.71, 5.76, "unconsciously"]], "response_mode": "offline", "answer": "[{\"word\":\"unconsciously\",\"start\":4.71,\"end\":5.76}]"} {"utterance_id": "5543-27761-0090", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0090.flac", "task_type": "lexical_detection", "query": "resign", "prompt": "After the audio ends, extract all occurrences of the word \"resign\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.23, "resign"]], "response_mode": "offline", "answer": "[{\"word\":\"resign\",\"start\":0.81,\"end\":1.23}]"} {"utterance_id": "3915-57461-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0014.flac", "task_type": "lexical_detection", "query": "stretched", "prompt": "When the utterance is complete, report all instances of the word \"stretched\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.55, 3.04, "stretched"]], "response_mode": "offline", "answer": "[{\"word\":\"stretched\",\"start\":2.55,\"end\":3.04}]"} {"utterance_id": "5543-27761-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0039.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.74, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":2.39,\"end\":2.74}]"} {"utterance_id": "8288-274162-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0066.flac", "task_type": "lexical_detection", "query": "tone", "prompt": "When the utterance is complete, report all instances of the word \"tone\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 1.93, "tone"]], "response_mode": "offline", "answer": "[{\"word\":\"tone\",\"start\":1.67,\"end\":1.93}]"} {"utterance_id": "1701-141759-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0029.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.03, 4.52, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":4.03,\"end\":4.52}]"} {"utterance_id": "4153-185072-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0015.flac", "task_type": "lexical_detection", "query": "secret", "prompt": "When the utterance is complete, report all instances of the word \"secret\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.23, "secret"]], "response_mode": "offline", "answer": "[{\"word\":\"secret\",\"start\":2.75,\"end\":3.23}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.51, 4.98, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.51,\"end\":4.98}]"} {"utterance_id": "700-122867-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0008.flac", "task_type": "lexical_detection", "query": "minute", "prompt": "Process the full audio offline. Find every instance of the word \"minute\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.88, "minute"]], "response_mode": "offline", "answer": "[{\"word\":\"minute\",\"start\":0.52,\"end\":0.88}]"} {"utterance_id": "4515-11057-0072", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0072.flac", "task_type": "lexical_detection", "query": "intended", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"intended\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.04, 2.79, "intended"]], "response_mode": "offline", "answer": "[{\"word\":\"intended\",\"start\":2.04,\"end\":2.79}]"} {"utterance_id": "1686-142278-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0005.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.81, "tea"]], "response_mode": "offline", "answer": "[{\"word\":\"tea\",\"start\":1.59,\"end\":1.81}]"} {"utterance_id": "700-122868-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0012.flac", "task_type": "lexical_detection", "query": "dead", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"dead\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.84, 2.12, "dead"]], "response_mode": "offline", "answer": "[{\"word\":\"dead\",\"start\":1.84,\"end\":2.12}]"} {"utterance_id": "7641-96670-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 0.7, "eetes"]], "response_mode": "offline", "answer": "[{\"word\":\"eetes\",\"start\":0.61,\"end\":0.7}]"} {"utterance_id": "4323-18416-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0018.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.19, 3.58, "one"], [8.49, 8.85, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.19,\"end\":3.58},{\"word\":\"two\",\"start\":8.49,\"end\":8.85}]"} {"utterance_id": "6267-65525-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0018.flac", "task_type": "lexical_detection", "query": "liked", "prompt": "After listening to the whole recording, output every instance of the word \"liked\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.31, "liked"]], "response_mode": "offline", "answer": "[{\"word\":\"liked\",\"start\":0.96,\"end\":1.31}]"} {"utterance_id": "3663-172528-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0030.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.65, "painter"]], "response_mode": "offline", "answer": "[{\"word\":\"painter\",\"start\":0.94,\"end\":1.65}]"} {"utterance_id": "1255-90413-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0028.flac", "task_type": "lexical_detection", "query": "father's", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"father's\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.37, 2.79, "father's"]], "response_mode": "offline", "answer": "[{\"word\":\"father's\",\"start\":2.37,\"end\":2.79}]"} {"utterance_id": "1686-142278-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0029.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.69, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.46,\"end\":0.69}]"} {"utterance_id": "7697-105815-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0019.flac", "task_type": "lexical_detection", "query": "case", "prompt": "Once the audio has finished, identify every instance of the word \"case\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.74, 1.12, "case"]], "response_mode": "offline", "answer": "[{\"word\":\"case\",\"start\":0.74,\"end\":1.12}]"} {"utterance_id": "6841-88294-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0030.flac", "task_type": "lexical_detection", "query": "wouldn't", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wouldn't\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.99, 4.2, "wouldn't"]], "response_mode": "offline", "answer": "[{\"word\":\"wouldn't\",\"start\":3.99,\"end\":4.2}]"} {"utterance_id": "4515-11057-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0040.flac", "task_type": "lexical_detection", "query": "scribbling", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"scribbling\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.53, "scribbling"]], "response_mode": "offline", "answer": "[{\"word\":\"scribbling\",\"start\":1.98,\"end\":2.53}]"} {"utterance_id": "8173-294714-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0045.flac", "task_type": "lexical_detection", "query": "prevent", "prompt": "When the utterance is complete, report all instances of the word \"prevent\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.41, "prevent"]], "response_mode": "offline", "answer": "[{\"word\":\"prevent\",\"start\":1.0,\"end\":1.41}]"} {"utterance_id": "8288-274162-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0042.flac", "task_type": "lexical_detection", "query": "disposed", "prompt": "After listening to the whole recording, output every instance of the word \"disposed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.66, 3.19, "disposed"]], "response_mode": "offline", "answer": "[{\"word\":\"disposed\",\"start\":2.66,\"end\":3.19}]"} {"utterance_id": "4570-102353-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0004.flac", "task_type": "lexical_detection", "query": "chosen", "prompt": "When the utterance is complete, report all instances of the word \"chosen\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.45, 3.0, "chosen"]], "response_mode": "offline", "answer": "[{\"word\":\"chosen\",\"start\":2.45,\"end\":3.0}]"} {"utterance_id": "4572-112375-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0001.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.08, 5.55, "golden"]], "response_mode": "offline", "answer": "[{\"word\":\"golden\",\"start\":5.08,\"end\":5.55}]"} {"utterance_id": "6123-59150-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0006.flac", "task_type": "lexical_counting", "query": "lust", "prompt": "After the audio ends, tally every instance of the word \"lust\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.84, 1.25, "one"], [3.09, 3.41, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.84,\"end\":1.25},{\"word\":\"two\",\"start\":3.09,\"end\":3.41}]"} {"utterance_id": "8288-274162-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0061.flac", "task_type": "lexical_detection", "query": "distress", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"distress\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.24, 0.84, "distress"]], "response_mode": "offline", "answer": "[{\"word\":\"distress\",\"start\":0.24,\"end\":0.84}]"} {"utterance_id": "8254-115543-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0010.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a country name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.61, 3.05, "india"]], "response_mode": "offline", "answer": "[{\"word\":\"india\",\"start\":2.61,\"end\":3.05}]"} {"utterance_id": "8254-115543-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0024.flac", "task_type": "lexical_detection", "query": "servant", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"servant\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.72, 2.06, "servant"]], "response_mode": "offline", "answer": "[{\"word\":\"servant\",\"start\":1.72,\"end\":2.06}]"} {"utterance_id": "1701-141760-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0025.flac", "task_type": "lexical_detection", "query": "acquaintance", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"acquaintance\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.21, 9.89, "acquaintance"]], "response_mode": "offline", "answer": "[{\"word\":\"acquaintance\",\"start\":9.21,\"end\":9.89}]"} {"utterance_id": "6455-67803-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.37, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":1.8,\"end\":2.37}]"} {"utterance_id": "1686-142278-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.89, 9.24, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":8.89,\"end\":9.24}]"} {"utterance_id": "3660-172182-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0004.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Process the full audio offline. Find every instance of a time of day and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.25, "midday"]], "response_mode": "offline", "answer": "[{\"word\":\"midday\",\"start\":0.84,\"end\":1.25}]"} {"utterance_id": "5849-50873-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0039.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Listen carefully to the entire clip, then report every instance of a state or province name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.41, 5.99, "rhode island"]], "response_mode": "offline", "answer": "[{\"word\":\"rhode island\",\"start\":5.41,\"end\":5.99}]"} {"utterance_id": "1650-167613-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0001.flac", "task_type": "lexical_detection", "query": "tea", "prompt": "Process the full audio offline. Find every instance of the word \"tea\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.43, "tea"]], "response_mode": "offline", "answer": "[{\"word\":\"tea\",\"start\":2.1,\"end\":2.43}]"} {"utterance_id": "3663-172528-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0047.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 4.05, "lorenzino"]], "response_mode": "offline", "answer": "[{\"word\":\"lorenzino\",\"start\":2.91,\"end\":4.05}]"} {"utterance_id": "4153-185072-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0005.flac", "task_type": "lexical_counting", "query": "came", "prompt": "After the audio ends, tally every instance of the word \"came\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[3.21, 3.48, "one"], [14.1, 14.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.21,\"end\":3.48},{\"word\":\"two\",\"start\":14.1,\"end\":14.42}]"} {"utterance_id": "7641-96670-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0009.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "When the utterance is complete, report all instances of a weather term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.24, 3.63, "winds"]], "response_mode": "offline", "answer": "[{\"word\":\"winds\",\"start\":3.24,\"end\":3.63}]"} {"utterance_id": "6267-65525-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0010.flac", "task_type": "lexical_detection", "query": "hated", "prompt": "When the utterance is complete, report all instances of the word \"hated\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.21, "hated"]], "response_mode": "offline", "answer": "[{\"word\":\"hated\",\"start\":0.8,\"end\":1.21}]"} {"utterance_id": "1585-131718-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0047.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen carefully to the entire clip, then report every instance of a city name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.34, 6.69, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":6.34,\"end\":6.69}]"} {"utterance_id": "1630-141772-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.79, 1.14, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":0.79,\"end\":1.14}]"} {"utterance_id": "1701-141759-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0033.flac", "task_type": "lexical_counting", "query": "always", "prompt": "Process the full audio offline. For every occurrence of the word \"always\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.79, "one"], [3.41, 3.89, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.39,\"end\":0.79},{\"word\":\"two\",\"start\":3.41,\"end\":3.89}]"} {"utterance_id": "1255-74899-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0009.flac", "task_type": "lexical_detection", "query": "poles", "prompt": "Once the audio has finished, identify every instance of the word \"poles\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.68, 1.07, "poles"]], "response_mode": "offline", "answer": "[{\"word\":\"poles\",\"start\":0.68,\"end\":1.07}]"} {"utterance_id": "7641-96670-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.15, 2.59, "medea"]], "response_mode": "offline", "answer": "[{\"word\":\"medea\",\"start\":2.15,\"end\":2.59}]"} {"utterance_id": "6467-97061-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.25, 0.45, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.25,\"end\":0.45}]"} {"utterance_id": "116-288048-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0003.flac", "task_type": "lexical_detection", "query": "designs", "prompt": "Process the full audio offline. Find every instance of the word \"designs\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.45, "designs"]], "response_mode": "offline", "answer": "[{\"word\":\"designs\",\"start\":3.05,\"end\":3.45}]"} {"utterance_id": "1630-102884-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0002.flac", "task_type": "lexical_detection", "query": "flattered", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"flattered\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.61, 1.2, "flattered"]], "response_mode": "offline", "answer": "[{\"word\":\"flattered\",\"start\":0.61,\"end\":1.2}]"} {"utterance_id": "2506-13150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.14, 2.42, "ten"]], "response_mode": "offline", "answer": "[{\"word\":\"ten\",\"start\":2.14,\"end\":2.42}]"} {"utterance_id": "1650-173552-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0007.flac", "task_type": "lexical_detection", "query": "monarch", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"monarch\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.13, 5.67, "monarch"]], "response_mode": "offline", "answer": "[{\"word\":\"monarch\",\"start\":5.13,\"end\":5.67}]"} {"utterance_id": "8173-294714-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0043.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 3.82, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":3.52,\"end\":3.82}]"} {"utterance_id": "3660-172182-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0033.flac", "task_type": "lexical_detection", "query": "journeying", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"journeying\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.36, 2.85, "journeying"]], "response_mode": "offline", "answer": "[{\"word\":\"journeying\",\"start\":2.36,\"end\":2.85}]"} {"utterance_id": "4570-56594-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.68, "uncle sam"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle sam\",\"start\":4.82,\"end\":5.68}]"} {"utterance_id": "8254-84205-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0003.flac", "task_type": "lexical_detection", "query": "suffer", "prompt": "After the audio ends, extract all occurrences of the word \"suffer\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.61, 7.07, "suffer"]], "response_mode": "offline", "answer": "[{\"word\":\"suffer\",\"start\":6.61,\"end\":7.07}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[2.89, 3.19, "rider"]], "response_mode": "offline", "answer": "[{\"word\":\"rider\",\"start\":2.89,\"end\":3.19}]"} {"utterance_id": "4153-185072-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0004.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.54, 3.01, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":2.54,\"end\":3.01}]"} {"utterance_id": "1630-141772-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0013.flac", "task_type": "lexical_detection", "query": "emperor's", "prompt": "After listening to the whole recording, output every instance of the word \"emperor's\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.95, 4.57, "emperor's"]], "response_mode": "offline", "answer": "[{\"word\":\"emperor's\",\"start\":3.95,\"end\":4.57}]"} {"utterance_id": "3660-172183-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0023.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 2.11, "repast"]], "response_mode": "offline", "answer": "[{\"word\":\"repast\",\"start\":1.45,\"end\":2.11}]"} {"utterance_id": "3660-6517-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0024.flac", "task_type": "lexical_detection", "query": "ponies", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ponies\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.09, 4.46, "ponies"]], "response_mode": "offline", "answer": "[{\"word\":\"ponies\",\"start\":4.09,\"end\":4.46}]"} {"utterance_id": "1650-167613-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0015.flac", "task_type": "lexical_detection", "query": "skillful", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"skillful\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.65, 2.31, "skillful"]], "response_mode": "offline", "answer": "[{\"word\":\"skillful\",\"start\":1.65,\"end\":2.31}]"} {"utterance_id": "7697-105817-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.43, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":1.99,\"end\":2.43}]"} {"utterance_id": "8254-84205-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0008.flac", "task_type": "lexical_detection", "query": "canter", "prompt": "Once the audio has finished, identify every instance of the word \"canter\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.78, 1.16, "canter"], [2.7, 3.43, "canter"]], "response_mode": "offline", "answer": "[{\"word\":\"canter\",\"start\":0.78,\"end\":1.16},{\"word\":\"canter\",\"start\":2.7,\"end\":3.43}]"} {"utterance_id": "1686-142278-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0005.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[0.5, 1.11, "one"], [3.64, 4.1, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":1.11},{\"word\":\"two\",\"start\":3.64,\"end\":4.1}]"} {"utterance_id": "1630-73710-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0003.flac", "task_type": "lexical_counting", "query": "like", "prompt": "After listening to the whole recording, output timed count words for the word \"like\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[4.57, 4.77, "one"], [10.24, 10.48, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.57,\"end\":4.77},{\"word\":\"two\",\"start\":10.24,\"end\":10.48}]"} {"utterance_id": "3915-57461-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0030.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.28, "priest"]], "response_mode": "offline", "answer": "[{\"word\":\"priest\",\"start\":1.83,\"end\":2.28}]"} {"utterance_id": "700-122868-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.13, 4.44, "elaine"]], "response_mode": "offline", "answer": "[{\"word\":\"elaine\",\"start\":4.13,\"end\":4.44}]"} {"utterance_id": "1255-138279-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0024.flac", "task_type": "semantic_counting", "query": "clock_time", "prompt": "Listen to the entire clip, then report the count sequence for a time of day as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.86, "one"], [2.07, 2.57, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.66,\"end\":0.86},{\"word\":\"two\",\"start\":2.07,\"end\":2.57}]"} {"utterance_id": "4153-186223-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0004.flac", "task_type": "lexical_detection", "query": "boy", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"boy\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.84, 2.13, "boy"]], "response_mode": "offline", "answer": "[{\"word\":\"boy\",\"start\":1.84,\"end\":2.13}]"} {"utterance_id": "7697-245712-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0017.flac", "task_type": "lexical_detection", "query": "seem", "prompt": "Process the full audio offline. Find every instance of the word \"seem\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 2.81, "seem"]], "response_mode": "offline", "answer": "[{\"word\":\"seem\",\"start\":2.48,\"end\":2.81}]"} {"utterance_id": "4153-186222-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0017.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.82, "elmhurst"]], "response_mode": "offline", "answer": "[{\"word\":\"elmhurst\",\"start\":2.38,\"end\":2.82}]"} {"utterance_id": "1255-90407-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0024.flac", "task_type": "lexical_detection", "query": "eat", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"eat\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.97, 3.29, "eat"]], "response_mode": "offline", "answer": "[{\"word\":\"eat\",\"start\":2.97,\"end\":3.29}]"} {"utterance_id": "700-122867-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0018.flac", "task_type": "lexical_detection", "query": "hadn't", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"hadn't\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.58, 1.03, "hadn't"]], "response_mode": "offline", "answer": "[{\"word\":\"hadn't\",\"start\":0.58,\"end\":1.03}]"} {"utterance_id": "6455-67803-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.92, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":1.45,\"end\":1.92}]"} {"utterance_id": "8288-274162-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.78, 2.35, "bragelonne"]], "response_mode": "offline", "answer": "[{\"word\":\"bragelonne\",\"start\":1.78,\"end\":2.35}]"} {"utterance_id": "1650-173552-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.6, 14.29, "boyars"]], "response_mode": "offline", "answer": "[{\"word\":\"boyars\",\"start\":13.6,\"end\":14.29}]"} {"utterance_id": "6267-65525-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0047.flac", "task_type": "lexical_detection", "query": "found", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"found\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.3, 0.71, "found"]], "response_mode": "offline", "answer": "[{\"word\":\"found\",\"start\":0.3,\"end\":0.71}]"} {"utterance_id": "3663-172528-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0031.flac", "task_type": "lexical_detection", "query": "fiercely", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fiercely\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.55, 1.23, "fiercely"]], "response_mode": "offline", "answer": "[{\"word\":\"fiercely\",\"start\":0.55,\"end\":1.23}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 1.23, "one"], [10.14, 10.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.34,\"end\":1.23},{\"word\":\"two\",\"start\":10.14,\"end\":10.46}]"} {"utterance_id": "1686-142278-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.77, 6.32, "mister hale"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hale\",\"start\":5.77,\"end\":6.32}]"} {"utterance_id": "7641-96684-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0000.flac", "task_type": "lexical_detection", "query": "bread", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"bread\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.47, "bread"]], "response_mode": "offline", "answer": "[{\"word\":\"bread\",\"start\":2.22,\"end\":2.47}]"} {"utterance_id": "7641-96252-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0018.flac", "task_type": "lexical_detection", "query": "weeping", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"weeping\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.15, 0.63, "weeping"]], "response_mode": "offline", "answer": "[{\"word\":\"weeping\",\"start\":0.15,\"end\":0.63}]"} {"utterance_id": "6467-94831-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0010.flac", "task_type": "lexical_detection", "query": "green", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"green\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.55, "green"]], "response_mode": "offline", "answer": "[{\"word\":\"green\",\"start\":1.14,\"end\":1.55}]"} {"utterance_id": "7641-96684-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0003.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.51, "one"], [2.59, 3.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.94,\"end\":1.51},{\"word\":\"two\",\"start\":2.59,\"end\":3.44}]"} {"utterance_id": "5849-50963-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0006.flac", "task_type": "lexical_detection", "query": "counted", "prompt": "Once the audio has finished, identify every instance of the word \"counted\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.65, 1.05, "counted"]], "response_mode": "offline", "answer": "[{\"word\":\"counted\",\"start\":0.65,\"end\":1.05}]"} {"utterance_id": "1686-142278-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0029.flac", "task_type": "lexical_detection", "query": "foundation", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"foundation\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.58, "foundation"]], "response_mode": "offline", "answer": "[{\"word\":\"foundation\",\"start\":1.02,\"end\":1.58}]"} {"utterance_id": "2506-11278-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0017.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After the audio ends, extract all occurrences of a family relation. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.83, "sisters"]], "response_mode": "offline", "answer": "[{\"word\":\"sisters\",\"start\":1.18,\"end\":1.83}]"} {"utterance_id": "7697-105817-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.67, "judge"]], "response_mode": "offline", "answer": "[{\"word\":\"judge\",\"start\":4.32,\"end\":4.67}]"} {"utterance_id": "8254-84205-0072", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0072.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "Listen to the entire clip, then report the count sequence for an animal name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.47, "one"], [3.54, 4.0, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.01,\"end\":2.47},{\"word\":\"two\",\"start\":3.54,\"end\":4.0}]"} {"utterance_id": "8288-274162-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0064.flac", "task_type": "lexical_detection", "query": "eagerly", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"eagerly\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.69, 2.0, "eagerly"]], "response_mode": "offline", "answer": "[{\"word\":\"eagerly\",\"start\":1.69,\"end\":2.0}]"} {"utterance_id": "1650-173552-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.93, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":2.39,\"end\":2.93}]"} {"utterance_id": "3663-172528-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0049.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.42, 8.82, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":8.42,\"end\":8.82}]"} {"utterance_id": "2506-11278-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0017.flac", "task_type": "lexical_detection", "query": "sisters", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sisters\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.83, "sisters"]], "response_mode": "offline", "answer": "[{\"word\":\"sisters\",\"start\":1.18,\"end\":1.83}]"} {"utterance_id": "4515-11057-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0043.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.74, 5.44, "countenance"]], "response_mode": "offline", "answer": "[{\"word\":\"countenance\",\"start\":4.74,\"end\":5.44}]"} {"utterance_id": "4515-11057-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0075.flac", "task_type": "lexical_detection", "query": "marian", "prompt": "Process the full audio offline. Find every instance of the word \"marian\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.45, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":2.06,\"end\":2.45}]"} {"utterance_id": "3915-98647-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0014.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of a country name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.85, 9.37, "japan"]], "response_mode": "offline", "answer": "[{\"word\":\"japan\",\"start\":8.85,\"end\":9.37}]"} {"utterance_id": "4831-29134-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0015.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.54, 1.03, "senator"]], "response_mode": "offline", "answer": "[{\"word\":\"senator\",\"start\":0.54,\"end\":1.03}]"} {"utterance_id": "7601-175351-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0007.flac", "task_type": "lexical_detection", "query": "towards", "prompt": "After the audio ends, extract all occurrences of the word \"towards\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 3.89, "towards"]], "response_mode": "offline", "answer": "[{\"word\":\"towards\",\"start\":3.52,\"end\":3.89}]"} {"utterance_id": "6455-67803-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.87, 1.31, "nostrils"], [2.48, 2.69, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"nostrils\",\"start\":0.87,\"end\":1.31},{\"word\":\"head\",\"start\":2.48,\"end\":2.69}]"} {"utterance_id": "1255-90407-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0015.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "After listening to the whole recording, output timed count words for a family relation (one, two, three, ...) as JSON. If none, reply [].", "labels": [[4.71, 4.95, "one"], [6.22, 6.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.71,\"end\":4.95},{\"word\":\"two\",\"start\":6.22,\"end\":6.6}]"} {"utterance_id": "1585-157660-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0002.flac", "task_type": "lexical_detection", "query": "flight", "prompt": "Once the audio has finished, identify every instance of the word \"flight\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[12.06, 12.55, "flight"]], "response_mode": "offline", "answer": "[{\"word\":\"flight\",\"start\":12.06,\"end\":12.55}]"} {"utterance_id": "4323-18416-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0007.flac", "task_type": "lexical_detection", "query": "chatterton", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"chatterton\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.31, "chatterton"]], "response_mode": "offline", "answer": "[{\"word\":\"chatterton\",\"start\":2.81,\"end\":3.31}]"} {"utterance_id": "4153-186223-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[3.96, 4.22, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":3.96,\"end\":4.22}]"} {"utterance_id": "116-288047-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0021.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.79, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":2.39,\"end\":2.79}]"} {"utterance_id": "5849-50964-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.23, 2.74, "dorothy"]], "response_mode": "offline", "answer": "[{\"word\":\"dorothy\",\"start\":2.23,\"end\":2.74}]"} {"utterance_id": "1630-96099-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0017.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen carefully to the entire clip, then report every instance of a plant name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.08, 6.4, "leaves"]], "response_mode": "offline", "answer": "[{\"word\":\"leaves\",\"start\":6.08,\"end\":6.4}]"} {"utterance_id": "1650-167613-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.15, "barclay"]], "response_mode": "offline", "answer": "[{\"word\":\"barclay\",\"start\":1.57,\"end\":2.15}]"} {"utterance_id": "4323-55228-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.5, 2.75, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.5,\"end\":2.75}]"} {"utterance_id": "6841-88291-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.63, "homer"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":0.34,\"end\":0.63}]"} {"utterance_id": "7697-105815-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0039.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[13.14, 13.47, "watchful"]], "response_mode": "offline", "answer": "[{\"word\":\"watchful\",\"start\":13.14,\"end\":13.47}]"} {"utterance_id": "5849-50873-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.48, 1.48, "captain wadsworth"]], "response_mode": "offline", "answer": "[{\"word\":\"captain wadsworth\",\"start\":0.48,\"end\":1.48}]"} {"utterance_id": "1585-131718-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0034.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.25, 0.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.25,\"end\":0.63}]"} {"utterance_id": "8254-115543-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0023.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Do not answer until the audio ends. Then list each instance of a color name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[11.0, 11.3, "coloured"]], "response_mode": "offline", "answer": "[{\"word\":\"coloured\",\"start\":11.0,\"end\":11.3}]"} {"utterance_id": "3915-98647-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0034.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.76, 5.42, "birds"]], "response_mode": "offline", "answer": "[{\"word\":\"birds\",\"start\":4.76,\"end\":5.42}]"} {"utterance_id": "5849-50873-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0037.flac", "task_type": "semantic_counting", "query": "plant", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a plant name (word, start, end). If none, reply [].", "labels": [[0.62, 0.9, "one"], [5.14, 5.83, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.62,\"end\":0.9},{\"word\":\"two\",\"start\":5.14,\"end\":5.83}]"} {"utterance_id": "6455-67803-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0007.flac", "task_type": "lexical_detection", "query": "understand", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"understand\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.73, "understand"]], "response_mode": "offline", "answer": "[{\"word\":\"understand\",\"start\":1.0,\"end\":1.73}]"} {"utterance_id": "6123-59150-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0031.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.33, 6.6, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.33,\"end\":6.6}]"} {"utterance_id": "4570-14911-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0002.flac", "task_type": "lexical_detection", "query": "useless", "prompt": "Process the full audio offline. Find every instance of the word \"useless\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.49, "useless"]], "response_mode": "offline", "answer": "[{\"word\":\"useless\",\"start\":0.99,\"end\":1.49}]"} {"utterance_id": "2506-13150-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0004.flac", "task_type": "lexical_detection", "query": "land", "prompt": "Process the full audio offline. Find every instance of the word \"land\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 2.08, "land"]], "response_mode": "offline", "answer": "[{\"word\":\"land\",\"start\":1.5,\"end\":2.08}]"} {"utterance_id": "1686-142278-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0067.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.84, 8.33, "hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"hundred\",\"start\":7.84,\"end\":8.33}]"} {"utterance_id": "3915-57461-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0028.flac", "task_type": "lexical_detection", "query": "spirit", "prompt": "After listening to the whole recording, output every instance of the word \"spirit\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.77, "spirit"]], "response_mode": "offline", "answer": "[{\"word\":\"spirit\",\"start\":4.23,\"end\":4.77}]"} {"utterance_id": "1651-136854-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0021.flac", "task_type": "lexical_detection", "query": "scarce", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"scarce\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.62, "scarce"]], "response_mode": "offline", "answer": "[{\"word\":\"scarce\",\"start\":3.15,\"end\":3.62}]"} {"utterance_id": "1701-141760-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0010.flac", "task_type": "lexical_detection", "query": "kiss", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"kiss\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.15, 3.45, "kiss"]], "response_mode": "offline", "answer": "[{\"word\":\"kiss\",\"start\":3.15,\"end\":3.45}]"} {"utterance_id": "3660-172183-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0003.flac", "task_type": "lexical_detection", "query": "confession", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"confession\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.52, 2.13, "confession"]], "response_mode": "offline", "answer": "[{\"word\":\"confession\",\"start\":1.52,\"end\":2.13}]"} {"utterance_id": "4572-112381-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0004.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.35, 3.89, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":3.35,\"end\":3.89}]"} {"utterance_id": "1255-138279-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0024.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "You will hear a complete utterance. When it ends, list every instance of a time of day you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.86, "night"], [2.07, 2.57, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":0.66,\"end\":0.86},{\"word\":\"night\",\"start\":2.07,\"end\":2.57}]"} {"utterance_id": "7697-245712-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0009.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, produce a running count of an animal name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.73, 1.13, "one"], [4.82, 5.23, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.73,\"end\":1.13},{\"word\":\"two\",\"start\":4.82,\"end\":5.23}]"} {"utterance_id": "8173-294714-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0002.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After listening to the whole recording, output every instance of a money or currency term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.51, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":1.22,\"end\":1.51}]"} {"utterance_id": "6841-88294-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0059.flac", "task_type": "lexical_detection", "query": "gratitude", "prompt": "Process the full audio offline. Find every instance of the word \"gratitude\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 2.05, "gratitude"]], "response_mode": "offline", "answer": "[{\"word\":\"gratitude\",\"start\":1.36,\"end\":2.05}]"} {"utterance_id": "1630-96099-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0021.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.91, "hundred"], [3.48, 4.11, "hundred"], [7.26, 7.83, "seventeen"]], "response_mode": "offline", "answer": "[{\"word\":\"hundred\",\"start\":1.25,\"end\":1.91},{\"word\":\"hundred\",\"start\":3.48,\"end\":4.11},{\"word\":\"seventeen\",\"start\":7.26,\"end\":7.83}]"} {"utterance_id": "1650-167613-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.93, 9.28, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.93,\"end\":9.28}]"} {"utterance_id": "8254-84205-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.73, 4.08, "limbs"]], "response_mode": "offline", "answer": "[{\"word\":\"limbs\",\"start\":3.73,\"end\":4.08}]"} {"utterance_id": "4153-186222-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0024.flac", "task_type": "lexical_detection", "query": "sure", "prompt": "Process the full audio offline. Find every instance of the word \"sure\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.07, "sure"]], "response_mode": "offline", "answer": "[{\"word\":\"sure\",\"start\":0.7,\"end\":1.07}]"} {"utterance_id": "6455-67804-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 2.08, "lord lansdowne"]], "response_mode": "offline", "answer": "[{\"word\":\"lord lansdowne\",\"start\":1.28,\"end\":2.08}]"} {"utterance_id": "6267-53049-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0030.flac", "task_type": "lexical_detection", "query": "patted", "prompt": "After listening to the whole recording, output every instance of the word \"patted\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.31, "patted"]], "response_mode": "offline", "answer": "[{\"word\":\"patted\",\"start\":0.77,\"end\":1.31}]"} {"utterance_id": "4831-18525-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0031.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.88, "one"], [2.11, 2.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.33,\"end\":0.88},{\"word\":\"two\",\"start\":2.11,\"end\":2.66}]"} {"utterance_id": "1630-73710-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0016.flac", "task_type": "lexical_detection", "query": "dare", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"dare\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.33, 0.56, "dare"]], "response_mode": "offline", "answer": "[{\"word\":\"dare\",\"start\":0.33,\"end\":0.56}]"} {"utterance_id": "6123-59150-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0012.flac", "task_type": "lexical_detection", "query": "wearing", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wearing\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.31, 3.65, "wearing"]], "response_mode": "offline", "answer": "[{\"word\":\"wearing\",\"start\":3.31,\"end\":3.65}]"} {"utterance_id": "1585-131718-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0050.flac", "task_type": "lexical_detection", "query": "winter", "prompt": "Once the audio has finished, identify every instance of the word \"winter\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.63, 1.04, "winter"]], "response_mode": "offline", "answer": "[{\"word\":\"winter\",\"start\":0.63,\"end\":1.04}]"} {"utterance_id": "1650-167613-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0054.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.17, 0.46, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.17,\"end\":0.46}]"} {"utterance_id": "1255-138279-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0021.flac", "task_type": "lexical_detection", "query": "live", "prompt": "After the audio ends, extract all occurrences of the word \"live\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 0.89, "live"]], "response_mode": "offline", "answer": "[{\"word\":\"live\",\"start\":0.71,\"end\":0.89}]"} {"utterance_id": "1701-141759-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0026.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.66, "ear"]], "response_mode": "offline", "answer": "[{\"word\":\"ear\",\"start\":1.45,\"end\":1.66}]"} {"utterance_id": "8254-115543-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.55, 0.86, "chola"]], "response_mode": "offline", "answer": "[{\"word\":\"chola\",\"start\":0.55,\"end\":0.86}]"} {"utterance_id": "4572-64670-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0009.flac", "task_type": "lexical_detection", "query": "enforced", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"enforced\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.26, "enforced"]], "response_mode": "offline", "answer": "[{\"word\":\"enforced\",\"start\":0.78,\"end\":1.26}]"} {"utterance_id": "4570-24733-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0003.flac", "task_type": "lexical_detection", "query": "necessarily", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"necessarily\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.91, 1.84, "necessarily"]], "response_mode": "offline", "answer": "[{\"word\":\"necessarily\",\"start\":0.91,\"end\":1.84}]"} {"utterance_id": "5849-50962-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0015.flac", "task_type": "lexical_detection", "query": "stranger", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"stranger\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.64, 5.12, "stranger"]], "response_mode": "offline", "answer": "[{\"word\":\"stranger\",\"start\":4.64,\"end\":5.12}]"} {"utterance_id": "4153-186222-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0003.flac", "task_type": "lexical_detection", "query": "asking", "prompt": "After the audio ends, extract all occurrences of the word \"asking\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.26, "asking"]], "response_mode": "offline", "answer": "[{\"word\":\"asking\",\"start\":0.79,\"end\":1.26}]"} {"utterance_id": "700-122867-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0032.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.27, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.09,\"end\":1.27}]"} {"utterance_id": "4323-13259-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0021.flac", "task_type": "lexical_counting", "query": "hear", "prompt": "Listen to the entire clip, then report the count sequence for the word \"hear\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.4, 4.56, "one"], [6.6, 6.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.4,\"end\":4.56},{\"word\":\"two\",\"start\":6.6,\"end\":6.87}]"} {"utterance_id": "7697-105815-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0025.flac", "task_type": "lexical_detection", "query": "cloak", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"cloak\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.17, 5.49, "cloak"]], "response_mode": "offline", "answer": "[{\"word\":\"cloak\",\"start\":5.17,\"end\":5.49}]"} {"utterance_id": "7641-96684-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0020.flac", "task_type": "lexical_detection", "query": "terrible", "prompt": "Process the full audio offline. Find every instance of the word \"terrible\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.62, "terrible"]], "response_mode": "offline", "answer": "[{\"word\":\"terrible\",\"start\":1.16,\"end\":1.62}]"} {"utterance_id": "6267-65525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.73, 0.99, "jane"], [5.09, 5.59, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"jane\",\"start\":0.73,\"end\":0.99},{\"word\":\"sidney\",\"start\":5.09,\"end\":5.59}]"} {"utterance_id": "3660-172183-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.96, "geraint"], [2.97, 3.3, "arthur's"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":0.52,\"end\":0.96},{\"word\":\"arthur's\",\"start\":2.97,\"end\":3.3}]"} {"utterance_id": "8254-115543-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0044.flac", "task_type": "lexical_detection", "query": "sacrifice", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"sacrifice\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.65, 1.26, "sacrifice"]], "response_mode": "offline", "answer": "[{\"word\":\"sacrifice\",\"start\":0.65,\"end\":1.26}]"} {"utterance_id": "1585-131718-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0021.flac", "task_type": "lexical_detection", "query": "genuine", "prompt": "Process the full audio offline. Find every instance of the word \"genuine\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.83, 4.34, "genuine"]], "response_mode": "offline", "answer": "[{\"word\":\"genuine\",\"start\":3.83,\"end\":4.34}]"} {"utterance_id": "4323-18416-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0032.flac", "task_type": "lexical_detection", "query": "you've", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"you've\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.0, 2.14, "you've"]], "response_mode": "offline", "answer": "[{\"word\":\"you've\",\"start\":2.0,\"end\":2.14}]"} {"utterance_id": "1585-157660-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0005.flac", "task_type": "lexical_detection", "query": "devil's", "prompt": "Once the audio has finished, identify every instance of the word \"devil's\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.62, 1.01, "devil's"]], "response_mode": "offline", "answer": "[{\"word\":\"devil's\",\"start\":0.62,\"end\":1.01}]"} {"utterance_id": "1650-167613-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0050.flac", "task_type": "lexical_detection", "query": "suddenly", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"suddenly\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.78, 3.33, "suddenly"]], "response_mode": "offline", "answer": "[{\"word\":\"suddenly\",\"start\":2.78,\"end\":3.33}]"} {"utterance_id": "1686-142278-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0039.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.2, 6.66, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":6.2,\"end\":6.66}]"} {"utterance_id": "4323-55228-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0043.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.32, 1.64, "billy's"], [3.12, 3.52, "bertram's"]], "response_mode": "offline", "answer": "[{\"word\":\"billy's\",\"start\":1.32,\"end\":1.64},{\"word\":\"bertram's\",\"start\":3.12,\"end\":3.52}]"} {"utterance_id": "116-288045-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.34, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.08,\"end\":1.34}]"} {"utterance_id": "3663-172528-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0047.flac", "task_type": "lexical_detection", "query": "reverse", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"reverse\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.83, 6.45, "reverse"]], "response_mode": "offline", "answer": "[{\"word\":\"reverse\",\"start\":5.83,\"end\":6.45}]"} {"utterance_id": "3660-172182-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0011.flac", "task_type": "lexical_counting", "query": "thee", "prompt": "When the utterance is complete, report how the word \"thee\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.36, "one"], [2.47, 2.58, "two"], [8.97, 9.19, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.26,\"end\":1.36},{\"word\":\"two\",\"start\":2.47,\"end\":2.58},{\"word\":\"three\",\"start\":8.97,\"end\":9.19}]"} {"utterance_id": "1701-141760-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0038.flac", "task_type": "lexical_detection", "query": "lie", "prompt": "Process the full audio offline. Find every instance of the word \"lie\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.67, 5.01, "lie"]], "response_mode": "offline", "answer": "[{\"word\":\"lie\",\"start\":4.67,\"end\":5.01}]"} {"utterance_id": "6841-88294-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0000.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Once the audio has finished, identify every instance of a weather term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.95, 2.36, "storm"]], "response_mode": "offline", "answer": "[{\"word\":\"storm\",\"start\":1.95,\"end\":2.36}]"} {"utterance_id": "6123-59186-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.2, 2.85, "mister hardy's"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy's\",\"start\":2.2,\"end\":2.85}]"} {"utterance_id": "4570-24733-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0004.flac", "task_type": "lexical_detection", "query": "hypothesis", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hypothesis\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[26.91, 27.8, "hypothesis"]], "response_mode": "offline", "answer": "[{\"word\":\"hypothesis\",\"start\":26.91,\"end\":27.8}]"} {"utterance_id": "7601-175351-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0016.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.45, 4.97, "seventh"]], "response_mode": "offline", "answer": "[{\"word\":\"seventh\",\"start\":4.45,\"end\":4.97}]"} {"utterance_id": "6467-97061-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.41, "forehead"], [7.32, 7.69, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"forehead\",\"start\":0.96,\"end\":1.41},{\"word\":\"hands\",\"start\":7.32,\"end\":7.69}]"} {"utterance_id": "4570-102353-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.41, "president"]], "response_mode": "offline", "answer": "[{\"word\":\"president\",\"start\":1.75,\"end\":2.41}]"} {"utterance_id": "8288-274162-0060", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0060.flac", "task_type": "lexical_detection", "query": "cat", "prompt": "Once the audio has finished, identify every instance of the word \"cat\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.07, 3.39, "cat"]], "response_mode": "offline", "answer": "[{\"word\":\"cat\",\"start\":3.07,\"end\":3.39}]"} {"utterance_id": "4831-29134-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0009.flac", "task_type": "lexical_detection", "query": "hardly", "prompt": "When the utterance is complete, report all instances of the word \"hardly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.03, 2.43, "hardly"]], "response_mode": "offline", "answer": "[{\"word\":\"hardly\",\"start\":2.03,\"end\":2.43}]"} {"utterance_id": "8173-294714-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0055.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.74, 4.67, "mister van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"mister van brandt\",\"start\":3.74,\"end\":4.67}]"} {"utterance_id": "4515-11057-0084", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0084.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Once the audio has finished, identify every instance of a plant name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.58, 9.91, "tree"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":9.58,\"end\":9.91}]"} {"utterance_id": "116-288048-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0016.flac", "task_type": "lexical_detection", "query": "saw", "prompt": "Once the audio has finished, identify every instance of the word \"saw\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.25, 1.54, "saw"]], "response_mode": "offline", "answer": "[{\"word\":\"saw\",\"start\":1.25,\"end\":1.54}]"} {"utterance_id": "4831-25894-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0010.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 1.01, "one"], [3.0, 3.5, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.63,\"end\":1.01},{\"word\":\"two\",\"start\":3.0,\"end\":3.5}]"} {"utterance_id": "4570-102353-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0003.flac", "task_type": "lexical_detection", "query": "inquiries", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"inquiries\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.24, "inquiries"]], "response_mode": "offline", "answer": "[{\"word\":\"inquiries\",\"start\":0.5,\"end\":1.24}]"} {"utterance_id": "1650-167613-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0009.flac", "task_type": "lexical_detection", "query": "pierre", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pierre\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.84, 2.29, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":1.84,\"end\":2.29}]"} {"utterance_id": "6467-94831-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0020.flac", "task_type": "lexical_detection", "query": "help", "prompt": "After listening to the whole recording, output every instance of the word \"help\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.84, "help"]], "response_mode": "offline", "answer": "[{\"word\":\"help\",\"start\":2.35,\"end\":2.84}]"} {"utterance_id": "1686-142278-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0047.flac", "task_type": "lexical_detection", "query": "mamma", "prompt": "After the audio ends, extract all occurrences of the word \"mamma\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.29, "mamma"]], "response_mode": "offline", "answer": "[{\"word\":\"mamma\",\"start\":0.85,\"end\":1.29}]"} {"utterance_id": "3663-172528-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0030.flac", "task_type": "lexical_detection", "query": "giorgio", "prompt": "After listening to the whole recording, output every instance of the word \"giorgio\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.29, 0.82, "giorgio"]], "response_mode": "offline", "answer": "[{\"word\":\"giorgio\",\"start\":0.29,\"end\":0.82}]"} {"utterance_id": "3915-98647-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0027.flac", "task_type": "lexical_detection", "query": "occur", "prompt": "After listening to the whole recording, output every instance of the word \"occur\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 1.88, "occur"]], "response_mode": "offline", "answer": "[{\"word\":\"occur\",\"start\":1.5,\"end\":1.88}]"} {"utterance_id": "6841-88291-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0018.flac", "task_type": "lexical_detection", "query": "calf", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"calf\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.6, 0.89, "calf"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":0.6,\"end\":0.89}]"} {"utterance_id": "1701-141759-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0015.flac", "task_type": "lexical_detection", "query": "shut", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"shut\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.72, "shut"]], "response_mode": "offline", "answer": "[{\"word\":\"shut\",\"start\":1.46,\"end\":1.72}]"} {"utterance_id": "4572-112375-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.85, 2.37, "tara"]], "response_mode": "offline", "answer": "[{\"word\":\"tara\",\"start\":1.85,\"end\":2.37}]"} {"utterance_id": "8173-294714-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0034.flac", "task_type": "lexical_detection", "query": "estimate", "prompt": "After the audio ends, extract all occurrences of the word \"estimate\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.61, "estimate"]], "response_mode": "offline", "answer": "[{\"word\":\"estimate\",\"start\":1.04,\"end\":1.61}]"} {"utterance_id": "8254-84205-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0035.flac", "task_type": "lexical_detection", "query": "animals", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"animals\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.15, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":0.69,\"end\":1.15}]"} {"utterance_id": "1630-102884-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0001.flac", "task_type": "lexical_detection", "query": "west", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"west\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.18, 3.75, "west"]], "response_mode": "offline", "answer": "[{\"word\":\"west\",\"start\":3.18,\"end\":3.75}]"} {"utterance_id": "4831-25894-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.36, 2.69, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":2.36,\"end\":2.69}]"} {"utterance_id": "700-122866-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0009.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.28, 2.98, "teens"]], "response_mode": "offline", "answer": "[{\"word\":\"teens\",\"start\":2.28,\"end\":2.98}]"} {"utterance_id": "5849-50962-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0018.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.16, 3.6, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":3.16,\"end\":3.6}]"} {"utterance_id": "6123-59150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0001.flac", "task_type": "lexical_detection", "query": "dostoevsky", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"dostoevsky\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.7, 1.35, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":0.7,\"end\":1.35}]"} {"utterance_id": "6123-59150-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0037.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.57, 2.48, "dostoevsky"], [4.46, 5.6, "peter's lord"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":1.57,\"end\":2.48},{\"word\":\"peter's lord\",\"start\":4.46,\"end\":5.6}]"} {"utterance_id": "7697-245712-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0005.flac", "task_type": "lexical_detection", "query": "universal", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"universal\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.97, 4.73, "universal"]], "response_mode": "offline", "answer": "[{\"word\":\"universal\",\"start\":3.97,\"end\":4.73}]"} {"utterance_id": "3663-172528-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0042.flac", "task_type": "lexical_detection", "query": "want", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"want\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.64, "want"]], "response_mode": "offline", "answer": "[{\"word\":\"want\",\"start\":0.33,\"end\":0.64}]"} {"utterance_id": "1701-141760-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0006.flac", "task_type": "lexical_detection", "query": "remarked", "prompt": "After the audio ends, extract all occurrences of the word \"remarked\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.37, "remarked"]], "response_mode": "offline", "answer": "[{\"word\":\"remarked\",\"start\":2.77,\"end\":3.37}]"} {"utterance_id": "6455-67803-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0001.flac", "task_type": "lexical_detection", "query": "octavia", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"octavia\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.37, "octavia"]], "response_mode": "offline", "answer": "[{\"word\":\"octavia\",\"start\":1.8,\"end\":2.37}]"} {"utterance_id": "4515-11057-0078", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0078.flac", "task_type": "lexical_detection", "query": "trembled", "prompt": "When the utterance is complete, report all instances of the word \"trembled\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.64, "trembled"]], "response_mode": "offline", "answer": "[{\"word\":\"trembled\",\"start\":3.06,\"end\":3.64}]"} {"utterance_id": "700-122868-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0021.flac", "task_type": "lexical_detection", "query": "somebody", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"somebody\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.89, 1.28, "somebody"]], "response_mode": "offline", "answer": "[{\"word\":\"somebody\",\"start\":0.89,\"end\":1.28}]"} {"utterance_id": "4831-25894-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0029.flac", "task_type": "lexical_detection", "query": "socks", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"socks\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.86, 2.27, "socks"]], "response_mode": "offline", "answer": "[{\"word\":\"socks\",\"start\":1.86,\"end\":2.27}]"} {"utterance_id": "1585-131718-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0008.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Once the audio has finished, identify every instance of an organization name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.48, 3.05, "companies"], [6.16, 6.72, "companies"]], "response_mode": "offline", "answer": "[{\"word\":\"companies\",\"start\":2.48,\"end\":3.05},{\"word\":\"companies\",\"start\":6.16,\"end\":6.72}]"} {"utterance_id": "6467-62797-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0006.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, list every instance of a plant name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.03, 8.58, "jak fruit"]], "response_mode": "offline", "answer": "[{\"word\":\"jak fruit\",\"start\":8.03,\"end\":8.58}]"} {"utterance_id": "1650-167613-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0017.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.77, "commander"]], "response_mode": "offline", "answer": "[{\"word\":\"commander\",\"start\":1.09,\"end\":1.77}]"} {"utterance_id": "116-288045-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.6, 2.22, "homer"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":1.6,\"end\":2.22}]"} {"utterance_id": "116-288045-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0016.flac", "task_type": "lexical_counting", "query": "heart", "prompt": "Once the audio has finished, list each occurrence of the word \"heart\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.8, 2.26, "one"], [6.72, 7.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.8,\"end\":2.26},{\"word\":\"two\",\"start\":6.72,\"end\":7.18}]"} {"utterance_id": "1255-90413-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0015.flac", "task_type": "lexical_detection", "query": "rather", "prompt": "Once the audio has finished, identify every instance of the word \"rather\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.89, 1.14, "rather"]], "response_mode": "offline", "answer": "[{\"word\":\"rather\",\"start\":0.89,\"end\":1.14}]"} {"utterance_id": "4153-186223-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0008.flac", "task_type": "lexical_detection", "query": "lad", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"lad\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.59, "lad"]], "response_mode": "offline", "answer": "[{\"word\":\"lad\",\"start\":2.31,\"end\":2.59}]"} {"utterance_id": "6455-66379-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0009.flac", "task_type": "lexical_detection", "query": "light", "prompt": "Once the audio has finished, identify every instance of the word \"light\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.85, "light"]], "response_mode": "offline", "answer": "[{\"word\":\"light\",\"start\":0.5,\"end\":0.85}]"} {"utterance_id": "116-288046-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0000.flac", "task_type": "lexical_counting", "query": "truth", "prompt": "Listen to the entire clip, then report the count sequence for the word \"truth\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.99, 4.32, "one"], [6.69, 6.97, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.99,\"end\":4.32},{\"word\":\"two\",\"start\":6.69,\"end\":6.97}]"} {"utterance_id": "1630-96099-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0017.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.91, 2.38, "chief"]], "response_mode": "offline", "answer": "[{\"word\":\"chief\",\"start\":1.91,\"end\":2.38}]"} {"utterance_id": "5849-50963-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0003.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Once the audio has finished, identify every instance of a disease name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.5, 6.01, "fever"]], "response_mode": "offline", "answer": "[{\"word\":\"fever\",\"start\":5.5,\"end\":6.01}]"} {"utterance_id": "8288-274162-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0047.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.04, 4.6, "carpenter"]], "response_mode": "offline", "answer": "[{\"word\":\"carpenter\",\"start\":4.04,\"end\":4.6}]"} {"utterance_id": "4570-102353-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0006.flac", "task_type": "lexical_detection", "query": "further", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"further\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.19, "further"]], "response_mode": "offline", "answer": "[{\"word\":\"further\",\"start\":0.92,\"end\":1.19}]"} {"utterance_id": "4570-56594-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0006.flac", "task_type": "semantic_detection", "query": "city", "prompt": "You will hear a complete utterance. When it ends, list every instance of a city name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 2.1, "orleans"]], "response_mode": "offline", "answer": "[{\"word\":\"orleans\",\"start\":1.26,\"end\":2.1}]"} {"utterance_id": "6267-65525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.63, 4.16, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.63,\"end\":4.16}]"} {"utterance_id": "7697-105815-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.49, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.86,\"end\":1.49}]"} {"utterance_id": "1585-131718-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0028.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.17, 8.6, "importer"]], "response_mode": "offline", "answer": "[{\"word\":\"importer\",\"start\":8.17,\"end\":8.6}]"} {"utterance_id": "8254-84205-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 1.91, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":1.68,\"end\":1.91}]"} {"utterance_id": "116-288048-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.58, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":1.21,\"end\":1.58}]"} {"utterance_id": "4515-11057-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0048.flac", "task_type": "lexical_detection", "query": "lucky", "prompt": "After listening to the whole recording, output every instance of the word \"lucky\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.77, "lucky"]], "response_mode": "offline", "answer": "[{\"word\":\"lucky\",\"start\":2.27,\"end\":2.77}]"} {"utterance_id": "6467-62797-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0006.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.48, 5.64, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.48,\"end\":5.64}]"} {"utterance_id": "6841-88294-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0039.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.2, "back"]], "response_mode": "offline", "answer": "[{\"word\":\"back\",\"start\":0.91,\"end\":1.2}]"} {"utterance_id": "3915-98647-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0004.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.61, "mother"], [8.1, 8.53, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":2.24,\"end\":2.61},{\"word\":\"son\",\"start\":8.1,\"end\":8.53}]"} {"utterance_id": "8288-274162-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0001.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.08, 4.55, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":4.08,\"end\":4.55}]"} {"utterance_id": "8254-84205-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0000.flac", "task_type": "lexical_detection", "query": "forty", "prompt": "When the utterance is complete, report all instances of the word \"forty\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.02, "forty"]], "response_mode": "offline", "answer": "[{\"word\":\"forty\",\"start\":0.7,\"end\":1.02}]"} {"utterance_id": "3660-172183-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0017.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.44, 5.84, "horses"]], "response_mode": "offline", "answer": "[{\"word\":\"horses\",\"start\":5.44,\"end\":5.84}]"} {"utterance_id": "4831-25894-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0015.flac", "task_type": "semantic_counting", "query": "money", "prompt": "After listening to the whole recording, output timed count words for a money or currency term (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.18, 1.61, "one"], [2.62, 2.9, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.18,\"end\":1.61},{\"word\":\"two\",\"start\":2.62,\"end\":2.9}]"} {"utterance_id": "1650-157641-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0011.flac", "task_type": "semantic_detection", "query": "country", "prompt": "When the utterance is complete, report all instances of a country name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.41, 0.97, "germany"], [3.65, 4.02, "france"]], "response_mode": "offline", "answer": "[{\"word\":\"germany\",\"start\":0.41,\"end\":0.97},{\"word\":\"france\",\"start\":3.65,\"end\":4.02}]"} {"utterance_id": "7641-96684-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0005.flac", "task_type": "lexical_detection", "query": "joy", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"joy\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.08, 3.46, "joy"]], "response_mode": "offline", "answer": "[{\"word\":\"joy\",\"start\":3.08,\"end\":3.46}]"} {"utterance_id": "7641-96684-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0021.flac", "task_type": "lexical_counting", "query": "words", "prompt": "Listen to the entire clip, then report the count sequence for the word \"words\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.93, "one"], [3.19, 3.67, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.62,\"end\":0.93},{\"word\":\"two\",\"start\":3.19,\"end\":3.67}]"} {"utterance_id": "1630-141772-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.76, 4.2, "larrey"]], "response_mode": "offline", "answer": "[{\"word\":\"larrey\",\"start\":3.76,\"end\":4.2}]"} {"utterance_id": "1701-141759-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.03, "petya"], [2.19, 2.75, "natasha"]], "response_mode": "offline", "answer": "[{\"word\":\"petya\",\"start\":0.5,\"end\":1.03},{\"word\":\"natasha\",\"start\":2.19,\"end\":2.75}]"} {"utterance_id": "8254-84205-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0061.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.89, 2.29, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":1.89,\"end\":2.29}]"} {"utterance_id": "700-122868-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.45, 4.0, "missus lynde"]], "response_mode": "offline", "answer": "[{\"word\":\"missus lynde\",\"start\":3.45,\"end\":4.0}]"} {"utterance_id": "2506-13150-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.63, "democritus"]], "response_mode": "offline", "answer": "[{\"word\":\"democritus\",\"start\":0.65,\"end\":1.63}]"} {"utterance_id": "1650-167613-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0005.flac", "task_type": "lexical_detection", "query": "anyone", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"anyone\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.89, "anyone"]], "response_mode": "offline", "answer": "[{\"word\":\"anyone\",\"start\":2.31,\"end\":2.89}]"} {"utterance_id": "4515-11057-0077", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0077.flac", "task_type": "lexical_detection", "query": "wish", "prompt": "After the audio ends, extract all occurrences of the word \"wish\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.49, "wish"]], "response_mode": "offline", "answer": "[{\"word\":\"wish\",\"start\":1.2,\"end\":1.49}]"} {"utterance_id": "8173-294714-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 2.28, "van brandt"]], "response_mode": "offline", "answer": "[{\"word\":\"van brandt\",\"start\":1.71,\"end\":2.28}]"} {"utterance_id": "5849-50962-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0000.flac", "task_type": "lexical_detection", "query": "edge", "prompt": "Process the full audio offline. Find every instance of the word \"edge\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.42, 4.77, "edge"]], "response_mode": "offline", "answer": "[{\"word\":\"edge\",\"start\":4.42,\"end\":4.77}]"} {"utterance_id": "1651-136854-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0031.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.53, 2.87, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.53,\"end\":2.87}]"} {"utterance_id": "6123-59150-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.36, 8.13, "dostoevsky s"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky s\",\"start\":7.36,\"end\":8.13}]"} {"utterance_id": "700-122868-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.87, 5.36, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"lips\",\"start\":4.87,\"end\":5.36}]"} {"utterance_id": "6267-53049-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.88, "penelope"], [6.11, 6.79, "doris"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.21,\"end\":0.88},{\"word\":\"doris\",\"start\":6.11,\"end\":6.79}]"} {"utterance_id": "2506-11278-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0007.flac", "task_type": "lexical_detection", "query": "honest", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"honest\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.62, 0.92, "honest"]], "response_mode": "offline", "answer": "[{\"word\":\"honest\",\"start\":0.62,\"end\":0.92}]"} {"utterance_id": "1630-141772-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0015.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.02, 2.88, "one"], [7.02, 8.1, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.02,\"end\":2.88},{\"word\":\"two\",\"start\":7.02,\"end\":8.1}]"} {"utterance_id": "4323-55228-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 0.65, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.31,\"end\":0.65}]"} {"utterance_id": "4515-11057-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0032.flac", "task_type": "lexical_detection", "query": "current", "prompt": "After listening to the whole recording, output every instance of the word \"current\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.68, "current"]], "response_mode": "offline", "answer": "[{\"word\":\"current\",\"start\":2.11,\"end\":2.68}]"} {"utterance_id": "4831-18525-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0029.flac", "task_type": "lexical_detection", "query": "books", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"books\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.41, 2.74, "books"]], "response_mode": "offline", "answer": "[{\"word\":\"books\",\"start\":2.41,\"end\":2.74}]"} {"utterance_id": "1585-157660-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0008.flac", "task_type": "lexical_detection", "query": "fails", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fails\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 2.19, "fails"]], "response_mode": "offline", "answer": "[{\"word\":\"fails\",\"start\":1.63,\"end\":2.19}]"} {"utterance_id": "7641-96684-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.89, 1.4, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":0.89,\"end\":1.4}]"} {"utterance_id": "3663-172528-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.31, 1.67, "maestro francesco"]], "response_mode": "offline", "answer": "[{\"word\":\"maestro francesco\",\"start\":0.31,\"end\":1.67}]"} {"utterance_id": "7601-291468-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[18.11, 18.52, "horse"], [24.76, 25.37, "horses"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":18.11,\"end\":18.52},{\"word\":\"horses\",\"start\":24.76,\"end\":25.37}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.98, "commander"], [7.62, 8.06, "governor"]], "response_mode": "offline", "answer": "[{\"word\":\"commander\",\"start\":1.47,\"end\":1.98},{\"word\":\"governor\",\"start\":7.62,\"end\":8.06}]"} {"utterance_id": "6267-53049-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0016.flac", "task_type": "lexical_detection", "query": "hungry", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"hungry\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.8, 5.4, "hungry"]], "response_mode": "offline", "answer": "[{\"word\":\"hungry\",\"start\":4.8,\"end\":5.4}]"} {"utterance_id": "7601-101622-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0003.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.07, 3.88, "gladiators"]], "response_mode": "offline", "answer": "[{\"word\":\"gladiators\",\"start\":3.07,\"end\":3.88}]"} {"utterance_id": "2506-169427-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0000.flac", "task_type": "semantic_counting", "query": "money", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a money or currency term (word, start, end). If none, reply [].", "labels": [[0.2, 0.76, "one"], [5.88, 6.33, "two"], [17.25, 17.84, "three"], [20.07, 20.5, "four"], [24.24, 25.28, "five"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.2,\"end\":0.76},{\"word\":\"two\",\"start\":5.88,\"end\":6.33},{\"word\":\"three\",\"start\":17.25,\"end\":17.84},{\"word\":\"four\",\"start\":20.07,\"end\":20.5},{\"word\":\"five\",\"start\":24.24,\"end\":25.28}]"} {"utterance_id": "3915-57461-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0008.flac", "task_type": "lexical_detection", "query": "ronin", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"ronin\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.08, "ronin"]], "response_mode": "offline", "answer": "[{\"word\":\"ronin\",\"start\":0.75,\"end\":1.08}]"} {"utterance_id": "1650-173552-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0008.flac", "task_type": "lexical_counting", "query": "gentlemen", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"gentlemen\" with word/start/end. If none, reply [].", "labels": [[3.58, 4.09, "one"], [5.54, 6.11, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.58,\"end\":4.09},{\"word\":\"two\",\"start\":5.54,\"end\":6.11}]"} {"utterance_id": "6123-59150-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.58, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":0.91,\"end\":1.58}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.42, 4.92, "legs"], [6.37, 6.7, "neck"]], "response_mode": "offline", "answer": "[{\"word\":\"legs\",\"start\":4.42,\"end\":4.92},{\"word\":\"neck\",\"start\":6.37,\"end\":6.7}]"} {"utterance_id": "1686-142278-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0013.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.32, 2.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.32,\"end\":2.6}]"} {"utterance_id": "8254-115543-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0030.flac", "task_type": "lexical_detection", "query": "supper", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"supper\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.75, "supper"]], "response_mode": "offline", "answer": "[{\"word\":\"supper\",\"start\":0.5,\"end\":0.75}]"} {"utterance_id": "6123-59150-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0018.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 1.95, "artist"]], "response_mode": "offline", "answer": "[{\"word\":\"artist\",\"start\":1.54,\"end\":1.95}]"} {"utterance_id": "4515-11057-0082", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0082.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.22, 3.72, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":3.22,\"end\":3.72}]"} {"utterance_id": "4153-61735-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0011.flac", "task_type": "lexical_detection", "query": "waiting", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"waiting\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.85, 1.16, "waiting"]], "response_mode": "offline", "answer": "[{\"word\":\"waiting\",\"start\":0.85,\"end\":1.16}]"} {"utterance_id": "1585-131718-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0053.flac", "task_type": "lexical_detection", "query": "enjoyed", "prompt": "After listening to the whole recording, output every instance of the word \"enjoyed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.3, "enjoyed"]], "response_mode": "offline", "answer": "[{\"word\":\"enjoyed\",\"start\":0.94,\"end\":1.3}]"} {"utterance_id": "8254-115543-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0002.flac", "task_type": "lexical_detection", "query": "blankets", "prompt": "After the audio ends, extract all occurrences of the word \"blankets\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.52, 6.06, "blankets"]], "response_mode": "offline", "answer": "[{\"word\":\"blankets\",\"start\":5.52,\"end\":6.06}]"} {"utterance_id": "6123-59186-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0013.flac", "task_type": "lexical_detection", "query": "sensitive", "prompt": "Once the audio has finished, identify every instance of the word \"sensitive\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.36, 1.96, "sensitive"]], "response_mode": "offline", "answer": "[{\"word\":\"sensitive\",\"start\":1.36,\"end\":1.96}]"} {"utterance_id": "6123-59186-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[5.09, 5.89, "mister hardy"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy\",\"start\":5.09,\"end\":5.89}]"} {"utterance_id": "1255-74899-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0018.flac", "task_type": "lexical_detection", "query": "alas", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"alas\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.72, 3.28, "alas"]], "response_mode": "offline", "answer": "[{\"word\":\"alas\",\"start\":2.72,\"end\":3.28}]"} {"utterance_id": "4515-11057-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0070.flac", "task_type": "lexical_detection", "query": "umbrella", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"umbrella\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.4, "umbrella"]], "response_mode": "offline", "answer": "[{\"word\":\"umbrella\",\"start\":0.9,\"end\":1.4}]"} {"utterance_id": "6267-53049-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.5, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":2.77,\"end\":3.5}]"} {"utterance_id": "1650-167613-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0000.flac", "task_type": "lexical_detection", "query": "apparently", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"apparently\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.39, "apparently"]], "response_mode": "offline", "answer": "[{\"word\":\"apparently\",\"start\":3.82,\"end\":4.39}]"} {"utterance_id": "4570-102353-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0004.flac", "task_type": "semantic_counting", "query": "organization", "prompt": "Once the audio has finished, list each occurrence of an organization name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.7, 2.31, "one"], [7.26, 7.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.7,\"end\":2.31},{\"word\":\"two\",\"start\":7.26,\"end\":7.6}]"} {"utterance_id": "4153-185072-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0002.flac", "task_type": "lexical_detection", "query": "protected", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"protected\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 1.2, "protected"]], "response_mode": "offline", "answer": "[{\"word\":\"protected\",\"start\":0.63,\"end\":1.2}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen carefully to the entire clip, then report every instance of a money or currency term as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.35, "cents"], [4.44, 4.75, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"cents\",\"start\":2.95,\"end\":3.35},{\"word\":\"money\",\"start\":4.44,\"end\":4.75}]"} {"utterance_id": "6123-59150-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0004.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.41, "author"]], "response_mode": "offline", "answer": "[{\"word\":\"author\",\"start\":1.9,\"end\":2.41}]"} {"utterance_id": "6123-59186-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.95, 5.65, "mister hardy's"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy's\",\"start\":4.95,\"end\":5.65}]"} {"utterance_id": "1701-141760-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0048.flac", "task_type": "lexical_counting", "query": "stories", "prompt": "Listen to the entire clip, then report the count sequence for the word \"stories\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.12, "one"], [2.99, 3.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.71,\"end\":1.12},{\"word\":\"two\",\"start\":2.99,\"end\":3.53}]"} {"utterance_id": "6455-67804-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0003.flac", "task_type": "lexical_detection", "query": "same", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"same\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.45, "same"]], "response_mode": "offline", "answer": "[{\"word\":\"same\",\"start\":1.18,\"end\":1.45}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[0.53, 1.41, "one"], [6.92, 7.98, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.53,\"end\":1.41},{\"word\":\"two\",\"start\":6.92,\"end\":7.98}]"} {"utterance_id": "6841-88294-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0005.flac", "task_type": "lexical_detection", "query": "breathing", "prompt": "After the audio ends, extract all occurrences of the word \"breathing\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.6, "breathing"]], "response_mode": "offline", "answer": "[{\"word\":\"breathing\",\"start\":2.27,\"end\":2.6}]"} {"utterance_id": "7601-101622-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0000.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After listening to the whole recording, output every instance of an organization name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.16, "romans"]], "response_mode": "offline", "answer": "[{\"word\":\"romans\",\"start\":0.69,\"end\":1.16}]"} {"utterance_id": "116-288045-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.19, "jesus"], [5.0, 5.45, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":1.83,\"end\":2.19},{\"word\":\"apollo\",\"start\":5.0,\"end\":5.45}]"} {"utterance_id": "1630-102884-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0013.flac", "task_type": "lexical_detection", "query": "permanent", "prompt": "After the audio ends, extract all occurrences of the word \"permanent\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.45, 3.03, "permanent"]], "response_mode": "offline", "answer": "[{\"word\":\"permanent\",\"start\":2.45,\"end\":3.03}]"} {"utterance_id": "4572-112381-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0014.flac", "task_type": "lexical_detection", "query": "wealth", "prompt": "When the utterance is complete, report all instances of the word \"wealth\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.23, 4.73, "wealth"]], "response_mode": "offline", "answer": "[{\"word\":\"wealth\",\"start\":4.23,\"end\":4.73}]"} {"utterance_id": "4831-25894-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0022.flac", "task_type": "lexical_detection", "query": "fourth", "prompt": "After listening to the whole recording, output every instance of the word \"fourth\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.58, 1.21, "fourth"]], "response_mode": "offline", "answer": "[{\"word\":\"fourth\",\"start\":0.58,\"end\":1.21}]"} {"utterance_id": "1686-142278-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0055.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.36, 4.77, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":4.36,\"end\":4.77}]"} {"utterance_id": "1255-138279-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0002.flac", "task_type": "lexical_detection", "query": "darker", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"darker\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.23, 1.58, "darker"]], "response_mode": "offline", "answer": "[{\"word\":\"darker\",\"start\":1.23,\"end\":1.58}]"} {"utterance_id": "4570-14911-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0005.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.41, 7.1, "beagle"]], "response_mode": "offline", "answer": "[{\"word\":\"beagle\",\"start\":6.41,\"end\":7.1}]"} {"utterance_id": "6467-56885-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0009.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.45, 2.99, "elephant"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant\",\"start\":2.45,\"end\":2.99}]"} {"utterance_id": "6455-67804-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0009.flac", "task_type": "lexical_detection", "query": "confounded", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"confounded\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.47, 1.97, "confounded"]], "response_mode": "offline", "answer": "[{\"word\":\"confounded\",\"start\":1.47,\"end\":1.97}]"} {"utterance_id": "4153-186222-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0017.flac", "task_type": "lexical_detection", "query": "elmhurst", "prompt": "After the audio ends, extract all occurrences of the word \"elmhurst\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.38, 2.82, "elmhurst"]], "response_mode": "offline", "answer": "[{\"word\":\"elmhurst\",\"start\":2.38,\"end\":2.82}]"} {"utterance_id": "116-288046-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0008.flac", "task_type": "lexical_detection", "query": "rights", "prompt": "After the audio ends, extract all occurrences of the word \"rights\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.39, "rights"]], "response_mode": "offline", "answer": "[{\"word\":\"rights\",\"start\":0.9,\"end\":1.39}]"} {"utterance_id": "6841-88291-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.14, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.95,\"end\":3.14}]"} {"utterance_id": "4153-186223-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0013.flac", "task_type": "lexical_detection", "query": "kenneth", "prompt": "Once the audio has finished, identify every instance of the word \"kenneth\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.38, 0.74, "kenneth"]], "response_mode": "offline", "answer": "[{\"word\":\"kenneth\",\"start\":0.38,\"end\":0.74}]"} {"utterance_id": "1630-141772-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0022.flac", "task_type": "lexical_detection", "query": "nervous", "prompt": "After the audio ends, extract all occurrences of the word \"nervous\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.55, "nervous"]], "response_mode": "offline", "answer": "[{\"word\":\"nervous\",\"start\":1.04,\"end\":1.55}]"} {"utterance_id": "5543-27761-0083", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0083.flac", "task_type": "lexical_detection", "query": "nightmare", "prompt": "Once the audio has finished, identify every instance of the word \"nightmare\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.58, 3.17, "nightmare"]], "response_mode": "offline", "answer": "[{\"word\":\"nightmare\",\"start\":2.58,\"end\":3.17}]"} {"utterance_id": "6841-88291-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0026.flac", "task_type": "lexical_detection", "query": "unnecessary", "prompt": "After listening to the whole recording, output every instance of the word \"unnecessary\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.5, "unnecessary"]], "response_mode": "offline", "answer": "[{\"word\":\"unnecessary\",\"start\":1.79,\"end\":2.5}]"} {"utterance_id": "1255-90407-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0019.flac", "task_type": "lexical_detection", "query": "force", "prompt": "When the utterance is complete, report all instances of the word \"force\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.74, "force"]], "response_mode": "offline", "answer": "[{\"word\":\"force\",\"start\":1.44,\"end\":1.74}]"} {"utterance_id": "7641-96252-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0005.flac", "task_type": "lexical_counting", "query": "knowledge", "prompt": "After listening to the whole recording, output timed count words for the word \"knowledge\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[6.04, 6.49, "one"], [10.13, 10.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.04,\"end\":6.49},{\"word\":\"two\",\"start\":10.13,\"end\":10.53}]"} {"utterance_id": "1686-142278-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0034.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[12.59, 13.04, "margaret's"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret's\",\"start\":12.59,\"end\":13.04}]"} {"utterance_id": "8173-294714-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0052.flac", "task_type": "lexical_detection", "query": "advice", "prompt": "When the utterance is complete, report all instances of the word \"advice\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.02, "advice"]], "response_mode": "offline", "answer": "[{\"word\":\"advice\",\"start\":1.57,\"end\":2.02}]"} {"utterance_id": "5849-50873-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0004.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.98, 8.42, "sixty"]], "response_mode": "offline", "answer": "[{\"word\":\"sixty\",\"start\":7.98,\"end\":8.42}]"} {"utterance_id": "1630-73710-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0015.flac", "task_type": "lexical_detection", "query": "grateful", "prompt": "When the utterance is complete, report all instances of the word \"grateful\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.23, 2.91, "grateful"]], "response_mode": "offline", "answer": "[{\"word\":\"grateful\",\"start\":2.23,\"end\":2.91}]"} {"utterance_id": "4323-55228-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 3.13, "miss billy"]], "response_mode": "offline", "answer": "[{\"word\":\"miss billy\",\"start\":2.58,\"end\":3.13}]"} {"utterance_id": "8288-274162-0062", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0062.flac", "task_type": "lexical_detection", "query": "wait", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wait\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.82, "wait"]], "response_mode": "offline", "answer": "[{\"word\":\"wait\",\"start\":0.53,\"end\":0.82}]"} {"utterance_id": "6267-65525-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0048.flac", "task_type": "lexical_detection", "query": "letters", "prompt": "After listening to the whole recording, output every instance of the word \"letters\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.06, "letters"]], "response_mode": "offline", "answer": "[{\"word\":\"letters\",\"start\":0.6,\"end\":1.06}]"} {"utterance_id": "4831-18525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0021.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.31, 4.87, "nose"]], "response_mode": "offline", "answer": "[{\"word\":\"nose\",\"start\":4.31,\"end\":4.87}]"} {"utterance_id": "6467-94831-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0025.flac", "task_type": "lexical_detection", "query": "sure", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"sure\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.13, 2.38, "sure"]], "response_mode": "offline", "answer": "[{\"word\":\"sure\",\"start\":2.13,\"end\":2.38}]"} {"utterance_id": "4153-185072-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0005.flac", "task_type": "lexical_detection", "query": "nettles", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"nettles\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.5, 10.05, "nettles"]], "response_mode": "offline", "answer": "[{\"word\":\"nettles\",\"start\":9.5,\"end\":10.05}]"} {"utterance_id": "4323-18416-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0019.flac", "task_type": "lexical_detection", "query": "wasn't", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wasn't\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.32, 2.64, "wasn't"]], "response_mode": "offline", "answer": "[{\"word\":\"wasn't\",\"start\":2.32,\"end\":2.64}]"} {"utterance_id": "4831-18525-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0004.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.39, 5.87, "paris"]], "response_mode": "offline", "answer": "[{\"word\":\"paris\",\"start\":5.39,\"end\":5.87}]"} {"utterance_id": "4572-64670-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.11, 7.55, "rama"]], "response_mode": "offline", "answer": "[{\"word\":\"rama\",\"start\":7.11,\"end\":7.55}]"} {"utterance_id": "6123-59150-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0036.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.04, 1.87, "one"], [3.02, 3.72, "two"], [6.16, 6.97, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.04,\"end\":1.87},{\"word\":\"two\",\"start\":3.02,\"end\":3.72},{\"word\":\"three\",\"start\":6.16,\"end\":6.97}]"} {"utterance_id": "7601-101622-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.79, 3.75, "king philip"], [5.57, 6.08, "john"]], "response_mode": "offline", "answer": "[{\"word\":\"king philip\",\"start\":2.79,\"end\":3.75},{\"word\":\"john\",\"start\":5.57,\"end\":6.08}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.18, 7.56, "bodies"], [8.79, 9.13, "arms"]], "response_mode": "offline", "answer": "[{\"word\":\"bodies\",\"start\":7.18,\"end\":7.56},{\"word\":\"arms\",\"start\":8.79,\"end\":9.13}]"} {"utterance_id": "3663-172528-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0018.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.58, 6.08, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":5.58,\"end\":6.08}]"} {"utterance_id": "116-288048-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0027.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.44, 1.96, "prophets"]], "response_mode": "offline", "answer": "[{\"word\":\"prophets\",\"start\":1.44,\"end\":1.96}]"} {"utterance_id": "4831-29134-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0014.flac", "task_type": "lexical_counting", "query": "position", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"position\" (word, start, end). If none, reply [].", "labels": [[2.32, 2.92, "one"], [4.06, 4.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.32,\"end\":2.92},{\"word\":\"two\",\"start\":4.06,\"end\":4.66}]"} {"utterance_id": "5849-50964-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0005.flac", "task_type": "lexical_detection", "query": "hat", "prompt": "After listening to the whole recording, output every instance of the word \"hat\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.72, 4.13, "hat"]], "response_mode": "offline", "answer": "[{\"word\":\"hat\",\"start\":3.72,\"end\":4.13}]"} {"utterance_id": "2506-169427-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0000.flac", "task_type": "lexical_counting", "query": "interest", "prompt": "Once the audio has finished, list each occurrence of the word \"interest\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.2, 0.76, "one"], [20.07, 20.5, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.2,\"end\":0.76},{\"word\":\"two\",\"start\":20.07,\"end\":20.5}]"} {"utterance_id": "1255-74899-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[3.26, 3.59, "hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"hundred\",\"start\":3.26,\"end\":3.59}]"} {"utterance_id": "2506-169427-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0003.flac", "task_type": "lexical_detection", "query": "uttermost", "prompt": "Process the full audio offline. Find every instance of the word \"uttermost\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.98, 13.91, "uttermost"]], "response_mode": "offline", "answer": "[{\"word\":\"uttermost\",\"start\":12.98,\"end\":13.91}]"} {"utterance_id": "116-288048-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0014.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen carefully to the entire clip, then report every instance of a city name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 1.86, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":1.52,\"end\":1.86}]"} {"utterance_id": "8288-274162-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0031.flac", "task_type": "lexical_detection", "query": "obey", "prompt": "When the utterance is complete, report all instances of the word \"obey\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.65, "obey"]], "response_mode": "offline", "answer": "[{\"word\":\"obey\",\"start\":4.32,\"end\":4.65}]"} {"utterance_id": "6599-38591-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0002.flac", "task_type": "semantic_counting", "query": "date", "prompt": "After listening to the whole recording, output timed count words for a date (one, two, three, ...) as JSON. If none, reply [].", "labels": [[2.61, 3.67, "one"], [4.69, 5.99, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.61,\"end\":3.67},{\"word\":\"two\",\"start\":4.69,\"end\":5.99}]"} {"utterance_id": "3915-98647-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0025.flac", "task_type": "semantic_detection", "query": "food", "prompt": "You will hear a complete utterance. When it ends, list every instance of a food name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.61, 5.91, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":5.61,\"end\":5.91}]"} {"utterance_id": "6123-59150-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0046.flac", "task_type": "lexical_detection", "query": "figure", "prompt": "When the utterance is complete, report all instances of the word \"figure\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.98, "figure"]], "response_mode": "offline", "answer": "[{\"word\":\"figure\",\"start\":0.68,\"end\":0.98}]"} {"utterance_id": "700-122867-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.31, 0.47, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":0.31,\"end\":0.47}]"} {"utterance_id": "8254-115543-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0014.flac", "task_type": "lexical_detection", "query": "sacred", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"sacred\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.87, 5.21, "sacred"]], "response_mode": "offline", "answer": "[{\"word\":\"sacred\",\"start\":4.87,\"end\":5.21}]"} {"utterance_id": "6841-88294-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0057.flac", "task_type": "lexical_detection", "query": "save", "prompt": "Process the full audio offline. Find every instance of the word \"save\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 0.8, "save"]], "response_mode": "offline", "answer": "[{\"word\":\"save\",\"start\":0.56,\"end\":0.8}]"} {"utterance_id": "8254-115543-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0018.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.66, 2.08, "farmer"]], "response_mode": "offline", "answer": "[{\"word\":\"farmer\",\"start\":1.66,\"end\":2.08}]"} {"utterance_id": "8254-84205-0070", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0070.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.05, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":0.73,\"end\":1.05}]"} {"utterance_id": "4572-112375-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0015.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.3, 1.86, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":1.3,\"end\":1.86}]"} {"utterance_id": "6841-88294-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0058.flac", "task_type": "lexical_detection", "query": "solomon", "prompt": "Once the audio has finished, identify every instance of the word \"solomon\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.7, 2.22, "solomon"]], "response_mode": "offline", "answer": "[{\"word\":\"solomon\",\"start\":1.7,\"end\":2.22}]"} {"utterance_id": "6123-59150-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0002.flac", "task_type": "lexical_detection", "query": "novelist", "prompt": "Process the full audio offline. Find every instance of the word \"novelist\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 1.91, "novelist"]], "response_mode": "offline", "answer": "[{\"word\":\"novelist\",\"start\":1.48,\"end\":1.91}]"} {"utterance_id": "1255-74899-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0017.flac", "task_type": "lexical_detection", "query": "really", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"really\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.03, "really"]], "response_mode": "offline", "answer": "[{\"word\":\"really\",\"start\":0.81,\"end\":1.03}]"} {"utterance_id": "3660-6517-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0030.flac", "task_type": "lexical_detection", "query": "roughly", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"roughly\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.43, "roughly"]], "response_mode": "offline", "answer": "[{\"word\":\"roughly\",\"start\":3.06,\"end\":3.43}]"} {"utterance_id": "4153-186223-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0013.flac", "task_type": "lexical_counting", "query": "stopped", "prompt": "When the utterance is complete, report how the word \"stopped\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.74, 1.04, "one"], [2.3, 2.73, "two"], [6.47, 6.88, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.74,\"end\":1.04},{\"word\":\"two\",\"start\":2.3,\"end\":2.73},{\"word\":\"three\",\"start\":6.47,\"end\":6.88}]"} {"utterance_id": "1686-142278-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0045.flac", "task_type": "lexical_detection", "query": "stay", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"stay\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.84, 1.09, "stay"]], "response_mode": "offline", "answer": "[{\"word\":\"stay\",\"start\":0.84,\"end\":1.09}]"} {"utterance_id": "4515-11057-0076", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0076.flac", "task_type": "lexical_detection", "query": "hour", "prompt": "When the utterance is complete, report all instances of the word \"hour\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 1.92, "hour"]], "response_mode": "offline", "answer": "[{\"word\":\"hour\",\"start\":1.51,\"end\":1.92}]"} {"utterance_id": "4831-29134-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0012.flac", "task_type": "lexical_counting", "query": "love", "prompt": "After listening to the whole recording, output timed count words for the word \"love\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.6, 0.77, "one"], [2.22, 2.46, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.6,\"end\":0.77},{\"word\":\"two\",\"start\":2.22,\"end\":2.46}]"} {"utterance_id": "8288-274162-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0011.flac", "task_type": "lexical_detection", "query": "say", "prompt": "After listening to the whole recording, output every instance of the word \"say\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.96, "say"]], "response_mode": "offline", "answer": "[{\"word\":\"say\",\"start\":1.47,\"end\":1.96}]"} {"utterance_id": "1650-167613-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0052.flac", "task_type": "lexical_detection", "query": "concluded", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"concluded\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 3.17, "concluded"]], "response_mode": "offline", "answer": "[{\"word\":\"concluded\",\"start\":2.4,\"end\":3.17}]"} {"utterance_id": "116-288048-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0011.flac", "task_type": "lexical_detection", "query": "matthew", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"matthew\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.26, "matthew"]], "response_mode": "offline", "answer": "[{\"word\":\"matthew\",\"start\":0.75,\"end\":1.26}]"} {"utterance_id": "1630-141772-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[16.57, 17.04, "mary"]], "response_mode": "offline", "answer": "[{\"word\":\"mary\",\"start\":16.57,\"end\":17.04}]"} {"utterance_id": "1650-167613-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0043.flac", "task_type": "lexical_counting", "query": "war", "prompt": "Process the full audio offline. For every occurrence of the word \"war\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.6, "one"], [3.4, 3.95, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.25,\"end\":1.6},{\"word\":\"two\",\"start\":3.4,\"end\":3.95}]"} {"utterance_id": "5543-27761-0083", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0083.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.78, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":0.49,\"end\":0.78}]"} {"utterance_id": "3915-98647-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0002.flac", "task_type": "lexical_detection", "query": "carefully", "prompt": "After the audio ends, extract all occurrences of the word \"carefully\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.88, "carefully"]], "response_mode": "offline", "answer": "[{\"word\":\"carefully\",\"start\":1.46,\"end\":1.88}]"} {"utterance_id": "4570-56594-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0006.flac", "task_type": "lexical_detection", "query": "git", "prompt": "When the utterance is complete, report all instances of the word \"git\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.19, 3.32, "git"]], "response_mode": "offline", "answer": "[{\"word\":\"git\",\"start\":3.19,\"end\":3.32}]"} {"utterance_id": "6455-66379-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0019.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.68, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.22,\"end\":1.68}]"} {"utterance_id": "7697-105815-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0010.flac", "task_type": "lexical_detection", "query": "grace", "prompt": "When the utterance is complete, report all instances of the word \"grace\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.11, 8.34, "grace"]], "response_mode": "offline", "answer": "[{\"word\":\"grace\",\"start\":8.11,\"end\":8.34}]"} {"utterance_id": "4153-186223-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0007.flac", "task_type": "lexical_detection", "query": "rich", "prompt": "Process the full audio offline. Find every instance of the word \"rich\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.46, "rich"]], "response_mode": "offline", "answer": "[{\"word\":\"rich\",\"start\":1.0,\"end\":1.46}]"} {"utterance_id": "3660-172182-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0035.flac", "task_type": "lexical_detection", "query": "desired", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"desired\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 2.0, "desired"]], "response_mode": "offline", "answer": "[{\"word\":\"desired\",\"start\":1.46,\"end\":2.0}]"} {"utterance_id": "4831-25894-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0006.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.33, 4.57, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":4.33,\"end\":4.57}]"} {"utterance_id": "4570-56594-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0003.flac", "task_type": "lexical_detection", "query": "gentleman", "prompt": "After listening to the whole recording, output every instance of the word \"gentleman\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.41, 2.84, "gentleman"]], "response_mode": "offline", "answer": "[{\"word\":\"gentleman\",\"start\":2.41,\"end\":2.84}]"} {"utterance_id": "1585-131718-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0039.flac", "task_type": "lexical_detection", "query": "result", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"result\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.96, 7.47, "result"]], "response_mode": "offline", "answer": "[{\"word\":\"result\",\"start\":6.96,\"end\":7.47}]"} {"utterance_id": "6455-67803-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0021.flac", "task_type": "lexical_detection", "query": "aroused", "prompt": "Process the full audio offline. Find every instance of the word \"aroused\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.95, "aroused"]], "response_mode": "offline", "answer": "[{\"word\":\"aroused\",\"start\":3.33,\"end\":3.95}]"} {"utterance_id": "3660-6517-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0033.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 0.73, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.57,\"end\":0.73}]"} {"utterance_id": "8254-115543-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0009.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.24, 3.66, "oxen"]], "response_mode": "offline", "answer": "[{\"word\":\"oxen\",\"start\":3.24,\"end\":3.66}]"} {"utterance_id": "1701-141760-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0033.flac", "task_type": "lexical_detection", "query": "berg", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"berg\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.22, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":3.82,\"end\":4.22}]"} {"utterance_id": "1650-167613-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0048.flac", "task_type": "lexical_detection", "query": "taste", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"taste\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.96, 2.35, "taste"]], "response_mode": "offline", "answer": "[{\"word\":\"taste\",\"start\":1.96,\"end\":2.35}]"} {"utterance_id": "700-122868-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0033.flac", "task_type": "lexical_detection", "query": "right", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"right\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 1.06, "right"]], "response_mode": "offline", "answer": "[{\"word\":\"right\",\"start\":0.71,\"end\":1.06}]"} {"utterance_id": "3660-6517-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0018.flac", "task_type": "lexical_detection", "query": "quantity", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"quantity\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.93, "quantity"]], "response_mode": "offline", "answer": "[{\"word\":\"quantity\",\"start\":0.48,\"end\":0.93}]"} {"utterance_id": "7641-96670-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0015.flac", "task_type": "lexical_detection", "query": "divided", "prompt": "When the utterance is complete, report all instances of the word \"divided\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.35, "divided"]], "response_mode": "offline", "answer": "[{\"word\":\"divided\",\"start\":0.85,\"end\":1.35}]"} {"utterance_id": "116-288048-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0021.flac", "task_type": "lexical_detection", "query": "decide", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"decide\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.27, 2.79, "decide"]], "response_mode": "offline", "answer": "[{\"word\":\"decide\",\"start\":2.27,\"end\":2.79}]"} {"utterance_id": "6841-88294-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0024.flac", "task_type": "lexical_detection", "query": "fit", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fit\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.38, 3.81, "fit"]], "response_mode": "offline", "answer": "[{\"word\":\"fit\",\"start\":3.38,\"end\":3.81}]"} {"utterance_id": "3660-172182-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.96, 2.38, "arthur"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur\",\"start\":1.96,\"end\":2.38}]"} {"utterance_id": "3663-172528-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0019.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen carefully to the entire clip, then report every instance of a time of day as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.46, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":0.95,\"end\":1.46}]"} {"utterance_id": "4515-11057-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0043.flac", "task_type": "lexical_detection", "query": "trouble", "prompt": "When the utterance is complete, report all instances of the word \"trouble\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.86, 3.33, "trouble"]], "response_mode": "offline", "answer": "[{\"word\":\"trouble\",\"start\":2.86,\"end\":3.33}]"} {"utterance_id": "6467-56885-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0010.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "When the utterance is complete, report all instances of a weather term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.72, 12.03, "mist"]], "response_mode": "offline", "answer": "[{\"word\":\"mist\",\"start\":11.72,\"end\":12.03}]"} {"utterance_id": "1630-96099-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0021.flac", "task_type": "lexical_counting", "query": "hundred", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"hundred\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.25, 1.91, "one"], [3.48, 4.11, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.25,\"end\":1.91},{\"word\":\"two\",\"start\":3.48,\"end\":4.11}]"} {"utterance_id": "4831-18525-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.47, "polly"], [2.63, 3.04, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":0.16,\"end\":0.47},{\"word\":\"jasper\",\"start\":2.63,\"end\":3.04}]"} {"utterance_id": "8173-294714-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0027.flac", "task_type": "lexical_detection", "query": "infatuation", "prompt": "Once the audio has finished, identify every instance of the word \"infatuation\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.46, 9.48, "infatuation"]], "response_mode": "offline", "answer": "[{\"word\":\"infatuation\",\"start\":8.46,\"end\":9.48}]"} {"utterance_id": "1650-167613-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0023.flac", "task_type": "lexical_detection", "query": "firmly", "prompt": "After the audio ends, extract all occurrences of the word \"firmly\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.28, 2.7, "firmly"]], "response_mode": "offline", "answer": "[{\"word\":\"firmly\",\"start\":2.28,\"end\":2.7}]"} {"utterance_id": "1585-131718-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0046.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After the audio ends, extract all occurrences of a time of day. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.94, 5.33, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":4.94,\"end\":5.33}]"} {"utterance_id": "6123-59150-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0021.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "When the utterance is complete, report all instances of an ordinal number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.28, 4.61, "eldest"]], "response_mode": "offline", "answer": "[{\"word\":\"eldest\",\"start\":4.28,\"end\":4.61}]"} {"utterance_id": "4515-11057-0107", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0107.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.4, 3.79, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.4,\"end\":3.79}]"} {"utterance_id": "4831-18525-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0003.flac", "task_type": "lexical_counting", "query": "tom", "prompt": "After the audio ends, tally every instance of the word \"tom\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.9, 2.25, "one"], [5.12, 5.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.9,\"end\":2.25},{\"word\":\"two\",\"start\":5.12,\"end\":5.56}]"} {"utterance_id": "4831-29134-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0010.flac", "task_type": "lexical_detection", "query": "wise", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"wise\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.83, "wise"]], "response_mode": "offline", "answer": "[{\"word\":\"wise\",\"start\":3.38,\"end\":3.83}]"} {"utterance_id": "1686-142278-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0053.flac", "task_type": "lexical_detection", "query": "milton", "prompt": "After the audio ends, extract all occurrences of the word \"milton\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.02, "milton"]], "response_mode": "offline", "answer": "[{\"word\":\"milton\",\"start\":0.5,\"end\":1.02}]"} {"utterance_id": "6599-38590-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0008.flac", "task_type": "lexical_counting", "query": "peace", "prompt": "Once the audio has finished, list each occurrence of the word \"peace\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.61, 0.88, "one"], [2.67, 2.95, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.61,\"end\":0.88},{\"word\":\"two\",\"start\":2.67,\"end\":2.95}]"} {"utterance_id": "5849-50962-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.07, 3.37, "nellie"]], "response_mode": "offline", "answer": "[{\"word\":\"nellie\",\"start\":3.07,\"end\":3.37}]"} {"utterance_id": "116-288045-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.03, 1.57, "homer"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":1.03,\"end\":1.57}]"} {"utterance_id": "3915-98647-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0032.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.69, 3.17, "flowers"]], "response_mode": "offline", "answer": "[{\"word\":\"flowers\",\"start\":2.69,\"end\":3.17}]"} {"utterance_id": "1630-141772-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0019.flac", "task_type": "lexical_detection", "query": "address", "prompt": "After listening to the whole recording, output every instance of the word \"address\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.43, 6.06, "address"]], "response_mode": "offline", "answer": "[{\"word\":\"address\",\"start\":5.43,\"end\":6.06}]"} {"utterance_id": "6123-59150-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0037.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.48, "one"], [4.46, 5.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.57,\"end\":2.48},{\"word\":\"two\",\"start\":4.46,\"end\":5.6}]"} {"utterance_id": "6267-65525-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0053.flac", "task_type": "lexical_detection", "query": "lincoln", "prompt": "When the utterance is complete, report all instances of the word \"lincoln\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.56, 1.01, "lincoln"]], "response_mode": "offline", "answer": "[{\"word\":\"lincoln\",\"start\":0.56,\"end\":1.01}]"} {"utterance_id": "6467-97061-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0014.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After the audio ends, extract all occurrences of an animal name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.13, 5.52, "swan"]], "response_mode": "offline", "answer": "[{\"word\":\"swan\",\"start\":5.13,\"end\":5.52}]"} {"utterance_id": "6841-88291-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0010.flac", "task_type": "lexical_detection", "query": "sweep", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sweep\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.37, "sweep"]], "response_mode": "offline", "answer": "[{\"word\":\"sweep\",\"start\":2.95,\"end\":3.37}]"} {"utterance_id": "116-288045-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0018.flac", "task_type": "lexical_detection", "query": "homer", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"homer\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.6, 2.22, "homer"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":1.6,\"end\":2.22}]"} {"utterance_id": "5849-50963-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0011.flac", "task_type": "lexical_detection", "query": "maybe", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"maybe\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.22, 0.6, "maybe"]], "response_mode": "offline", "answer": "[{\"word\":\"maybe\",\"start\":0.22,\"end\":0.6}]"} {"utterance_id": "4572-64670-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.9, 7.46, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":6.9,\"end\":7.46}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[13.3, 13.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":13.3,\"end\":13.63}]"} {"utterance_id": "4323-55228-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0002.flac", "task_type": "lexical_detection", "query": "billy's", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"billy's\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.5, 6.84, "billy's"]], "response_mode": "offline", "answer": "[{\"word\":\"billy's\",\"start\":6.5,\"end\":6.84}]"} {"utterance_id": "5849-50963-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0009.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "You will hear a complete utterance. When it ends, list every instance of an organization name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.75, 8.1, "firm"]], "response_mode": "offline", "answer": "[{\"word\":\"firm\",\"start\":7.75,\"end\":8.1}]"} {"utterance_id": "8254-115543-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0021.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a country name with fields word, start, end (seconds). If none, reply [].", "labels": [[9.8, 10.18, "india"]], "response_mode": "offline", "answer": "[{\"word\":\"india\",\"start\":9.8,\"end\":10.18}]"} {"utterance_id": "116-288045-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0003.flac", "task_type": "lexical_detection", "query": "other's", "prompt": "Process the full audio offline. Find every instance of the word \"other's\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.63, "other's"]], "response_mode": "offline", "answer": "[{\"word\":\"other's\",\"start\":2.39,\"end\":2.63}]"} {"utterance_id": "1701-141760-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0030.flac", "task_type": "lexical_detection", "query": "presented", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"presented\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.07, 4.56, "presented"]], "response_mode": "offline", "answer": "[{\"word\":\"presented\",\"start\":4.07,\"end\":4.56}]"} {"utterance_id": "7641-96670-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0010.flac", "task_type": "lexical_detection", "query": "heroes", "prompt": "When the utterance is complete, report all instances of the word \"heroes\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.31, 2.61, "heroes"]], "response_mode": "offline", "answer": "[{\"word\":\"heroes\",\"start\":2.31,\"end\":2.61}]"} {"utterance_id": "8173-294714-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0041.flac", "task_type": "lexical_detection", "query": "excellent", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"excellent\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.66, 6.18, "excellent"]], "response_mode": "offline", "answer": "[{\"word\":\"excellent\",\"start\":5.66,\"end\":6.18}]"} {"utterance_id": "7601-101619-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0003.flac", "task_type": "lexical_counting", "query": "serve", "prompt": "Listen to the full audio. Count every occurrence of the word \"serve\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.27, 8.65, "one"], [9.95, 10.29, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.27,\"end\":8.65},{\"word\":\"two\",\"start\":9.95,\"end\":10.29}]"} {"utterance_id": "4323-13259-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0007.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.79, 1.06, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.79,\"end\":1.06}]"} {"utterance_id": "8254-84205-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0006.flac", "task_type": "lexical_detection", "query": "pony", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"pony\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.16, "pony"]], "response_mode": "offline", "answer": "[{\"word\":\"pony\",\"start\":0.83,\"end\":1.16}]"} {"utterance_id": "1630-73710-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0007.flac", "task_type": "lexical_detection", "query": "day", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"day\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.26, "day"]], "response_mode": "offline", "answer": "[{\"word\":\"day\",\"start\":1.9,\"end\":2.26}]"} {"utterance_id": "1255-90407-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0013.flac", "task_type": "lexical_detection", "query": "evening", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"evening\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.94, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":1.56,\"end\":1.94}]"} {"utterance_id": "1651-136854-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0010.flac", "task_type": "lexical_counting", "query": "different", "prompt": "Listen to the full audio. Count every occurrence of the word \"different\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.89, "one"], [2.69, 3.01, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.51,\"end\":0.89},{\"word\":\"two\",\"start\":2.69,\"end\":3.01}]"} {"utterance_id": "4323-18416-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0012.flac", "task_type": "lexical_detection", "query": "dear", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"dear\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.07, "dear"]], "response_mode": "offline", "answer": "[{\"word\":\"dear\",\"start\":0.86,\"end\":1.07}]"} {"utterance_id": "6455-66379-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0015.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "You will hear a complete utterance. When it ends, list every instance of a time of day you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.86, 4.3, "daylight"]], "response_mode": "offline", "answer": "[{\"word\":\"daylight\",\"start\":3.86,\"end\":4.3}]"} {"utterance_id": "1651-136854-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0029.flac", "task_type": "lexical_detection", "query": "wonderfully", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"wonderfully\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.83, 5.37, "wonderfully"]], "response_mode": "offline", "answer": "[{\"word\":\"wonderfully\",\"start\":4.83,\"end\":5.37}]"} {"utterance_id": "3915-98647-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0001.flac", "task_type": "lexical_detection", "query": "rice", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"rice\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.74, 2.12, "rice"]], "response_mode": "offline", "answer": "[{\"word\":\"rice\",\"start\":1.74,\"end\":2.12}]"} {"utterance_id": "6267-65525-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0029.flac", "task_type": "lexical_detection", "query": "sidney", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"sidney\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.2, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":1.55,\"end\":2.2}]"} {"utterance_id": "1701-141760-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0015.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After listening to the whole recording, output every instance of a country name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.53, 2.13, "poland"]], "response_mode": "offline", "answer": "[{\"word\":\"poland\",\"start\":1.53,\"end\":2.13}]"} {"utterance_id": "1686-142278-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0006.flac", "task_type": "lexical_detection", "query": "offer", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"offer\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.45, 2.86, "offer"]], "response_mode": "offline", "answer": "[{\"word\":\"offer\",\"start\":2.45,\"end\":2.86}]"} {"utterance_id": "8288-274162-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0040.flac", "task_type": "lexical_detection", "query": "stupid", "prompt": "Once the audio has finished, identify every instance of the word \"stupid\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.89, 1.31, "stupid"]], "response_mode": "offline", "answer": "[{\"word\":\"stupid\",\"start\":0.89,\"end\":1.31}]"} {"utterance_id": "3660-172182-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0030.flac", "task_type": "lexical_detection", "query": "mission", "prompt": "After listening to the whole recording, output every instance of the word \"mission\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.25, "mission"]], "response_mode": "offline", "answer": "[{\"word\":\"mission\",\"start\":1.83,\"end\":2.25}]"} {"utterance_id": "5849-50964-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0001.flac", "task_type": "lexical_detection", "query": "nan", "prompt": "After listening to the whole recording, output every instance of the word \"nan\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.55, "nan"]], "response_mode": "offline", "answer": "[{\"word\":\"nan\",\"start\":2.26,\"end\":2.55}]"} {"utterance_id": "3663-172528-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0009.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.7, 2.35, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":1.7,\"end\":2.35}]"} {"utterance_id": "6841-88291-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.46, 2.92, "horn"]], "response_mode": "offline", "answer": "[{\"word\":\"horn\",\"start\":2.46,\"end\":2.92}]"} {"utterance_id": "1630-141772-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.0, 11.43, "chest"]], "response_mode": "offline", "answer": "[{\"word\":\"chest\",\"start\":11.0,\"end\":11.43}]"} {"utterance_id": "6455-66379-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.15, "pilot's"]], "response_mode": "offline", "answer": "[{\"word\":\"pilot's\",\"start\":0.77,\"end\":1.15}]"} {"utterance_id": "3660-172182-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0028.flac", "task_type": "lexical_detection", "query": "nephew", "prompt": "After listening to the whole recording, output every instance of the word \"nephew\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.19, "nephew"]], "response_mode": "offline", "answer": "[{\"word\":\"nephew\",\"start\":2.81,\"end\":3.19}]"} {"utterance_id": "4323-18416-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0031.flac", "task_type": "lexical_detection", "query": "pickering", "prompt": "When the utterance is complete, report all instances of the word \"pickering\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.82, "pickering"]], "response_mode": "offline", "answer": "[{\"word\":\"pickering\",\"start\":2.39,\"end\":2.82}]"} {"utterance_id": "4323-18416-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.51, 2.0, "mister king"]], "response_mode": "offline", "answer": "[{\"word\":\"mister king\",\"start\":1.51,\"end\":2.0}]"} {"utterance_id": "7697-245715-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[11.45, 11.79, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":11.45,\"end\":11.79}]"} {"utterance_id": "4831-18525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.39, 6.68, "heels"]], "response_mode": "offline", "answer": "[{\"word\":\"heels\",\"start\":6.39,\"end\":6.68}]"} {"utterance_id": "7697-245712-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0019.flac", "task_type": "lexical_counting", "query": "sense", "prompt": "Once the audio has finished, list each occurrence of the word \"sense\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.75, 1.12, "one"], [2.91, 3.2, "two"], [5.27, 5.64, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.75,\"end\":1.12},{\"word\":\"two\",\"start\":2.91,\"end\":3.2},{\"word\":\"three\",\"start\":5.27,\"end\":5.64}]"} {"utterance_id": "7697-245712-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0012.flac", "task_type": "lexical_detection", "query": "dominion", "prompt": "After the audio ends, extract all occurrences of the word \"dominion\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.22, 6.66, "dominion"]], "response_mode": "offline", "answer": "[{\"word\":\"dominion\",\"start\":6.22,\"end\":6.66}]"} {"utterance_id": "7697-105815-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0045.flac", "task_type": "lexical_detection", "query": "mount", "prompt": "When the utterance is complete, report all instances of the word \"mount\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.42, 5.72, "mount"]], "response_mode": "offline", "answer": "[{\"word\":\"mount\",\"start\":5.42,\"end\":5.72}]"} {"utterance_id": "4323-55228-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0036.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Process the full audio offline. Find every instance of a time of day and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.98, 7.32, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":6.98,\"end\":7.32}]"} {"utterance_id": "1630-73710-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0012.flac", "task_type": "lexical_detection", "query": "bless", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"bless\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.75, 1.12, "bless"]], "response_mode": "offline", "answer": "[{\"word\":\"bless\",\"start\":0.75,\"end\":1.12}]"} {"utterance_id": "2506-11278-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0012.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.41, 5.06, "victuals"]], "response_mode": "offline", "answer": "[{\"word\":\"victuals\",\"start\":4.41,\"end\":5.06}]"} {"utterance_id": "4323-18416-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.46, 3.31, "charlotte chatterton"]], "response_mode": "offline", "answer": "[{\"word\":\"charlotte chatterton\",\"start\":2.46,\"end\":3.31}]"} {"utterance_id": "7641-96684-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0033.flac", "task_type": "lexical_detection", "query": "held", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"held\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.9, 1.2, "held"]], "response_mode": "offline", "answer": "[{\"word\":\"held\",\"start\":0.9,\"end\":1.2}]"} {"utterance_id": "6267-65525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0013.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.12, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.79,\"end\":1.12}]"} {"utterance_id": "4323-55228-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0034.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.28, 2.63, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":2.28,\"end\":2.63}]"} {"utterance_id": "6455-67804-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0008.flac", "task_type": "lexical_detection", "query": "afraid", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"afraid\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.4, "afraid"]], "response_mode": "offline", "answer": "[{\"word\":\"afraid\",\"start\":1.05,\"end\":1.4}]"} {"utterance_id": "1630-73710-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0020.flac", "task_type": "lexical_detection", "query": "restless", "prompt": "After listening to the whole recording, output every instance of the word \"restless\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.65, "restless"]], "response_mode": "offline", "answer": "[{\"word\":\"restless\",\"start\":2.93,\"end\":3.65}]"} {"utterance_id": "6467-94831-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0012.flac", "task_type": "lexical_detection", "query": "worst", "prompt": "When the utterance is complete, report all instances of the word \"worst\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.29, 3.61, "worst"]], "response_mode": "offline", "answer": "[{\"word\":\"worst\",\"start\":3.29,\"end\":3.61}]"} {"utterance_id": "700-122866-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0031.flac", "task_type": "lexical_detection", "query": "allowed", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"allowed\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.05, 1.41, "allowed"]], "response_mode": "offline", "answer": "[{\"word\":\"allowed\",\"start\":1.05,\"end\":1.41}]"} {"utterance_id": "8288-274162-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0000.flac", "task_type": "lexical_detection", "query": "buried", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"buried\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.7, "buried"]], "response_mode": "offline", "answer": "[{\"word\":\"buried\",\"start\":1.36,\"end\":1.7}]"} {"utterance_id": "3660-172183-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.48, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.14,\"end\":1.48}]"} {"utterance_id": "1686-142278-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0015.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.16, 1.61, "papa"]], "response_mode": "offline", "answer": "[{\"word\":\"papa\",\"start\":1.16,\"end\":1.61}]"} {"utterance_id": "7697-105815-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0039.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.27, 4.8, "beasts"]], "response_mode": "offline", "answer": "[{\"word\":\"beasts\",\"start\":4.27,\"end\":4.8}]"} {"utterance_id": "5849-50873-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0024.flac", "task_type": "lexical_detection", "query": "candles", "prompt": "Process the full audio offline. Find every instance of the word \"candles\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.54, "candles"]], "response_mode": "offline", "answer": "[{\"word\":\"candles\",\"start\":0.86,\"end\":1.54}]"} {"utterance_id": "1686-142278-0086", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0086.flac", "task_type": "lexical_detection", "query": "deed", "prompt": "After listening to the whole recording, output every instance of the word \"deed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.06, 7.26, "deed"]], "response_mode": "offline", "answer": "[{\"word\":\"deed\",\"start\":7.06,\"end\":7.26}]"} {"utterance_id": "1650-167613-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0003.flac", "task_type": "lexical_detection", "query": "troops", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"troops\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.91, 3.38, "troops"]], "response_mode": "offline", "answer": "[{\"word\":\"troops\",\"start\":2.91,\"end\":3.38}]"} {"utterance_id": "1686-142278-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0049.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.62, 4.09, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":3.62,\"end\":4.09}]"} {"utterance_id": "1686-142278-0056", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0056.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.49, 1.93, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":1.49,\"end\":1.93}]"} {"utterance_id": "7641-96684-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0009.flac", "task_type": "semantic_detection", "query": "color", "prompt": "You will hear a complete utterance. When it ends, list every instance of a color name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.89, 6.15, "blue"], [7.56, 7.86, "green"]], "response_mode": "offline", "answer": "[{\"word\":\"blue\",\"start\":5.89,\"end\":6.15},{\"word\":\"green\",\"start\":7.56,\"end\":7.86}]"} {"utterance_id": "1585-131718-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0052.flac", "task_type": "lexical_detection", "query": "cure", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"cure\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.04, 4.42, "cure"]], "response_mode": "offline", "answer": "[{\"word\":\"cure\",\"start\":4.04,\"end\":4.42}]"} {"utterance_id": "1650-167613-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0032.flac", "task_type": "lexical_detection", "query": "andrew", "prompt": "After listening to the whole recording, output every instance of the word \"andrew\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.57, "andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"andrew\",\"start\":2.14,\"end\":2.57}]"} {"utterance_id": "1255-138279-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0006.flac", "task_type": "lexical_detection", "query": "sudden", "prompt": "After listening to the whole recording, output every instance of the word \"sudden\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.38, 5.84, "sudden"]], "response_mode": "offline", "answer": "[{\"word\":\"sudden\",\"start\":5.38,\"end\":5.84}]"} {"utterance_id": "1255-90407-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0008.flac", "task_type": "lexical_detection", "query": "residence", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"residence\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.57, 3.02, "residence"]], "response_mode": "offline", "answer": "[{\"word\":\"residence\",\"start\":2.57,\"end\":3.02}]"} {"utterance_id": "3660-6517-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0022.flac", "task_type": "lexical_detection", "query": "generally", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"generally\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.86, "generally"]], "response_mode": "offline", "answer": "[{\"word\":\"generally\",\"start\":0.52,\"end\":0.86}]"} {"utterance_id": "4323-13259-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0011.flac", "task_type": "lexical_detection", "query": "meetings", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"meetings\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.6, 5.96, "meetings"]], "response_mode": "offline", "answer": "[{\"word\":\"meetings\",\"start\":5.6,\"end\":5.96}]"} {"utterance_id": "7601-101622-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.29, 2.11, "blood"]], "response_mode": "offline", "answer": "[{\"word\":\"blood\",\"start\":1.29,\"end\":2.11}]"} {"utterance_id": "3915-98647-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0004.flac", "task_type": "semantic_counting", "query": "kinship", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a family relation with word/start/end. If none, reply [].", "labels": [[2.24, 2.61, "one"], [8.1, 8.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.24,\"end\":2.61},{\"word\":\"two\",\"start\":8.1,\"end\":8.53}]"} {"utterance_id": "700-122868-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0028.flac", "task_type": "lexical_detection", "query": "explained", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"explained\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.49, "explained"]], "response_mode": "offline", "answer": "[{\"word\":\"explained\",\"start\":1.95,\"end\":2.49}]"} {"utterance_id": "6455-67803-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0011.flac", "task_type": "lexical_detection", "query": "knew", "prompt": "After listening to the whole recording, output every instance of the word \"knew\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.39, "knew"]], "response_mode": "offline", "answer": "[{\"word\":\"knew\",\"start\":1.25,\"end\":1.39}]"} {"utterance_id": "1651-136854-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0015.flac", "task_type": "lexical_detection", "query": "crushing", "prompt": "Process the full audio offline. Find every instance of the word \"crushing\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.85, "crushing"]], "response_mode": "offline", "answer": "[{\"word\":\"crushing\",\"start\":2.35,\"end\":2.85}]"} {"utterance_id": "1650-173551-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 1.88, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":1.54,\"end\":1.88}]"} {"utterance_id": "7697-105815-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.14, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"christian\",\"start\":0.65,\"end\":1.14}]"} {"utterance_id": "6841-88294-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0066.flac", "task_type": "lexical_detection", "query": "larger", "prompt": "After listening to the whole recording, output every instance of the word \"larger\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.51, "larger"]], "response_mode": "offline", "answer": "[{\"word\":\"larger\",\"start\":1.19,\"end\":1.51}]"} {"utterance_id": "1630-141772-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 1.36, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":0.48,\"end\":1.36}]"} {"utterance_id": "116-288046-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0003.flac", "task_type": "lexical_detection", "query": "freedom", "prompt": "After listening to the whole recording, output every instance of the word \"freedom\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.08, 3.55, "freedom"]], "response_mode": "offline", "answer": "[{\"word\":\"freedom\",\"start\":3.08,\"end\":3.55}]"} {"utterance_id": "116-288047-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0000.flac", "task_type": "lexical_detection", "query": "phenomenon", "prompt": "Once the audio has finished, identify every instance of the word \"phenomenon\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.21, 2.91, "phenomenon"]], "response_mode": "offline", "answer": "[{\"word\":\"phenomenon\",\"start\":2.21,\"end\":2.91}]"} {"utterance_id": "4570-24733-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.94, 6.51, "god"], [9.56, 10.12, "god"], [24.85, 25.23, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":5.94,\"end\":6.51},{\"word\":\"god\",\"start\":9.56,\"end\":10.12},{\"word\":\"god\",\"start\":24.85,\"end\":25.23}]"} {"utterance_id": "7641-96670-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0009.flac", "task_type": "lexical_detection", "query": "thereafter", "prompt": "After the audio ends, extract all occurrences of the word \"thereafter\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.28, "thereafter"]], "response_mode": "offline", "answer": "[{\"word\":\"thereafter\",\"start\":1.77,\"end\":2.28}]"} {"utterance_id": "8254-115543-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0036.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Do not answer until the audio ends. Then list each instance of an animal name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.94, 8.54, "spirits"]], "response_mode": "offline", "answer": "[{\"word\":\"spirits\",\"start\":7.94,\"end\":8.54}]"} {"utterance_id": "3660-6517-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0033.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.48, 1.86, "waist"]], "response_mode": "offline", "answer": "[{\"word\":\"waist\",\"start\":1.48,\"end\":1.86}]"} {"utterance_id": "4153-186223-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.94, 8.21, "arm"], [9.53, 9.78, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":7.94,\"end\":8.21},{\"word\":\"hand\",\"start\":9.53,\"end\":9.78}]"} {"utterance_id": "6467-56885-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0017.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.41, 5.7, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":5.41,\"end\":5.7}]"} {"utterance_id": "5849-50963-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0011.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the full audio. Count every occurrence of a person's name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.17, "one"], [4.35, 4.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.6,\"end\":1.17},{\"word\":\"two\",\"start\":4.35,\"end\":4.7}]"} {"utterance_id": "4323-18416-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 3.84, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":3.52,\"end\":3.84}]"} {"utterance_id": "1650-173551-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0005.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "When the utterance is complete, report how a body part accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.45, 5.75, "one"], [8.77, 9.03, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.45,\"end\":5.75},{\"word\":\"two\",\"start\":8.77,\"end\":9.03}]"} {"utterance_id": "4323-55228-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.91, "aunt hannah"], [5.53, 5.94, "pete"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt hannah\",\"start\":0.52,\"end\":0.91},{\"word\":\"pete\",\"start\":5.53,\"end\":5.94}]"} {"utterance_id": "4323-55228-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.23, 0.7, "dong ling"], [5.07, 5.47, "bertram's"]], "response_mode": "offline", "answer": "[{\"word\":\"dong ling\",\"start\":0.23,\"end\":0.7},{\"word\":\"bertram's\",\"start\":5.07,\"end\":5.47}]"} {"utterance_id": "4831-18525-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0008.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.99, 3.6, "grandpapa"]], "response_mode": "offline", "answer": "[{\"word\":\"grandpapa\",\"start\":2.99,\"end\":3.6}]"} {"utterance_id": "8288-274162-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0001.flac", "task_type": "lexical_detection", "query": "friend's", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"friend's\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.65, 4.08, "friend's"]], "response_mode": "offline", "answer": "[{\"word\":\"friend's\",\"start\":3.65,\"end\":4.08}]"} {"utterance_id": "6267-65525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0030.flac", "task_type": "lexical_detection", "query": "soundly", "prompt": "When the utterance is complete, report all instances of the word \"soundly\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.56, 9.24, "soundly"]], "response_mode": "offline", "answer": "[{\"word\":\"soundly\",\"start\":8.56,\"end\":9.24}]"} {"utterance_id": "8254-84205-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0057.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Process the full audio offline. Find every instance of a time of day and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.33, 4.66, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":4.33,\"end\":4.66}]"} {"utterance_id": "1650-173551-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0001.flac", "task_type": "lexical_detection", "query": "passage", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"passage\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 2.18, "passage"]], "response_mode": "offline", "answer": "[{\"word\":\"passage\",\"start\":1.52,\"end\":2.18}]"} {"utterance_id": "1255-90413-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0021.flac", "task_type": "lexical_detection", "query": "pretty", "prompt": "After listening to the whole recording, output every instance of the word \"pretty\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.97, 2.18, "pretty"]], "response_mode": "offline", "answer": "[{\"word\":\"pretty\",\"start\":1.97,\"end\":2.18}]"} {"utterance_id": "4570-56594-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0005.flac", "task_type": "lexical_detection", "query": "veil", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"veil\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.25, 3.71, "veil"]], "response_mode": "offline", "answer": "[{\"word\":\"veil\",\"start\":3.25,\"end\":3.71}]"} {"utterance_id": "5849-50962-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0000.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.26, 2.54, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":2.26,\"end\":2.54}]"} {"utterance_id": "6455-67803-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0023.flac", "task_type": "lexical_detection", "query": "shaking", "prompt": "After listening to the whole recording, output every instance of the word \"shaking\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.58, "shaking"]], "response_mode": "offline", "answer": "[{\"word\":\"shaking\",\"start\":1.21,\"end\":1.58}]"} {"utterance_id": "7641-96684-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0013.flac", "task_type": "lexical_detection", "query": "courtyard", "prompt": "When the utterance is complete, report all instances of the word \"courtyard\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.69, 7.28, "courtyard"]], "response_mode": "offline", "answer": "[{\"word\":\"courtyard\",\"start\":6.69,\"end\":7.28}]"} {"utterance_id": "116-288047-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0008.flac", "task_type": "lexical_detection", "query": "thinks", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"thinks\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.89, "thinks"]], "response_mode": "offline", "answer": "[{\"word\":\"thinks\",\"start\":0.5,\"end\":0.89}]"} {"utterance_id": "700-122867-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0038.flac", "task_type": "lexical_detection", "query": "tie", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"tie\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.98, 3.27, "tie"]], "response_mode": "offline", "answer": "[{\"word\":\"tie\",\"start\":2.98,\"end\":3.27}]"} {"utterance_id": "8173-294714-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0023.flac", "task_type": "lexical_detection", "query": "men", "prompt": "Process the full audio offline. Find every instance of the word \"men\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.42, "men"]], "response_mode": "offline", "answer": "[{\"word\":\"men\",\"start\":1.12,\"end\":1.42}]"} {"utterance_id": "4323-18416-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.19, 2.64, "charlotte"]], "response_mode": "offline", "answer": "[{\"word\":\"charlotte\",\"start\":2.19,\"end\":2.64}]"} {"utterance_id": "8173-294714-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0036.flac", "task_type": "lexical_counting", "query": "speaking", "prompt": "Listen to the entire clip, then report the count sequence for the word \"speaking\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.13, "one"], [2.17, 2.59, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.69,\"end\":1.13},{\"word\":\"two\",\"start\":2.17,\"end\":2.59}]"} {"utterance_id": "5849-50963-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0008.flac", "task_type": "lexical_detection", "query": "strangers", "prompt": "When the utterance is complete, report all instances of the word \"strangers\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.4, 3.92, "strangers"]], "response_mode": "offline", "answer": "[{\"word\":\"strangers\",\"start\":3.4,\"end\":3.92}]"} {"utterance_id": "5849-50873-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0038.flac", "task_type": "lexical_detection", "query": "fruit", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fruit\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.27, 2.63, "fruit"]], "response_mode": "offline", "answer": "[{\"word\":\"fruit\",\"start\":2.27,\"end\":2.63}]"} {"utterance_id": "4153-186222-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0010.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.35, 1.93, "cousin"]], "response_mode": "offline", "answer": "[{\"word\":\"cousin\",\"start\":1.35,\"end\":1.93}]"} {"utterance_id": "700-122866-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0002.flac", "task_type": "lexical_detection", "query": "marilla", "prompt": "After listening to the whole recording, output every instance of the word \"marilla\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 1.92, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":1.54,\"end\":1.92}]"} {"utterance_id": "4515-11057-0102", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0102.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Once the audio has finished, identify every instance of an organization name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.39, 2.02, "museum"]], "response_mode": "offline", "answer": "[{\"word\":\"museum\",\"start\":1.39,\"end\":2.02}]"} {"utterance_id": "6467-97061-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0008.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 2.02, "father"], [7.82, 8.23, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.51,\"end\":2.02},{\"word\":\"daughter\",\"start\":7.82,\"end\":8.23}]"} {"utterance_id": "6123-59150-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0014.flac", "task_type": "lexical_detection", "query": "nightmare", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"nightmare\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.28, 4.69, "nightmare"]], "response_mode": "offline", "answer": "[{\"word\":\"nightmare\",\"start\":4.28,\"end\":4.69}]"} {"utterance_id": "6467-94831-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0001.flac", "task_type": "lexical_detection", "query": "others", "prompt": "After the audio ends, extract all occurrences of the word \"others\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.55, 4.93, "others"]], "response_mode": "offline", "answer": "[{\"word\":\"others\",\"start\":4.55,\"end\":4.93}]"} {"utterance_id": "7601-175351-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0018.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.42, 2.8, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":2.42,\"end\":2.8}]"} {"utterance_id": "1585-131718-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0035.flac", "task_type": "lexical_detection", "query": "sixty", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"sixty\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.99, 2.27, "sixty"]], "response_mode": "offline", "answer": "[{\"word\":\"sixty\",\"start\":1.99,\"end\":2.27}]"} {"utterance_id": "1686-142278-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0032.flac", "task_type": "lexical_detection", "query": "kind", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"kind\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.53, "kind"]], "response_mode": "offline", "answer": "[{\"word\":\"kind\",\"start\":2.13,\"end\":2.53}]"} {"utterance_id": "4323-18416-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0030.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 0.99, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":0.73,\"end\":0.99}]"} {"utterance_id": "1585-131718-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0044.flac", "task_type": "lexical_detection", "query": "violin", "prompt": "After the audio ends, extract all occurrences of the word \"violin\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.72, 3.3, "violin"]], "response_mode": "offline", "answer": "[{\"word\":\"violin\",\"start\":2.72,\"end\":3.3}]"} {"utterance_id": "4831-25894-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0009.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.22, 14.5, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":14.22,\"end\":14.5}]"} {"utterance_id": "3663-172528-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0054.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Do not answer until the audio ends. Then list each instance of a city name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.3, 7.71, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":7.3,\"end\":7.71}]"} {"utterance_id": "2506-13150-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.04, 3.75, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.04,\"end\":3.75}]"} {"utterance_id": "6455-66379-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0010.flac", "task_type": "lexical_detection", "query": "captain", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"captain\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.18, 1.58, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.18,\"end\":1.58}]"} {"utterance_id": "4153-186222-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0027.flac", "task_type": "lexical_counting", "query": "girls", "prompt": "Once the audio has finished, list each occurrence of the word \"girls\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.64, 2.09, "one"], [5.26, 5.59, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.64,\"end\":2.09},{\"word\":\"two\",\"start\":5.26,\"end\":5.59}]"} {"utterance_id": "5849-50962-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0005.flac", "task_type": "lexical_detection", "query": "languages", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"languages\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.56, 2.22, "languages"]], "response_mode": "offline", "answer": "[{\"word\":\"languages\",\"start\":1.56,\"end\":2.22}]"} {"utterance_id": "8288-274162-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0047.flac", "task_type": "lexical_detection", "query": "carpenter", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"carpenter\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.04, 4.6, "carpenter"]], "response_mode": "offline", "answer": "[{\"word\":\"carpenter\",\"start\":4.04,\"end\":4.6}]"} {"utterance_id": "6841-88291-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0031.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.17, 4.42, "ten"]], "response_mode": "offline", "answer": "[{\"word\":\"ten\",\"start\":4.17,\"end\":4.42}]"} {"utterance_id": "7601-175351-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0025.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a number with word/start/end. If none, reply [].", "labels": [[4.82, 5.14, "one"], [7.88, 8.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.82,\"end\":5.14},{\"word\":\"two\",\"start\":7.88,\"end\":8.44}]"} {"utterance_id": "1701-141760-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0019.flac", "task_type": "lexical_detection", "query": "added", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"added\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.68, 3.12, "added"]], "response_mode": "offline", "answer": "[{\"word\":\"added\",\"start\":2.68,\"end\":3.12}]"} {"utterance_id": "1650-157641-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0012.flac", "task_type": "lexical_detection", "query": "look", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"look\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.37, 1.68, "look"]], "response_mode": "offline", "answer": "[{\"word\":\"look\",\"start\":1.37,\"end\":1.68}]"} {"utterance_id": "4831-25894-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0000.flac", "task_type": "lexical_detection", "query": "small", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"small\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.6, 2.01, "small"]], "response_mode": "offline", "answer": "[{\"word\":\"small\",\"start\":1.6,\"end\":2.01}]"} {"utterance_id": "1255-90407-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0003.flac", "task_type": "lexical_detection", "query": "knowledge", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"knowledge\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.51, 3.81, "knowledge"]], "response_mode": "offline", "answer": "[{\"word\":\"knowledge\",\"start\":3.51,\"end\":3.81}]"} {"utterance_id": "3660-172182-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0040.flac", "task_type": "lexical_detection", "query": "followers", "prompt": "After the audio ends, extract all occurrences of the word \"followers\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.52, "followers"]], "response_mode": "offline", "answer": "[{\"word\":\"followers\",\"start\":1.06,\"end\":1.52}]"} {"utterance_id": "6599-38591-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.65, 9.37, "queen eleanor"]], "response_mode": "offline", "answer": "[{\"word\":\"queen eleanor\",\"start\":8.65,\"end\":9.37}]"} {"utterance_id": "700-122868-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.42, 0.9, "jane"]], "response_mode": "offline", "answer": "[{\"word\":\"jane\",\"start\":0.42,\"end\":0.9}]"} {"utterance_id": "4323-18416-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0019.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.14, "grandpapa"], [3.43, 3.76, "aunt"]], "response_mode": "offline", "answer": "[{\"word\":\"grandpapa\",\"start\":0.5,\"end\":1.14},{\"word\":\"aunt\",\"start\":3.43,\"end\":3.76}]"} {"utterance_id": "3660-172182-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.92, "guenever"]], "response_mode": "offline", "answer": "[{\"word\":\"guenever\",\"start\":0.5,\"end\":0.92}]"} {"utterance_id": "5849-50873-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0001.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 2.96, "one"], [6.76, 7.61, "two"], [9.01, 9.95, "three"], [21.62, 22.37, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.08,\"end\":2.96},{\"word\":\"two\",\"start\":6.76,\"end\":7.61},{\"word\":\"three\",\"start\":9.01,\"end\":9.95},{\"word\":\"four\",\"start\":21.62,\"end\":22.37}]"} {"utterance_id": "6455-67803-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0014.flac", "task_type": "lexical_detection", "query": "lucia", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"lucia\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.5, 0.84, "lucia"]], "response_mode": "offline", "answer": "[{\"word\":\"lucia\",\"start\":0.5,\"end\":0.84}]"} {"utterance_id": "1701-141760-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0040.flac", "task_type": "lexical_counting", "query": "young", "prompt": "After listening to the whole recording, output timed count words for the word \"young\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.98, 2.18, "one"], [11.01, 11.22, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.98,\"end\":2.18},{\"word\":\"two\",\"start\":11.01,\"end\":11.22}]"} {"utterance_id": "3660-172182-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 1.88, "arthur"], [3.13, 3.6, "guenever"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur\",\"start\":1.47,\"end\":1.88},{\"word\":\"guenever\",\"start\":3.13,\"end\":3.6}]"} {"utterance_id": "1686-142278-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0016.flac", "task_type": "lexical_detection", "query": "minister", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"minister\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.75, 2.3, "minister"]], "response_mode": "offline", "answer": "[{\"word\":\"minister\",\"start\":1.75,\"end\":2.3}]"} {"utterance_id": "116-288047-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.91, "god's"]], "response_mode": "offline", "answer": "[{\"word\":\"god's\",\"start\":0.62,\"end\":0.91}]"} {"utterance_id": "4323-18416-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0005.flac", "task_type": "lexical_detection", "query": "fine", "prompt": "When the utterance is complete, report all instances of the word \"fine\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.49, 2.76, "fine"]], "response_mode": "offline", "answer": "[{\"word\":\"fine\",\"start\":2.49,\"end\":2.76}]"} {"utterance_id": "4323-55228-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.79, 5.6, "miss winthrop"]], "response_mode": "offline", "answer": "[{\"word\":\"miss winthrop\",\"start\":4.79,\"end\":5.6}]"} {"utterance_id": "4323-55228-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0049.flac", "task_type": "lexical_detection", "query": "glanced", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"glanced\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.08, "glanced"]], "response_mode": "offline", "answer": "[{\"word\":\"glanced\",\"start\":0.73,\"end\":1.08}]"} {"utterance_id": "1701-141759-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0001.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.82, 5.14, "son's"]], "response_mode": "offline", "answer": "[{\"word\":\"son's\",\"start\":4.82,\"end\":5.14}]"} {"utterance_id": "4323-13259-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0012.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.66, 1.98, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.66,\"end\":1.98}]"} {"utterance_id": "4323-55228-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0038.flac", "task_type": "lexical_detection", "query": "sir", "prompt": "After listening to the whole recording, output every instance of the word \"sir\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 1.83, "sir"]], "response_mode": "offline", "answer": "[{\"word\":\"sir\",\"start\":1.58,\"end\":1.83}]"} {"utterance_id": "1630-96099-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0008.flac", "task_type": "lexical_detection", "query": "rate", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"rate\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.12, "rate"]], "response_mode": "offline", "answer": "[{\"word\":\"rate\",\"start\":0.83,\"end\":1.12}]"} {"utterance_id": "6841-88294-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0012.flac", "task_type": "lexical_detection", "query": "song", "prompt": "Once the audio has finished, identify every instance of the word \"song\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.57, 1.91, "song"]], "response_mode": "offline", "answer": "[{\"word\":\"song\",\"start\":1.57,\"end\":1.91}]"} {"utterance_id": "1686-142278-0073", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0073.flac", "task_type": "lexical_counting", "query": "mister", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"mister\" with word/start/end. If none, reply [].", "labels": [[0.5, 0.77, "one"], [8.26, 8.55, "two"], [10.47, 10.78, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":0.77},{\"word\":\"two\",\"start\":8.26,\"end\":8.55},{\"word\":\"three\",\"start\":10.47,\"end\":10.78}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.82, 10.51, "officers"]], "response_mode": "offline", "answer": "[{\"word\":\"officers\",\"start\":9.82,\"end\":10.51}]"} {"utterance_id": "6599-38591-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0005.flac", "task_type": "lexical_detection", "query": "signed", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"signed\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.1, 1.38, "signed"]], "response_mode": "offline", "answer": "[{\"word\":\"signed\",\"start\":1.1,\"end\":1.38}]"} {"utterance_id": "4515-11057-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0026.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.74, 4.07, "mind"]], "response_mode": "offline", "answer": "[{\"word\":\"mind\",\"start\":3.74,\"end\":4.07}]"} {"utterance_id": "1255-138279-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0024.flac", "task_type": "lexical_detection", "query": "night", "prompt": "After the audio ends, extract all occurrences of the word \"night\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.86, "night"], [2.07, 2.57, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":0.66,\"end\":0.86},{\"word\":\"night\",\"start\":2.07,\"end\":2.57}]"} {"utterance_id": "4515-11057-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0054.flac", "task_type": "lexical_detection", "query": "chance", "prompt": "Process the full audio offline. Find every instance of the word \"chance\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.21, 2.81, "chance"]], "response_mode": "offline", "answer": "[{\"word\":\"chance\",\"start\":2.21,\"end\":2.81}]"} {"utterance_id": "1585-131718-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0031.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Do not answer until the audio ends. Then list each instance of an organization name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.79, 4.44, "edison electric light company"]], "response_mode": "offline", "answer": "[{\"word\":\"edison electric light company\",\"start\":2.79,\"end\":4.44}]"} {"utterance_id": "2506-169427-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.86, 10.59, "eugenius s"]], "response_mode": "offline", "answer": "[{\"word\":\"eugenius s\",\"start\":9.86,\"end\":10.59}]"} {"utterance_id": "6467-94831-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0041.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.13, 1.44, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.13,\"end\":1.44}]"} {"utterance_id": "700-122868-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[6.77, 7.28, "mister barry"]], "response_mode": "offline", "answer": "[{\"word\":\"mister barry\",\"start\":6.77,\"end\":7.28}]"} {"utterance_id": "4323-18416-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0016.flac", "task_type": "lexical_detection", "query": "couple", "prompt": "Process the full audio offline. Find every instance of the word \"couple\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.41, "couple"]], "response_mode": "offline", "answer": "[{\"word\":\"couple\",\"start\":2.16,\"end\":2.41}]"} {"utterance_id": "4515-11057-0112", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0112.flac", "task_type": "lexical_detection", "query": "sex", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"sex\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 3.11, "sex"]], "response_mode": "offline", "answer": "[{\"word\":\"sex\",\"start\":2.58,\"end\":3.11}]"} {"utterance_id": "1255-90407-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0015.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.71, 4.95, "son"], [6.22, 6.6, "daughter"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":4.71,\"end\":4.95},{\"word\":\"daughter\",\"start\":6.22,\"end\":6.6}]"} {"utterance_id": "2506-13150-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 1.23, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":0.53,\"end\":1.23}]"} {"utterance_id": "4515-11057-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0041.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.58, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":3.21,\"end\":3.58}]"} {"utterance_id": "4515-11057-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0065.flac", "task_type": "lexical_detection", "query": "support", "prompt": "After the audio ends, extract all occurrences of the word \"support\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.19, 6.69, "support"]], "response_mode": "offline", "answer": "[{\"word\":\"support\",\"start\":6.19,\"end\":6.69}]"} {"utterance_id": "4831-25894-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0028.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.86, 8.17, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":7.86,\"end\":8.17}]"} {"utterance_id": "6267-53049-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.22, 1.1, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":0.22,\"end\":1.1}]"} {"utterance_id": "4831-18525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a city name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.29, 4.87, "dresden"]], "response_mode": "offline", "answer": "[{\"word\":\"dresden\",\"start\":4.29,\"end\":4.87}]"} {"utterance_id": "6467-94831-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0021.flac", "task_type": "lexical_detection", "query": "begun", "prompt": "When the utterance is complete, report all instances of the word \"begun\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.13, 2.54, "begun"]], "response_mode": "offline", "answer": "[{\"word\":\"begun\",\"start\":2.13,\"end\":2.54}]"} {"utterance_id": "6841-88294-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0053.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.67, 1.97, "dozen"]], "response_mode": "offline", "answer": "[{\"word\":\"dozen\",\"start\":1.67,\"end\":1.97}]"} {"utterance_id": "8254-84205-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0045.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.48, 4.97, "wilton"]], "response_mode": "offline", "answer": "[{\"word\":\"wilton\",\"start\":4.48,\"end\":4.97}]"} {"utterance_id": "7697-105815-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.1, 3.46, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":3.1,\"end\":3.46}]"} {"utterance_id": "116-288047-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0002.flac", "task_type": "lexical_detection", "query": "snap", "prompt": "When the utterance is complete, report all instances of the word \"snap\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.15, "snap"]], "response_mode": "offline", "answer": "[{\"word\":\"snap\",\"start\":1.79,\"end\":2.15}]"} {"utterance_id": "4831-25894-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0003.flac", "task_type": "lexical_detection", "query": "merry", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"merry\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.57, 3.87, "merry"]], "response_mode": "offline", "answer": "[{\"word\":\"merry\",\"start\":3.57,\"end\":3.87}]"} {"utterance_id": "1585-131718-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0037.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.98, 3.38, "cigar"]], "response_mode": "offline", "answer": "[{\"word\":\"cigar\",\"start\":2.98,\"end\":3.38}]"} {"utterance_id": "7601-101619-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0003.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a profession: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[3.67, 4.1, "one"], [17.41, 17.99, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.67,\"end\":4.1},{\"word\":\"two\",\"start\":17.41,\"end\":17.99}]"} {"utterance_id": "1630-96099-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0013.flac", "task_type": "lexical_detection", "query": "monsieur", "prompt": "After listening to the whole recording, output every instance of the word \"monsieur\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.79, "monsieur"]], "response_mode": "offline", "answer": "[{\"word\":\"monsieur\",\"start\":1.25,\"end\":1.79}]"} {"utterance_id": "6467-94831-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0038.flac", "task_type": "lexical_detection", "query": "demanded", "prompt": "After the audio ends, extract all occurrences of the word \"demanded\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.14, 0.59, "demanded"]], "response_mode": "offline", "answer": "[{\"word\":\"demanded\",\"start\":0.14,\"end\":0.59}]"} {"utterance_id": "4153-186223-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0016.flac", "task_type": "lexical_detection", "query": "happen", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"happen\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.06, 6.36, "happen"]], "response_mode": "offline", "answer": "[{\"word\":\"happen\",\"start\":6.06,\"end\":6.36}]"} {"utterance_id": "1650-167613-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0014.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.0, 5.64, "minister"]], "response_mode": "offline", "answer": "[{\"word\":\"minister\",\"start\":5.0,\"end\":5.64}]"} {"utterance_id": "700-122866-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0035.flac", "task_type": "lexical_detection", "query": "allan", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"allan\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.05, 2.29, "allan"]], "response_mode": "offline", "answer": "[{\"word\":\"allan\",\"start\":2.05,\"end\":2.29}]"} {"utterance_id": "1630-96099-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0014.flac", "task_type": "lexical_detection", "query": "sight", "prompt": "After listening to the whole recording, output every instance of the word \"sight\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.96, "sight"]], "response_mode": "offline", "answer": "[{\"word\":\"sight\",\"start\":1.38,\"end\":1.96}]"} {"utterance_id": "700-122866-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0007.flac", "task_type": "lexical_detection", "query": "weeks", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"weeks\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.98, 1.3, "weeks"]], "response_mode": "offline", "answer": "[{\"word\":\"weeks\",\"start\":0.98,\"end\":1.3}]"} {"utterance_id": "7641-96670-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0006.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "Listen carefully to the entire clip, then report every instance of a weather term as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.38, 0.71, "mist"]], "response_mode": "offline", "answer": "[{\"word\":\"mist\",\"start\":0.38,\"end\":0.71}]"} {"utterance_id": "4323-55228-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0044.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.93, 4.2, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":3.93,\"end\":4.2}]"} {"utterance_id": "1701-141760-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0007.flac", "task_type": "lexical_detection", "query": "opened", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"opened\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.59, 2.15, "opened"]], "response_mode": "offline", "answer": "[{\"word\":\"opened\",\"start\":1.59,\"end\":2.15}]"} {"utterance_id": "8288-274150-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0006.flac", "task_type": "lexical_detection", "query": "ceased", "prompt": "After listening to the whole recording, output every instance of the word \"ceased\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.54, 2.97, "ceased"]], "response_mode": "offline", "answer": "[{\"word\":\"ceased\",\"start\":2.54,\"end\":2.97}]"} {"utterance_id": "7641-96252-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0009.flac", "task_type": "lexical_counting", "query": "conscience", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"conscience\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.81, 2.56, "one"], [7.04, 7.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.81,\"end\":2.56},{\"word\":\"two\",\"start\":7.04,\"end\":7.66}]"} {"utterance_id": "7641-96684-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0036.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.66, 3.21, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":2.66,\"end\":3.21}]"} {"utterance_id": "8254-84205-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.17, 3.68, "chris lee's"]], "response_mode": "offline", "answer": "[{\"word\":\"chris lee's\",\"start\":3.17,\"end\":3.68}]"} {"utterance_id": "6599-38591-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0002.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Once the audio has finished, identify every instance of a date. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.61, 3.67, "fifteen twenty nine"], [4.69, 5.99, "fifteen thirty six"]], "response_mode": "offline", "answer": "[{\"word\":\"fifteen twenty nine\",\"start\":2.61,\"end\":3.67},{\"word\":\"fifteen thirty six\",\"start\":4.69,\"end\":5.99}]"} {"utterance_id": "1255-90413-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.55, 2.46, "missus pine avon's"]], "response_mode": "offline", "answer": "[{\"word\":\"missus pine avon's\",\"start\":1.55,\"end\":2.46}]"} {"utterance_id": "8288-274162-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0008.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Do not answer until the audio ends. Then list each instance of a country name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.78, 4.22, "england"]], "response_mode": "offline", "answer": "[{\"word\":\"england\",\"start\":3.78,\"end\":4.22}]"} {"utterance_id": "5543-27761-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0075.flac", "task_type": "lexical_detection", "query": "prayed", "prompt": "After the audio ends, extract all occurrences of the word \"prayed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.73, 1.07, "prayed"]], "response_mode": "offline", "answer": "[{\"word\":\"prayed\",\"start\":0.73,\"end\":1.07}]"} {"utterance_id": "6123-59186-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0022.flac", "task_type": "lexical_detection", "query": "kneeling", "prompt": "Once the audio has finished, identify every instance of the word \"kneeling\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[7.86, 8.34, "kneeling"]], "response_mode": "offline", "answer": "[{\"word\":\"kneeling\",\"start\":7.86,\"end\":8.34}]"} {"utterance_id": "8288-274162-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0034.flac", "task_type": "lexical_detection", "query": "quarrel", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"quarrel\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.23, 2.6, "quarrel"]], "response_mode": "offline", "answer": "[{\"word\":\"quarrel\",\"start\":2.23,\"end\":2.6}]"} {"utterance_id": "8173-294714-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0044.flac", "task_type": "lexical_detection", "query": "single", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"single\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.52, 2.89, "single"]], "response_mode": "offline", "answer": "[{\"word\":\"single\",\"start\":2.52,\"end\":2.89}]"} {"utterance_id": "4153-186222-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.59, 2.9, "beth"]], "response_mode": "offline", "answer": "[{\"word\":\"beth\",\"start\":2.59,\"end\":2.9}]"} {"utterance_id": "6841-88291-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0036.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.35, 1.57, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.35,\"end\":1.57}]"} {"utterance_id": "116-288045-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0032.flac", "task_type": "lexical_detection", "query": "proves", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"proves\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.8, 3.11, "proves"]], "response_mode": "offline", "answer": "[{\"word\":\"proves\",\"start\":2.8,\"end\":3.11}]"} {"utterance_id": "6267-65525-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0007.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.76, "cousins"]], "response_mode": "offline", "answer": "[{\"word\":\"cousins\",\"start\":5.09,\"end\":5.76}]"} {"utterance_id": "3660-172183-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0019.flac", "task_type": "lexical_detection", "query": "meadows", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"meadows\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.0, "meadows"], [4.86, 5.4, "meadows"]], "response_mode": "offline", "answer": "[{\"word\":\"meadows\",\"start\":2.61,\"end\":3.0},{\"word\":\"meadows\",\"start\":4.86,\"end\":5.4}]"} {"utterance_id": "3915-98647-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0032.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.33, 7.6, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":7.33,\"end\":7.6}]"} {"utterance_id": "4831-25894-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 2.85, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":2.48,\"end\":2.85}]"} {"utterance_id": "5849-50962-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0013.flac", "task_type": "lexical_detection", "query": "nellie", "prompt": "Once the audio has finished, identify every instance of the word \"nellie\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.07, 3.37, "nellie"]], "response_mode": "offline", "answer": "[{\"word\":\"nellie\",\"start\":3.07,\"end\":3.37}]"} {"utterance_id": "8173-294714-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0010.flac", "task_type": "lexical_detection", "query": "van", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"van\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.71, 1.96, "van"]], "response_mode": "offline", "answer": "[{\"word\":\"van\",\"start\":1.71,\"end\":1.96}]"} {"utterance_id": "6267-65525-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0004.flac", "task_type": "lexical_detection", "query": "downstairs", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"downstairs\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.45, "downstairs"]], "response_mode": "offline", "answer": "[{\"word\":\"downstairs\",\"start\":1.88,\"end\":2.45}]"} {"utterance_id": "6841-88294-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0037.flac", "task_type": "lexical_detection", "query": "pulling", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pulling\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.15, 1.45, "pulling"]], "response_mode": "offline", "answer": "[{\"word\":\"pulling\",\"start\":1.15,\"end\":1.45}]"} {"utterance_id": "6455-67804-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0014.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Once the audio has finished, identify every instance of a time of day. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.57, 2.96, "evening"]], "response_mode": "offline", "answer": "[{\"word\":\"evening\",\"start\":2.57,\"end\":2.96}]"} {"utterance_id": "7641-96684-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0035.flac", "task_type": "lexical_detection", "query": "veiled", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"veiled\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.14, "veiled"]], "response_mode": "offline", "answer": "[{\"word\":\"veiled\",\"start\":0.65,\"end\":1.14}]"} {"utterance_id": "6267-65525-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.8, "sidney"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney\",\"start\":0.42,\"end\":0.8}]"} {"utterance_id": "6455-67804-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0020.flac", "task_type": "lexical_detection", "query": "she's", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"she's\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.2, 0.35, "she's"]], "response_mode": "offline", "answer": "[{\"word\":\"she's\",\"start\":0.2,\"end\":0.35}]"} {"utterance_id": "6267-53049-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.13, 1.24, "doris hunter"]], "response_mode": "offline", "answer": "[{\"word\":\"doris hunter\",\"start\":0.13,\"end\":1.24}]"} {"utterance_id": "7697-105817-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0006.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Do not answer until the audio ends. Then list each instance of an ordinal number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.07, 3.78, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":3.07,\"end\":3.78}]"} {"utterance_id": "1650-157641-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0013.flac", "task_type": "lexical_detection", "query": "drawing", "prompt": "After listening to the whole recording, output every instance of the word \"drawing\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 3.97, "drawing"]], "response_mode": "offline", "answer": "[{\"word\":\"drawing\",\"start\":3.36,\"end\":3.97}]"} {"utterance_id": "7641-96252-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0010.flac", "task_type": "lexical_detection", "query": "individuals", "prompt": "Process the full audio offline. Find every instance of the word \"individuals\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.86, "individuals"]], "response_mode": "offline", "answer": "[{\"word\":\"individuals\",\"start\":2.09,\"end\":2.86}]"} {"utterance_id": "6267-65525-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0057.flac", "task_type": "lexical_detection", "query": "letters", "prompt": "After the audio ends, extract all occurrences of the word \"letters\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 2.16, "letters"]], "response_mode": "offline", "answer": "[{\"word\":\"letters\",\"start\":1.59,\"end\":2.16}]"} {"utterance_id": "6599-38591-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0013.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After listening to the whole recording, output every instance of a country name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.23, 3.54, "france"]], "response_mode": "offline", "answer": "[{\"word\":\"france\",\"start\":3.23,\"end\":3.54}]"} {"utterance_id": "1701-141760-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0016.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.44, 0.82, "dinners"]], "response_mode": "offline", "answer": "[{\"word\":\"dinners\",\"start\":0.44,\"end\":0.82}]"} {"utterance_id": "116-288048-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0001.flac", "task_type": "lexical_detection", "query": "evidence", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"evidence\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.95, 1.36, "evidence"]], "response_mode": "offline", "answer": "[{\"word\":\"evidence\",\"start\":0.95,\"end\":1.36}]"} {"utterance_id": "1630-73710-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0002.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.51, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":1.11,\"end\":1.51}]"} {"utterance_id": "116-288047-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0006.flac", "task_type": "lexical_detection", "query": "myth", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"myth\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.13, 1.36, "myth"]], "response_mode": "offline", "answer": "[{\"word\":\"myth\",\"start\":1.13,\"end\":1.36}]"} {"utterance_id": "1701-141759-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0010.flac", "task_type": "lexical_detection", "query": "natasha", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"natasha\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.07, "natasha"]], "response_mode": "offline", "answer": "[{\"word\":\"natasha\",\"start\":0.52,\"end\":1.07}]"} {"utterance_id": "1686-142278-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0067.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.52, 0.97, "one"], [2.31, 2.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.52,\"end\":0.97},{\"word\":\"two\",\"start\":2.31,\"end\":2.72}]"} {"utterance_id": "4515-11057-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0036.flac", "task_type": "lexical_detection", "query": "surmise", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"surmise\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.69, 1.3, "surmise"]], "response_mode": "offline", "answer": "[{\"word\":\"surmise\",\"start\":0.69,\"end\":1.3}]"} {"utterance_id": "1701-141760-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0052.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.87, 4.45, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":3.87,\"end\":4.45}]"} {"utterance_id": "6267-53049-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.38, 6.19, "penelope"]], "response_mode": "offline", "answer": "[{\"word\":\"penelope\",\"start\":5.38,\"end\":6.19}]"} {"utterance_id": "4831-29134-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0013.flac", "task_type": "lexical_detection", "query": "governor", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"governor\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.35, "governor"]], "response_mode": "offline", "answer": "[{\"word\":\"governor\",\"start\":1.84,\"end\":2.35}]"} {"utterance_id": "5849-50962-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0013.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.22, 4.65, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":4.22,\"end\":4.65}]"} {"utterance_id": "3660-6517-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0004.flac", "task_type": "lexical_detection", "query": "tent", "prompt": "Once the audio has finished, identify every instance of the word \"tent\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.02, 1.24, "tent"]], "response_mode": "offline", "answer": "[{\"word\":\"tent\",\"start\":1.02,\"end\":1.24}]"} {"utterance_id": "1701-141759-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0027.flac", "task_type": "lexical_counting", "query": "voices", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"voices\" with word/start/end. If none, reply [].", "labels": [[2.61, 3.33, "one"], [11.9, 12.33, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.61,\"end\":3.33},{\"word\":\"two\",\"start\":11.9,\"end\":12.33}]"} {"utterance_id": "6467-97061-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0002.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.27, 4.75, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":4.27,\"end\":4.75}]"} {"utterance_id": "3660-172182-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0010.flac", "task_type": "lexical_detection", "query": "gate", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"gate\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.0, "gate"]], "response_mode": "offline", "answer": "[{\"word\":\"gate\",\"start\":1.74,\"end\":2.0}]"} {"utterance_id": "3915-57461-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0000.flac", "task_type": "lexical_detection", "query": "bronze", "prompt": "After listening to the whole recording, output every instance of the word \"bronze\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.89, 7.31, "bronze"]], "response_mode": "offline", "answer": "[{\"word\":\"bronze\",\"start\":6.89,\"end\":7.31}]"} {"utterance_id": "6267-65525-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0049.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.07, "four"]], "response_mode": "offline", "answer": "[{\"word\":\"four\",\"start\":0.77,\"end\":1.07}]"} {"utterance_id": "8254-115543-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0010.flac", "task_type": "lexical_detection", "query": "lovely", "prompt": "After the audio ends, extract all occurrences of the word \"lovely\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.04, "lovely"]], "response_mode": "offline", "answer": "[{\"word\":\"lovely\",\"start\":0.67,\"end\":1.04}]"} {"utterance_id": "6123-59186-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.96, 2.47, "singer"]], "response_mode": "offline", "answer": "[{\"word\":\"singer\",\"start\":1.96,\"end\":2.47}]"} {"utterance_id": "5849-50962-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0003.flac", "task_type": "lexical_counting", "query": "board", "prompt": "Listen to the full audio. Count every occurrence of the word \"board\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.17, "one"], [3.08, 3.47, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.86,\"end\":1.17},{\"word\":\"two\",\"start\":3.08,\"end\":3.47}]"} {"utterance_id": "4570-56594-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0017.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.92, 2.36, "salt"]], "response_mode": "offline", "answer": "[{\"word\":\"salt\",\"start\":1.92,\"end\":2.36}]"} {"utterance_id": "5849-50964-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0007.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.32, 0.6, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":0.32,\"end\":0.6}]"} {"utterance_id": "6267-65525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0030.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.63, 12.99, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":12.63,\"end\":12.99}]"} {"utterance_id": "4153-186222-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0015.flac", "task_type": "lexical_detection", "query": "delights", "prompt": "When the utterance is complete, report all instances of the word \"delights\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.85, 3.41, "delights"]], "response_mode": "offline", "answer": "[{\"word\":\"delights\",\"start\":2.85,\"end\":3.41}]"} {"utterance_id": "8254-84205-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.48, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":2.11,\"end\":2.48}]"} {"utterance_id": "8288-274162-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0035.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.79, 6.16, "brains"]], "response_mode": "offline", "answer": "[{\"word\":\"brains\",\"start\":5.79,\"end\":6.16}]"} {"utterance_id": "1686-142278-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0011.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.33, 4.76, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":4.33,\"end\":4.76}]"} {"utterance_id": "4831-29134-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.02, 3.46, "faces"]], "response_mode": "offline", "answer": "[{\"word\":\"faces\",\"start\":3.02,\"end\":3.46}]"} {"utterance_id": "116-288045-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0003.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.63, 3.16, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":2.63,\"end\":3.16}]"} {"utterance_id": "1585-131718-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0038.flac", "task_type": "lexical_detection", "query": "suggested", "prompt": "After the audio ends, extract all occurrences of the word \"suggested\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.09, "suggested"]], "response_mode": "offline", "answer": "[{\"word\":\"suggested\",\"start\":0.54,\"end\":1.09}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.47, "saybrook"]], "response_mode": "offline", "answer": "[{\"word\":\"saybrook\",\"start\":3.82,\"end\":4.47}]"} {"utterance_id": "1585-157660-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0013.flac", "task_type": "lexical_detection", "query": "joking", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"joking\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.28, 3.71, "joking"]], "response_mode": "offline", "answer": "[{\"word\":\"joking\",\"start\":3.28,\"end\":3.71}]"} {"utterance_id": "4515-11057-0064", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0064.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.11, 5.32, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.11,\"end\":5.32}]"} {"utterance_id": "1686-142278-0077", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0077.flac", "task_type": "lexical_detection", "query": "private", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"private\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.13, "private"]], "response_mode": "offline", "answer": "[{\"word\":\"private\",\"start\":0.59,\"end\":1.13}]"} {"utterance_id": "6123-59150-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.08, 4.76, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":4.08,\"end\":4.76}]"} {"utterance_id": "6841-88294-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0040.flac", "task_type": "lexical_detection", "query": "asked", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"asked\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.29, 5.59, "asked"]], "response_mode": "offline", "answer": "[{\"word\":\"asked\",\"start\":5.29,\"end\":5.59}]"} {"utterance_id": "116-288046-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0001.flac", "task_type": "lexical_detection", "query": "multitude", "prompt": "After the audio ends, extract all occurrences of the word \"multitude\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.02, 6.72, "multitude"]], "response_mode": "offline", "answer": "[{\"word\":\"multitude\",\"start\":6.02,\"end\":6.72}]"} {"utterance_id": "1650-167613-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0016.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 4.18, "commander"]], "response_mode": "offline", "answer": "[{\"word\":\"commander\",\"start\":3.54,\"end\":4.18}]"} {"utterance_id": "4831-25894-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.21, 1.53, "ranza"]], "response_mode": "offline", "answer": "[{\"word\":\"ranza\",\"start\":1.21,\"end\":1.53}]"} {"utterance_id": "6267-65525-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0045.flac", "task_type": "lexical_detection", "query": "empty", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"empty\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.97, 7.47, "empty"]], "response_mode": "offline", "answer": "[{\"word\":\"empty\",\"start\":6.97,\"end\":7.47}]"} {"utterance_id": "7697-105815-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0004.flac", "task_type": "lexical_detection", "query": "gold", "prompt": "Process the full audio offline. Find every instance of the word \"gold\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.58, 3.95, "gold"]], "response_mode": "offline", "answer": "[{\"word\":\"gold\",\"start\":3.58,\"end\":3.95}]"} {"utterance_id": "4515-11057-0094", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0094.flac", "task_type": "lexical_detection", "query": "disregarded", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"disregarded\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.54, 2.45, "disregarded"]], "response_mode": "offline", "answer": "[{\"word\":\"disregarded\",\"start\":1.54,\"end\":2.45}]"} {"utterance_id": "4570-56594-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.37, "preacher"]], "response_mode": "offline", "answer": "[{\"word\":\"preacher\",\"start\":0.97,\"end\":1.37}]"} {"utterance_id": "3663-172528-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.29, 0.82, "giorgio"], [6.43, 7.86, "maestro agostino"]], "response_mode": "offline", "answer": "[{\"word\":\"giorgio\",\"start\":0.29,\"end\":0.82},{\"word\":\"maestro agostino\",\"start\":6.43,\"end\":7.86}]"} {"utterance_id": "3660-6517-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0035.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.12, 2.33, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.12,\"end\":2.33}]"} {"utterance_id": "6455-67803-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0016.flac", "task_type": "lexical_detection", "query": "tightly", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"tightly\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.16, 3.62, "tightly"]], "response_mode": "offline", "answer": "[{\"word\":\"tightly\",\"start\":3.16,\"end\":3.62}]"} {"utterance_id": "5849-50963-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.73, 2.21, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.73,\"end\":2.21}]"} {"utterance_id": "4515-11057-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0051.flac", "task_type": "lexical_detection", "query": "don't", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"don't\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.27, "don't"]], "response_mode": "offline", "answer": "[{\"word\":\"don't\",\"start\":1.83,\"end\":2.27}]"} {"utterance_id": "700-122867-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0029.flac", "task_type": "lexical_detection", "query": "yes", "prompt": "After listening to the whole recording, output every instance of the word \"yes\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.86, "yes"]], "response_mode": "offline", "answer": "[{\"word\":\"yes\",\"start\":0.49,\"end\":0.86}]"} {"utterance_id": "6455-67804-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0005.flac", "task_type": "lexical_detection", "query": "gown", "prompt": "Once the audio has finished, identify every instance of the word \"gown\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.92, 6.19, "gown"]], "response_mode": "offline", "answer": "[{\"word\":\"gown\",\"start\":5.92,\"end\":6.19}]"} {"utterance_id": "1701-141760-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0033.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen carefully to the entire clip, then report every instance of an animal name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 2.26, "devils"]], "response_mode": "offline", "answer": "[{\"word\":\"devils\",\"start\":1.48,\"end\":2.26}]"} {"utterance_id": "1701-141760-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.58, "gabriel"]], "response_mode": "offline", "answer": "[{\"word\":\"gabriel\",\"start\":1.18,\"end\":1.58}]"} {"utterance_id": "8254-115543-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0026.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 2.84, "elephant"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant\",\"start\":2.22,\"end\":2.84}]"} {"utterance_id": "1701-141759-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0027.flac", "task_type": "lexical_detection", "query": "voices", "prompt": "After listening to the whole recording, output every instance of the word \"voices\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.33, "voices"], [11.9, 12.33, "voices"]], "response_mode": "offline", "answer": "[{\"word\":\"voices\",\"start\":2.61,\"end\":3.33},{\"word\":\"voices\",\"start\":11.9,\"end\":12.33}]"} {"utterance_id": "3915-57461-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0030.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.65, 7.04, "beak"], [9.99, 10.45, "mind"]], "response_mode": "offline", "answer": "[{\"word\":\"beak\",\"start\":6.65,\"end\":7.04},{\"word\":\"mind\",\"start\":9.99,\"end\":10.45}]"} {"utterance_id": "1650-167613-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0021.flac", "task_type": "lexical_counting", "query": "position", "prompt": "When the utterance is complete, report how the word \"position\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.06, 4.83, "one"], [9.51, 10.22, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.06,\"end\":4.83},{\"word\":\"two\",\"start\":9.51,\"end\":10.22}]"} {"utterance_id": "1585-157660-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.97, 1.37, "jack"]], "response_mode": "offline", "answer": "[{\"word\":\"jack\",\"start\":0.97,\"end\":1.37}]"} {"utterance_id": "6841-88291-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0025.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.3, "swallow"]], "response_mode": "offline", "answer": "[{\"word\":\"swallow\",\"start\":0.93,\"end\":1.3}]"} {"utterance_id": "4572-64670-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0000.flac", "task_type": "lexical_counting", "query": "thy", "prompt": "Listen to the full audio. Count every occurrence of the word \"thy\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.08, "one"], [9.4, 9.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.9,\"end\":1.08},{\"word\":\"two\",\"start\":9.4,\"end\":9.6}]"} {"utterance_id": "7641-96670-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0021.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 5.08, "dragon"]], "response_mode": "offline", "answer": "[{\"word\":\"dragon\",\"start\":4.52,\"end\":5.08}]"} {"utterance_id": "1686-142278-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0007.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.78, 2.11, "mother's"]], "response_mode": "offline", "answer": "[{\"word\":\"mother's\",\"start\":1.78,\"end\":2.11}]"} {"utterance_id": "1630-141772-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0016.flac", "task_type": "lexical_detection", "query": "thin", "prompt": "Process the full audio offline. Find every instance of the word \"thin\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.54, 7.96, "thin"]], "response_mode": "offline", "answer": "[{\"word\":\"thin\",\"start\":7.54,\"end\":7.96}]"} {"utterance_id": "6123-59150-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0045.flac", "task_type": "lexical_counting", "query": "novels", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"novels\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[3.57, 4.09, "one"], [7.23, 7.8, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.57,\"end\":4.09},{\"word\":\"two\",\"start\":7.23,\"end\":7.8}]"} {"utterance_id": "7641-96684-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0015.flac", "task_type": "lexical_detection", "query": "near", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"near\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.7, 6.93, "near"]], "response_mode": "offline", "answer": "[{\"word\":\"near\",\"start\":6.7,\"end\":6.93}]"} {"utterance_id": "3660-172182-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0020.flac", "task_type": "lexical_detection", "query": "arthur's", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"arthur's\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.31, 7.79, "arthur's"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur's\",\"start\":7.31,\"end\":7.79}]"} {"utterance_id": "6455-67803-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0008.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Process the full audio offline. Find every instance of a money or currency term and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.54, 8.92, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":8.54,\"end\":8.92}]"} {"utterance_id": "1650-173551-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0002.flac", "task_type": "lexical_detection", "query": "porter", "prompt": "Once the audio has finished, identify every instance of the word \"porter\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.82, 1.27, "porter"]], "response_mode": "offline", "answer": "[{\"word\":\"porter\",\"start\":0.82,\"end\":1.27}]"} {"utterance_id": "4572-112383-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 2.13, "sita"]], "response_mode": "offline", "answer": "[{\"word\":\"sita\",\"start\":1.45,\"end\":2.13}]"} {"utterance_id": "6467-62797-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.02, 6.51, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":6.02,\"end\":6.51}]"} {"utterance_id": "1686-142278-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0027.flac", "task_type": "lexical_detection", "query": "moment", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"moment\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.58, 2.0, "moment"]], "response_mode": "offline", "answer": "[{\"word\":\"moment\",\"start\":1.58,\"end\":2.0}]"} {"utterance_id": "6599-38591-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0011.flac", "task_type": "lexical_detection", "query": "seven", "prompt": "When the utterance is complete, report all instances of the word \"seven\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.84, "seven"]], "response_mode": "offline", "answer": "[{\"word\":\"seven\",\"start\":0.46,\"end\":0.84}]"} {"utterance_id": "8254-84205-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "When the utterance is complete, report all instances of an animal name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.16, "pony"]], "response_mode": "offline", "answer": "[{\"word\":\"pony\",\"start\":0.83,\"end\":1.16}]"} {"utterance_id": "3660-172182-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0007.flac", "task_type": "lexical_detection", "query": "aspect", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"aspect\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.76, 6.2, "aspect"]], "response_mode": "offline", "answer": "[{\"word\":\"aspect\",\"start\":5.76,\"end\":6.2}]"} {"utterance_id": "7697-245712-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0003.flac", "task_type": "lexical_detection", "query": "providence", "prompt": "After the audio ends, extract all occurrences of the word \"providence\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.73, 3.34, "providence"]], "response_mode": "offline", "answer": "[{\"word\":\"providence\",\"start\":2.73,\"end\":3.34}]"} {"utterance_id": "4831-18525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.41, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":1.99,\"end\":2.41}]"} {"utterance_id": "6267-65525-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0049.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.77, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":0.49,\"end\":0.77}]"} {"utterance_id": "6267-65525-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.55, 2.76, "sidney richmond"]], "response_mode": "offline", "answer": "[{\"word\":\"sidney richmond\",\"start\":1.55,\"end\":2.76}]"} {"utterance_id": "3663-172528-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0043.flac", "task_type": "lexical_detection", "query": "exquisite", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"exquisite\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.35, 3.21, "exquisite"]], "response_mode": "offline", "answer": "[{\"word\":\"exquisite\",\"start\":2.35,\"end\":3.21}]"} {"utterance_id": "4831-18525-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0025.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After the audio ends, extract all occurrences of a food name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.42, 6.9, "gingerbread"]], "response_mode": "offline", "answer": "[{\"word\":\"gingerbread\",\"start\":6.42,\"end\":6.9}]"} {"utterance_id": "7697-105815-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0023.flac", "task_type": "lexical_detection", "query": "climb", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"climb\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.23, 3.64, "climb"]], "response_mode": "offline", "answer": "[{\"word\":\"climb\",\"start\":3.23,\"end\":3.64}]"} {"utterance_id": "5849-50962-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0009.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.42, 3.81, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":3.42,\"end\":3.81}]"} {"utterance_id": "6599-38591-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 1.38, "fifteen sixty two"]], "response_mode": "offline", "answer": "[{\"word\":\"fifteen sixty two\",\"start\":0.35,\"end\":1.38}]"} {"utterance_id": "2506-11267-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0015.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.79, 2.25, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":1.79,\"end\":2.25}]"} {"utterance_id": "4153-186223-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0010.flac", "task_type": "lexical_detection", "query": "turned", "prompt": "Once the audio has finished, identify every instance of the word \"turned\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.45, 3.82, "turned"]], "response_mode": "offline", "answer": "[{\"word\":\"turned\",\"start\":3.45,\"end\":3.82}]"} {"utterance_id": "4153-186222-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0030.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 2.08, "lawyer"]], "response_mode": "offline", "answer": "[{\"word\":\"lawyer\",\"start\":1.67,\"end\":2.08}]"} {"utterance_id": "6123-59186-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0033.flac", "task_type": "lexical_detection", "query": "expressive", "prompt": "Process the full audio offline. Find every instance of the word \"expressive\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.82, 5.44, "expressive"]], "response_mode": "offline", "answer": "[{\"word\":\"expressive\",\"start\":4.82,\"end\":5.44}]"} {"utterance_id": "6123-59150-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 2.09, "dmitri"]], "response_mode": "offline", "answer": "[{\"word\":\"dmitri\",\"start\":1.67,\"end\":2.09}]"} {"utterance_id": "6467-56885-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0016.flac", "task_type": "lexical_detection", "query": "amusing", "prompt": "Once the audio has finished, identify every instance of the word \"amusing\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.4, 2.81, "amusing"]], "response_mode": "offline", "answer": "[{\"word\":\"amusing\",\"start\":2.4,\"end\":2.81}]"} {"utterance_id": "700-122867-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0033.flac", "task_type": "lexical_detection", "query": "scarecrow", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"scarecrow\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.89, 2.46, "scarecrow"]], "response_mode": "offline", "answer": "[{\"word\":\"scarecrow\",\"start\":1.89,\"end\":2.46}]"} {"utterance_id": "4515-11057-0105", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0105.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.8, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":2.35,\"end\":2.8}]"} {"utterance_id": "1630-73710-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0000.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.36, "one"], [12.35, 12.91, "two"], [16.61, 17.18, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.05,\"end\":1.36},{\"word\":\"two\",\"start\":12.35,\"end\":12.91},{\"word\":\"three\",\"start\":16.61,\"end\":17.18}]"} {"utterance_id": "4572-112383-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0003.flac", "task_type": "lexical_detection", "query": "eve", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"eve\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.48, "eve"]], "response_mode": "offline", "answer": "[{\"word\":\"eve\",\"start\":1.06,\"end\":1.48}]"} {"utterance_id": "8254-84205-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.28, 1.64, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.28,\"end\":1.64}]"} {"utterance_id": "6123-59150-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0045.flac", "task_type": "lexical_detection", "query": "portrait", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"portrait\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.02, 1.45, "portrait"], [5.12, 5.52, "portrait"]], "response_mode": "offline", "answer": "[{\"word\":\"portrait\",\"start\":1.02,\"end\":1.45},{\"word\":\"portrait\",\"start\":5.12,\"end\":5.52}]"} {"utterance_id": "1651-136854-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0008.flac", "task_type": "lexical_detection", "query": "confident", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"confident\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 1.39, "confident"]], "response_mode": "offline", "answer": "[{\"word\":\"confident\",\"start\":0.5,\"end\":1.39}]"} {"utterance_id": "6841-88294-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0028.flac", "task_type": "lexical_detection", "query": "coat", "prompt": "Once the audio has finished, identify every instance of the word \"coat\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.84, 1.34, "coat"]], "response_mode": "offline", "answer": "[{\"word\":\"coat\",\"start\":0.84,\"end\":1.34}]"} {"utterance_id": "1686-142278-0077", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0077.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.13, 1.72, "tutor"]], "response_mode": "offline", "answer": "[{\"word\":\"tutor\",\"start\":1.13,\"end\":1.72}]"} {"utterance_id": "4831-25894-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0016.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[10.74, 11.0, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":10.74,\"end\":11.0}]"} {"utterance_id": "4515-11057-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0003.flac", "task_type": "lexical_detection", "query": "raw", "prompt": "When the utterance is complete, report all instances of the word \"raw\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.13, 13.62, "raw"]], "response_mode": "offline", "answer": "[{\"word\":\"raw\",\"start\":13.13,\"end\":13.62}]"} {"utterance_id": "7641-96684-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0018.flac", "task_type": "lexical_detection", "query": "attend", "prompt": "Once the audio has finished, identify every instance of the word \"attend\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.0, 4.32, "attend"]], "response_mode": "offline", "answer": "[{\"word\":\"attend\",\"start\":4.0,\"end\":4.32}]"} {"utterance_id": "6467-56885-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0017.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.91, 6.35, "bride"]], "response_mode": "offline", "answer": "[{\"word\":\"bride\",\"start\":5.91,\"end\":6.35}]"} {"utterance_id": "6467-97061-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0016.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Once the audio has finished, identify every instance of a color name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.47, 4.72, "silver"]], "response_mode": "offline", "answer": "[{\"word\":\"silver\",\"start\":4.47,\"end\":4.72}]"} {"utterance_id": "4831-25894-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.35, 1.68, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":1.35,\"end\":1.68}]"} {"utterance_id": "1686-142278-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0021.flac", "task_type": "lexical_detection", "query": "tell", "prompt": "After listening to the whole recording, output every instance of the word \"tell\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.68, "tell"]], "response_mode": "offline", "answer": "[{\"word\":\"tell\",\"start\":1.45,\"end\":1.68}]"} {"utterance_id": "8288-274150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0000.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.52, 3.11, "one"], [6.18, 6.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.52,\"end\":3.11},{\"word\":\"two\",\"start\":6.18,\"end\":6.91}]"} {"utterance_id": "1630-102884-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.87, 9.36, "steps"]], "response_mode": "offline", "answer": "[{\"word\":\"steps\",\"start\":8.87,\"end\":9.36}]"} {"utterance_id": "1255-74899-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0011.flac", "task_type": "lexical_detection", "query": "aristocracy", "prompt": "After listening to the whole recording, output every instance of the word \"aristocracy\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 4.2, "aristocracy"]], "response_mode": "offline", "answer": "[{\"word\":\"aristocracy\",\"start\":3.52,\"end\":4.2}]"} {"utterance_id": "6841-88291-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0001.flac", "task_type": "lexical_detection", "query": "shadow", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"shadow\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.96, 6.33, "shadow"]], "response_mode": "offline", "answer": "[{\"word\":\"shadow\",\"start\":5.96,\"end\":6.33}]"} {"utterance_id": "6455-67804-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0007.flac", "task_type": "lexical_detection", "query": "afraid", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"afraid\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.8, 2.17, "afraid"]], "response_mode": "offline", "answer": "[{\"word\":\"afraid\",\"start\":1.8,\"end\":2.17}]"} {"utterance_id": "4572-112381-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0017.flac", "task_type": "lexical_detection", "query": "foe", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"foe\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.92, 4.31, "foe"]], "response_mode": "offline", "answer": "[{\"word\":\"foe\",\"start\":3.92,\"end\":4.31}]"} {"utterance_id": "4572-64670-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0016.flac", "task_type": "lexical_detection", "query": "rites", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"rites\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.74, 9.11, "rites"]], "response_mode": "offline", "answer": "[{\"word\":\"rites\",\"start\":8.74,\"end\":9.11}]"} {"utterance_id": "3660-172183-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0010.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.01, 2.35, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":2.01,\"end\":2.35}]"} {"utterance_id": "5849-50963-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.15, 11.7, "athlete"]], "response_mode": "offline", "answer": "[{\"word\":\"athlete\",\"start\":11.15,\"end\":11.7}]"} {"utterance_id": "5543-27761-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0024.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.51, 4.81, "heart"], [9.44, 9.75, "hair"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":4.51,\"end\":4.81},{\"word\":\"hair\",\"start\":9.44,\"end\":9.75}]"} {"utterance_id": "4323-18416-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0009.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.11, "grandpapa"]], "response_mode": "offline", "answer": "[{\"word\":\"grandpapa\",\"start\":0.54,\"end\":1.11}]"} {"utterance_id": "1701-141760-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0014.flac", "task_type": "lexical_detection", "query": "rode", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"rode\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.75, 3.05, "rode"]], "response_mode": "offline", "answer": "[{\"word\":\"rode\",\"start\":2.75,\"end\":3.05}]"} {"utterance_id": "3915-98647-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0021.flac", "task_type": "lexical_detection", "query": "safely", "prompt": "After the audio ends, extract all occurrences of the word \"safely\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.26, 4.75, "safely"]], "response_mode": "offline", "answer": "[{\"word\":\"safely\",\"start\":4.26,\"end\":4.75}]"} {"utterance_id": "4153-186222-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 0.78, "beth"]], "response_mode": "offline", "answer": "[{\"word\":\"beth\",\"start\":0.53,\"end\":0.78}]"} {"utterance_id": "6467-94831-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0028.flac", "task_type": "lexical_detection", "query": "straight", "prompt": "When the utterance is complete, report all instances of the word \"straight\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.29, "straight"]], "response_mode": "offline", "answer": "[{\"word\":\"straight\",\"start\":0.9,\"end\":1.29}]"} {"utterance_id": "6467-94831-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0006.flac", "task_type": "lexical_detection", "query": "fifteen", "prompt": "When the utterance is complete, report all instances of the word \"fifteen\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.02, "fifteen"]], "response_mode": "offline", "answer": "[{\"word\":\"fifteen\",\"start\":0.65,\"end\":1.02}]"} {"utterance_id": "7601-175351-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0021.flac", "task_type": "lexical_detection", "query": "amongst", "prompt": "Once the audio has finished, identify every instance of the word \"amongst\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.54, 2.07, "amongst"]], "response_mode": "offline", "answer": "[{\"word\":\"amongst\",\"start\":1.54,\"end\":2.07}]"} {"utterance_id": "6455-67804-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0018.flac", "task_type": "lexical_detection", "query": "color", "prompt": "After the audio ends, extract all occurrences of the word \"color\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.29, "color"]], "response_mode": "offline", "answer": "[{\"word\":\"color\",\"start\":0.92,\"end\":1.29}]"} {"utterance_id": "1650-167613-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0055.flac", "task_type": "lexical_detection", "query": "joyfully", "prompt": "Once the audio has finished, identify every instance of the word \"joyfully\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.51, 3.22, "joyfully"]], "response_mode": "offline", "answer": "[{\"word\":\"joyfully\",\"start\":2.51,\"end\":3.22}]"} {"utterance_id": "6467-94831-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0017.flac", "task_type": "lexical_detection", "query": "murmured", "prompt": "After listening to the whole recording, output every instance of the word \"murmured\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.83, 2.21, "murmured"]], "response_mode": "offline", "answer": "[{\"word\":\"murmured\",\"start\":1.83,\"end\":2.21}]"} {"utterance_id": "1650-167613-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0037.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.37, "one"], [4.27, 5.04, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.9,\"end\":2.37},{\"word\":\"two\",\"start\":4.27,\"end\":5.04}]"} {"utterance_id": "8254-115543-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0013.flac", "task_type": "semantic_detection", "query": "country", "prompt": "You will hear a complete utterance. When it ends, list every instance of a country name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.47, 3.89, "india"]], "response_mode": "offline", "answer": "[{\"word\":\"india\",\"start\":3.47,\"end\":3.89}]"} {"utterance_id": "6467-94831-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0007.flac", "task_type": "lexical_detection", "query": "pop", "prompt": "Process the full audio offline. Find every instance of the word \"pop\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.38, 4.92, "pop"]], "response_mode": "offline", "answer": "[{\"word\":\"pop\",\"start\":4.38,\"end\":4.92}]"} {"utterance_id": "1630-96099-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0016.flac", "task_type": "lexical_detection", "query": "behind", "prompt": "After the audio ends, extract all occurrences of the word \"behind\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 2.0, "behind"]], "response_mode": "offline", "answer": "[{\"word\":\"behind\",\"start\":1.39,\"end\":2.0}]"} {"utterance_id": "700-122867-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0038.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.47, 1.67, "hair"], [4.78, 5.06, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"hair\",\"start\":1.47,\"end\":1.67},{\"word\":\"head\",\"start\":4.78,\"end\":5.06}]"} {"utterance_id": "3663-172005-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0006.flac", "task_type": "lexical_detection", "query": "boat", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"boat\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.72, "boat"]], "response_mode": "offline", "answer": "[{\"word\":\"boat\",\"start\":1.33,\"end\":1.72}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "lexical_detection", "query": "louvre", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"louvre\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.84, 7.14, "louvre"], [8.49, 8.86, "louvre"]], "response_mode": "offline", "answer": "[{\"word\":\"louvre\",\"start\":6.84,\"end\":7.14},{\"word\":\"louvre\",\"start\":8.49,\"end\":8.86}]"} {"utterance_id": "7601-175351-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0013.flac", "task_type": "lexical_detection", "query": "weighed", "prompt": "Once the audio has finished, identify every instance of the word \"weighed\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.52, 4.07, "weighed"]], "response_mode": "offline", "answer": "[{\"word\":\"weighed\",\"start\":3.52,\"end\":4.07}]"} {"utterance_id": "8254-115543-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0020.flac", "task_type": "lexical_detection", "query": "cloud", "prompt": "Process the full audio offline. Find every instance of the word \"cloud\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.89, 3.18, "cloud"]], "response_mode": "offline", "answer": "[{\"word\":\"cloud\",\"start\":2.89,\"end\":3.18}]"} {"utterance_id": "6455-67803-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0010.flac", "task_type": "lexical_detection", "query": "won't", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"won't\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.37, "won't"]], "response_mode": "offline", "answer": "[{\"word\":\"won't\",\"start\":2.18,\"end\":2.37}]"} {"utterance_id": "3660-172182-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0032.flac", "task_type": "lexical_detection", "query": "forth", "prompt": "Process the full audio offline. Find every instance of the word \"forth\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.64, 5.0, "forth"]], "response_mode": "offline", "answer": "[{\"word\":\"forth\",\"start\":4.64,\"end\":5.0}]"} {"utterance_id": "5849-50963-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0013.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.52, 4.81, "one"], [7.01, 7.37, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.52,\"end\":4.81},{\"word\":\"two\",\"start\":7.01,\"end\":7.37}]"} {"utterance_id": "7697-105817-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0001.flac", "task_type": "lexical_detection", "query": "spot", "prompt": "After listening to the whole recording, output every instance of the word \"spot\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.52, 4.11, "spot"]], "response_mode": "offline", "answer": "[{\"word\":\"spot\",\"start\":3.52,\"end\":4.11}]"} {"utterance_id": "1650-167613-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0002.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen carefully to the entire clip, then report every instance of an organization name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.8, 8.26, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"army\",\"start\":7.8,\"end\":8.26}]"} {"utterance_id": "1585-131718-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0042.flac", "task_type": "lexical_detection", "query": "notice", "prompt": "After listening to the whole recording, output every instance of the word \"notice\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.65, "notice"]], "response_mode": "offline", "answer": "[{\"word\":\"notice\",\"start\":2.11,\"end\":2.65}]"} {"utterance_id": "5849-50962-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0007.flac", "task_type": "lexical_detection", "query": "clear", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"clear\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.15, "clear"]], "response_mode": "offline", "answer": "[{\"word\":\"clear\",\"start\":0.8,\"end\":1.15}]"} {"utterance_id": "1651-136854-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0027.flac", "task_type": "lexical_detection", "query": "mention", "prompt": "Once the audio has finished, identify every instance of the word \"mention\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.56, 1.98, "mention"]], "response_mode": "offline", "answer": "[{\"word\":\"mention\",\"start\":1.56,\"end\":1.98}]"} {"utterance_id": "8173-294714-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0058.flac", "task_type": "lexical_detection", "query": "resolution", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"resolution\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.43, "resolution"]], "response_mode": "offline", "answer": "[{\"word\":\"resolution\",\"start\":1.74,\"end\":2.43}]"} {"utterance_id": "7641-96670-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0020.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.28, 4.43, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":4.28,\"end\":4.43}]"} {"utterance_id": "1255-74899-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0007.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 1.76, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":1.43,\"end\":1.76}]"} {"utterance_id": "1686-142278-0087", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0087.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 1.03, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.51,\"end\":1.03}]"} {"utterance_id": "1585-131718-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0030.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[13.01, 14.53, "wooster street new york"]], "response_mode": "offline", "answer": "[{\"word\":\"wooster street new york\",\"start\":13.01,\"end\":14.53}]"} {"utterance_id": "6455-67803-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0003.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.8, "grand uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"grand uncle\",\"start\":1.08,\"end\":1.8}]"} {"utterance_id": "700-122868-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.27, "anne shirley"]], "response_mode": "offline", "answer": "[{\"word\":\"anne shirley\",\"start\":0.52,\"end\":1.27}]"} {"utterance_id": "4831-18525-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0020.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.39, 1.97, "granddaddy"]], "response_mode": "offline", "answer": "[{\"word\":\"granddaddy\",\"start\":1.39,\"end\":1.97}]"} {"utterance_id": "4570-14911-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0014.flac", "task_type": "lexical_detection", "query": "bone", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"bone\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.3, 6.88, "bone"]], "response_mode": "offline", "answer": "[{\"word\":\"bone\",\"start\":6.3,\"end\":6.88}]"} {"utterance_id": "1630-73710-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0000.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "When the utterance is complete, report all instances of a plant name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.74, "laurels"]], "response_mode": "offline", "answer": "[{\"word\":\"laurels\",\"start\":3.06,\"end\":3.74}]"} {"utterance_id": "3660-172182-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0003.flac", "task_type": "lexical_counting", "query": "best", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"best\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.06, 2.44, "one"], [5.91, 6.4, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.06,\"end\":2.44},{\"word\":\"two\",\"start\":5.91,\"end\":6.4}]"} {"utterance_id": "4153-186223-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0014.flac", "task_type": "lexical_detection", "query": "john", "prompt": "When the utterance is complete, report all instances of the word \"john\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.91, "john"]], "response_mode": "offline", "answer": "[{\"word\":\"john\",\"start\":0.33,\"end\":0.91}]"} {"utterance_id": "4515-11057-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.52, 1.86, "biffen"]], "response_mode": "offline", "answer": "[{\"word\":\"biffen\",\"start\":1.52,\"end\":1.86}]"} {"utterance_id": "1701-141760-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0003.flac", "task_type": "lexical_detection", "query": "marched", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"marched\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.33, 2.72, "marched"]], "response_mode": "offline", "answer": "[{\"word\":\"marched\",\"start\":2.33,\"end\":2.72}]"} {"utterance_id": "4323-55228-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0014.flac", "task_type": "lexical_counting", "query": "hall", "prompt": "When the utterance is complete, report how the word \"hall\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 0.93, "one"], [4.19, 4.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.66,\"end\":0.93},{\"word\":\"two\",\"start\":4.19,\"end\":4.42}]"} {"utterance_id": "4572-112375-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.25, 4.59, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":4.25,\"end\":4.59}]"} {"utterance_id": "4831-25894-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0021.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 0.91, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.7,\"end\":0.91}]"} {"utterance_id": "700-122868-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0032.flac", "task_type": "lexical_detection", "query": "friends", "prompt": "After listening to the whole recording, output every instance of the word \"friends\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.57, 3.86, "friends"]], "response_mode": "offline", "answer": "[{\"word\":\"friends\",\"start\":3.57,\"end\":3.86}]"} {"utterance_id": "3663-172528-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0033.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.16, 4.61, "bird"]], "response_mode": "offline", "answer": "[{\"word\":\"bird\",\"start\":4.16,\"end\":4.61}]"} {"utterance_id": "1651-136854-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.95, "johnson's"]], "response_mode": "offline", "answer": "[{\"word\":\"johnson's\",\"start\":0.47,\"end\":0.95}]"} {"utterance_id": "6841-88294-0063", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0063.flac", "task_type": "lexical_detection", "query": "somebody", "prompt": "Once the audio has finished, identify every instance of the word \"somebody\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.74, 5.24, "somebody"]], "response_mode": "offline", "answer": "[{\"word\":\"somebody\",\"start\":4.74,\"end\":5.24}]"} {"utterance_id": "116-288047-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0022.flac", "task_type": "lexical_detection", "query": "millions", "prompt": "Process the full audio offline. Find every instance of the word \"millions\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.44, "millions"]], "response_mode": "offline", "answer": "[{\"word\":\"millions\",\"start\":5.09,\"end\":5.44}]"} {"utterance_id": "1686-142278-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.93, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.5,\"end\":0.93}]"} {"utterance_id": "4323-55228-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0023.flac", "task_type": "lexical_detection", "query": "wrote", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wrote\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.37, "wrote"]], "response_mode": "offline", "answer": "[{\"word\":\"wrote\",\"start\":1.14,\"end\":1.37}]"} {"utterance_id": "2506-11267-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 2.28, "nathan houndsditch"]], "response_mode": "offline", "answer": "[{\"word\":\"nathan houndsditch\",\"start\":1.27,\"end\":2.28}]"} {"utterance_id": "7601-291468-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0006.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[27.94, 28.42, "wind"]], "response_mode": "offline", "answer": "[{\"word\":\"wind\",\"start\":27.94,\"end\":28.42}]"} {"utterance_id": "4323-55228-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0020.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.34, 2.58, "arm"]], "response_mode": "offline", "answer": "[{\"word\":\"arm\",\"start\":2.34,\"end\":2.58}]"} {"utterance_id": "7641-96684-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0013.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.25, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":1.99,\"end\":2.25}]"} {"utterance_id": "6455-66379-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0014.flac", "task_type": "lexical_detection", "query": "cords", "prompt": "After the audio ends, extract all occurrences of the word \"cords\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.46, "cords"]], "response_mode": "offline", "answer": "[{\"word\":\"cords\",\"start\":1.07,\"end\":1.46}]"} {"utterance_id": "4572-112375-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0001.flac", "task_type": "lexical_detection", "query": "swift", "prompt": "Process the full audio offline. Find every instance of the word \"swift\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.99, "swift"]], "response_mode": "offline", "answer": "[{\"word\":\"swift\",\"start\":0.49,\"end\":0.99}]"} {"utterance_id": "700-122867-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0032.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.53, 0.8, "one"], [7.1, 7.77, "two"], [10.45, 10.76, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.53,\"end\":0.8},{\"word\":\"two\",\"start\":7.1,\"end\":7.77},{\"word\":\"three\",\"start\":10.45,\"end\":10.76}]"} {"utterance_id": "4323-55228-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 0.69, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.33,\"end\":0.69}]"} {"utterance_id": "4515-11057-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0025.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.0, 2.43, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":2.0,\"end\":2.43}]"} {"utterance_id": "3660-6517-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0021.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Do not answer until the audio ends. Then list each instance of an animal name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.88, 6.15, "beast"]], "response_mode": "offline", "answer": "[{\"word\":\"beast\",\"start\":5.88,\"end\":6.15}]"} {"utterance_id": "3663-172528-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.57, 3.56, "bernardino"]], "response_mode": "offline", "answer": "[{\"word\":\"bernardino\",\"start\":2.57,\"end\":3.56}]"} {"utterance_id": "4572-112375-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.55, 1.1, "tara's"]], "response_mode": "offline", "answer": "[{\"word\":\"tara's\",\"start\":0.55,\"end\":1.1}]"} {"utterance_id": "7641-96670-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0018.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.64, 3.08, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":2.64,\"end\":3.08}]"} {"utterance_id": "4570-24733-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.09, 5.62, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":5.09,\"end\":5.62}]"} {"utterance_id": "1255-138279-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0001.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Process the full audio offline. Find every instance of a plant name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.3, 13.81, "buds"], [14.83, 15.27, "leaf"]], "response_mode": "offline", "answer": "[{\"word\":\"buds\",\"start\":13.3,\"end\":13.81},{\"word\":\"leaf\",\"start\":14.83,\"end\":15.27}]"} {"utterance_id": "8254-115543-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0025.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "After the audio ends, tally every instance of an animal name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.72, 2.17, "one"], [3.26, 3.65, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.72,\"end\":2.17},{\"word\":\"two\",\"start\":3.26,\"end\":3.65}]"} {"utterance_id": "700-122866-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0021.flac", "task_type": "lexical_detection", "query": "story", "prompt": "Once the audio has finished, identify every instance of the word \"story\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.53, 2.04, "story"]], "response_mode": "offline", "answer": "[{\"word\":\"story\",\"start\":1.53,\"end\":2.04}]"} {"utterance_id": "1651-136854-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0009.flac", "task_type": "lexical_detection", "query": "honours", "prompt": "Process the full audio offline. Find every instance of the word \"honours\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.45, "honours"]], "response_mode": "offline", "answer": "[{\"word\":\"honours\",\"start\":2.76,\"end\":3.45}]"} {"utterance_id": "1585-131718-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0011.flac", "task_type": "lexical_counting", "query": "lighting", "prompt": "When the utterance is complete, report how the word \"lighting\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.33, 4.67, "one"], [8.46, 8.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.33,\"end\":4.67},{\"word\":\"two\",\"start\":8.46,\"end\":8.92}]"} {"utterance_id": "6467-94831-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.6, 11.92, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":11.6,\"end\":11.92}]"} {"utterance_id": "7641-96252-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0017.flac", "task_type": "lexical_detection", "query": "socrates", "prompt": "After the audio ends, extract all occurrences of the word \"socrates\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.75, "socrates"]], "response_mode": "offline", "answer": "[{\"word\":\"socrates\",\"start\":1.09,\"end\":1.75}]"} {"utterance_id": "1650-167613-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.29, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":1.84,\"end\":2.29}]"} {"utterance_id": "1585-131718-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0003.flac", "task_type": "lexical_detection", "query": "invention", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"invention\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.49, "invention"]], "response_mode": "offline", "answer": "[{\"word\":\"invention\",\"start\":2.06,\"end\":2.49}]"} {"utterance_id": "4570-14911-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 1.06, "fore legs"], [6.3, 6.88, "bone"]], "response_mode": "offline", "answer": "[{\"word\":\"fore legs\",\"start\":0.53,\"end\":1.06},{\"word\":\"bone\",\"start\":6.3,\"end\":6.88}]"} {"utterance_id": "116-288045-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0020.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.51, 1.74, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":1.51,\"end\":1.74}]"} {"utterance_id": "1686-142278-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.71, 0.97, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":0.71,\"end\":0.97}]"} {"utterance_id": "700-122866-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.22, 3.01, "ruby gillis"]], "response_mode": "offline", "answer": "[{\"word\":\"ruby gillis\",\"start\":2.22,\"end\":3.01}]"} {"utterance_id": "5849-50962-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0024.flac", "task_type": "lexical_detection", "query": "hal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"hal\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.63, 0.9, "hal"]], "response_mode": "offline", "answer": "[{\"word\":\"hal\",\"start\":0.63,\"end\":0.9}]"} {"utterance_id": "6467-56885-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0010.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a body part with word/start/end. If none, reply [].", "labels": [[2.0, 2.51, "one"], [13.42, 13.89, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.0,\"end\":2.51},{\"word\":\"two\",\"start\":13.42,\"end\":13.89}]"} {"utterance_id": "1255-90407-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0023.flac", "task_type": "lexical_detection", "query": "bloom", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"bloom\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.45, 6.9, "bloom"]], "response_mode": "offline", "answer": "[{\"word\":\"bloom\",\"start\":6.45,\"end\":6.9}]"} {"utterance_id": "3663-172528-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0028.flac", "task_type": "lexical_detection", "query": "inconvenience", "prompt": "After listening to the whole recording, output every instance of the word \"inconvenience\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.83, 8.86, "inconvenience"]], "response_mode": "offline", "answer": "[{\"word\":\"inconvenience\",\"start\":7.83,\"end\":8.86}]"} {"utterance_id": "1701-141759-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0002.flac", "task_type": "lexical_detection", "query": "receiving", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"receiving\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.65, 1.16, "receiving"]], "response_mode": "offline", "answer": "[{\"word\":\"receiving\",\"start\":0.65,\"end\":1.16}]"} {"utterance_id": "7697-105815-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0007.flac", "task_type": "lexical_detection", "query": "means", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"means\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.76, 2.1, "means"]], "response_mode": "offline", "answer": "[{\"word\":\"means\",\"start\":1.76,\"end\":2.1}]"} {"utterance_id": "1650-173551-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0005.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.45, 5.75, "head"], [8.77, 9.03, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":5.45,\"end\":5.75},{\"word\":\"head\",\"start\":8.77,\"end\":9.03}]"} {"utterance_id": "1701-141760-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0012.flac", "task_type": "lexical_detection", "query": "society", "prompt": "When the utterance is complete, report all instances of the word \"society\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.32, 11.05, "society"]], "response_mode": "offline", "answer": "[{\"word\":\"society\",\"start\":10.32,\"end\":11.05}]"} {"utterance_id": "4572-64670-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0000.flac", "task_type": "lexical_detection", "query": "host", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"host\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.6, "host"]], "response_mode": "offline", "answer": "[{\"word\":\"host\",\"start\":1.08,\"end\":1.6}]"} {"utterance_id": "4570-14911-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0011.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.97, "one"], [8.54, 8.88, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.54,\"end\":0.97},{\"word\":\"two\",\"start\":8.54,\"end\":8.88}]"} {"utterance_id": "700-122866-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0036.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 0.71, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.57,\"end\":0.71}]"} {"utterance_id": "6599-38591-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0002.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.01, 2.51, "cambrai"]], "response_mode": "offline", "answer": "[{\"word\":\"cambrai\",\"start\":2.01,\"end\":2.51}]"} {"utterance_id": "4831-18525-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0024.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.47, "one"], [2.63, 3.04, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.16,\"end\":0.47},{\"word\":\"two\",\"start\":2.63,\"end\":3.04}]"} {"utterance_id": "3660-6517-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0011.flac", "task_type": "lexical_counting", "query": "done", "prompt": "Once the audio has finished, list each occurrence of the word \"done\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[9.82, 10.1, "one"], [11.63, 11.89, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.82,\"end\":10.1},{\"word\":\"two\",\"start\":11.63,\"end\":11.89}]"} {"utterance_id": "8254-84205-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0035.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.69, 1.15, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":0.69,\"end\":1.15}]"} {"utterance_id": "6841-88291-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0046.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.13, 1.46, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":1.13,\"end\":1.46}]"} {"utterance_id": "6599-38591-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0000.flac", "task_type": "lexical_detection", "query": "horseback", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"horseback\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.7, 6.15, "horseback"]], "response_mode": "offline", "answer": "[{\"word\":\"horseback\",\"start\":5.7,\"end\":6.15}]"} {"utterance_id": "7641-96684-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0032.flac", "task_type": "lexical_detection", "query": "temple", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"temple\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.79, 3.17, "temple"]], "response_mode": "offline", "answer": "[{\"word\":\"temple\",\"start\":2.79,\"end\":3.17}]"} {"utterance_id": "6841-88294-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0015.flac", "task_type": "lexical_detection", "query": "pay", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pay\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.66, 0.84, "pay"]], "response_mode": "offline", "answer": "[{\"word\":\"pay\",\"start\":0.66,\"end\":0.84}]"} {"utterance_id": "6455-67803-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0003.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Once the audio has finished, identify every instance of a money or currency term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.28, 5.65, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":5.28,\"end\":5.65}]"} {"utterance_id": "1650-167613-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0036.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen carefully to the entire clip, then report every instance of a city name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.43, 6.09, "tilsit"]], "response_mode": "offline", "answer": "[{\"word\":\"tilsit\",\"start\":5.43,\"end\":6.09}]"} {"utterance_id": "8254-84205-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0026.flac", "task_type": "lexical_detection", "query": "griggs", "prompt": "After listening to the whole recording, output every instance of the word \"griggs\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.64, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":1.28,\"end\":1.64}]"} {"utterance_id": "8254-115543-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0012.flac", "task_type": "lexical_detection", "query": "accompanied", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"accompanied\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 3.12, "accompanied"]], "response_mode": "offline", "answer": "[{\"word\":\"accompanied\",\"start\":2.58,\"end\":3.12}]"} {"utterance_id": "5849-50873-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.37, "charles"]], "response_mode": "offline", "answer": "[{\"word\":\"charles\",\"start\":0.87,\"end\":1.37}]"} {"utterance_id": "1686-142278-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0019.flac", "task_type": "lexical_detection", "query": "longer", "prompt": "Process the full audio offline. Find every instance of the word \"longer\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.19, "longer"]], "response_mode": "offline", "answer": "[{\"word\":\"longer\",\"start\":0.91,\"end\":1.19}]"} {"utterance_id": "3660-172182-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0016.flac", "task_type": "lexical_detection", "query": "grant", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"grant\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 3.66, "grant"]], "response_mode": "offline", "answer": "[{\"word\":\"grant\",\"start\":3.36,\"end\":3.66}]"} {"utterance_id": "8254-84205-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0051.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.25, 3.56, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":3.25,\"end\":3.56}]"} {"utterance_id": "700-122867-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.55, 3.02, "marilla's"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla's\",\"start\":2.55,\"end\":3.02}]"} {"utterance_id": "3915-57461-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0021.flac", "task_type": "lexical_detection", "query": "mistaken", "prompt": "Process the full audio offline. Find every instance of the word \"mistaken\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.05, 5.73, "mistaken"]], "response_mode": "offline", "answer": "[{\"word\":\"mistaken\",\"start\":5.05,\"end\":5.73}]"} {"utterance_id": "4515-11057-0107", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0107.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 0.92, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":0.54,\"end\":0.92}]"} {"utterance_id": "6455-67804-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0001.flac", "task_type": "lexical_detection", "query": "joyous", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"joyous\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.03, 1.3, "joyous"]], "response_mode": "offline", "answer": "[{\"word\":\"joyous\",\"start\":1.03,\"end\":1.3}]"} {"utterance_id": "5849-50873-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0008.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.06, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.57,\"end\":2.06}]"} {"utterance_id": "4153-186222-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0026.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 4.03, "nieces"]], "response_mode": "offline", "answer": "[{\"word\":\"nieces\",\"start\":3.56,\"end\":4.03}]"} {"utterance_id": "6841-88291-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 1.07, "mister frost's"]], "response_mode": "offline", "answer": "[{\"word\":\"mister frost's\",\"start\":0.34,\"end\":1.07}]"} {"utterance_id": "1650-167613-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 3.43, "barclay de tolly"]], "response_mode": "offline", "answer": "[{\"word\":\"barclay de tolly\",\"start\":2.1,\"end\":3.43}]"} {"utterance_id": "6123-59186-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0036.flac", "task_type": "lexical_detection", "query": "lovers", "prompt": "After listening to the whole recording, output every instance of the word \"lovers\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.76, "lovers"]], "response_mode": "offline", "answer": "[{\"word\":\"lovers\",\"start\":0.4,\"end\":0.76}]"} {"utterance_id": "4323-55228-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0016.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.58, 5.89, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.58,\"end\":5.89}]"} {"utterance_id": "8288-274162-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0020.flac", "task_type": "lexical_detection", "query": "whose", "prompt": "After listening to the whole recording, output every instance of the word \"whose\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.87, 5.08, "whose"]], "response_mode": "offline", "answer": "[{\"word\":\"whose\",\"start\":4.87,\"end\":5.08}]"} {"utterance_id": "4570-24733-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0008.flac", "task_type": "lexical_detection", "query": "attributes", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"attributes\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.4, 7.21, "attributes"], [10.0, 10.82, "attributes"]], "response_mode": "offline", "answer": "[{\"word\":\"attributes\",\"start\":6.4,\"end\":7.21},{\"word\":\"attributes\",\"start\":10.0,\"end\":10.82}]"} {"utterance_id": "4570-14911-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0002.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.08, 4.45, "colour"]], "response_mode": "offline", "answer": "[{\"word\":\"colour\",\"start\":4.08,\"end\":4.45}]"} {"utterance_id": "1686-142278-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0035.flac", "task_type": "lexical_detection", "query": "borne", "prompt": "Once the audio has finished, identify every instance of the word \"borne\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.78, 1.07, "borne"]], "response_mode": "offline", "answer": "[{\"word\":\"borne\",\"start\":0.78,\"end\":1.07}]"} {"utterance_id": "3660-172183-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0026.flac", "task_type": "lexical_detection", "query": "drank", "prompt": "After listening to the whole recording, output every instance of the word \"drank\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.68, "drank"]], "response_mode": "offline", "answer": "[{\"word\":\"drank\",\"start\":1.28,\"end\":1.68}]"} {"utterance_id": "5849-50963-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0006.flac", "task_type": "lexical_counting", "query": "last", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"last\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.75, 2.1, "one"], [3.13, 3.44, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.75,\"end\":2.1},{\"word\":\"two\",\"start\":3.13,\"end\":3.44}]"} {"utterance_id": "2506-11278-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0034.flac", "task_type": "lexical_detection", "query": "natures", "prompt": "After the audio ends, extract all occurrences of the word \"natures\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 2.13, "natures"]], "response_mode": "offline", "answer": "[{\"word\":\"natures\",\"start\":1.51,\"end\":2.13}]"} {"utterance_id": "700-122867-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0040.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[0.66, 0.87, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":0.66,\"end\":0.87}]"} {"utterance_id": "5849-50873-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0022.flac", "task_type": "lexical_detection", "query": "possible", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"possible\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.8, 5.45, "possible"]], "response_mode": "offline", "answer": "[{\"word\":\"possible\",\"start\":4.8,\"end\":5.45}]"} {"utterance_id": "4831-18525-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0021.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an ordinal number with fields word, start, end (seconds). If none, reply [].", "labels": [[3.01, 3.4, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":3.01,\"end\":3.4}]"} {"utterance_id": "1255-74899-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0015.flac", "task_type": "lexical_detection", "query": "ended", "prompt": "After listening to the whole recording, output every instance of the word \"ended\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.84, 7.26, "ended"]], "response_mode": "offline", "answer": "[{\"word\":\"ended\",\"start\":6.84,\"end\":7.26}]"} {"utterance_id": "4323-55228-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0048.flac", "task_type": "semantic_detection", "query": "date", "prompt": "When the utterance is complete, report all instances of a date including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.72, "october"]], "response_mode": "offline", "answer": "[{\"word\":\"october\",\"start\":2.02,\"end\":2.72}]"} {"utterance_id": "1701-141760-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0033.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Listen to the full audio. After it finishes, find every instance of a state or province name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.63, 3.39, "siberia"]], "response_mode": "offline", "answer": "[{\"word\":\"siberia\",\"start\":2.63,\"end\":3.39}]"} {"utterance_id": "1630-141772-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0002.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.65, 1.19, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.65,\"end\":1.19}]"} {"utterance_id": "4831-25894-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0030.flac", "task_type": "lexical_detection", "query": "warm", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"warm\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.45, 7.82, "warm"]], "response_mode": "offline", "answer": "[{\"word\":\"warm\",\"start\":7.45,\"end\":7.82}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.63, 4.01, "polly"], [9.13, 9.57, "adela"], [10.7, 11.46, "polly pepper"]], "response_mode": "offline", "answer": "[{\"word\":\"polly\",\"start\":3.63,\"end\":4.01},{\"word\":\"adela\",\"start\":9.13,\"end\":9.57},{\"word\":\"polly pepper\",\"start\":10.7,\"end\":11.46}]"} {"utterance_id": "3663-172528-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0044.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 3.23, "duke alessandro"], [4.86, 5.83, "pope clement"]], "response_mode": "offline", "answer": "[{\"word\":\"duke alessandro\",\"start\":1.95,\"end\":3.23},{\"word\":\"pope clement\",\"start\":4.86,\"end\":5.83}]"} {"utterance_id": "8254-84205-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.27, 3.49, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":3.27,\"end\":3.49}]"} {"utterance_id": "700-122868-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0004.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a body part with word/start/end. If none, reply [].", "labels": [[0.7, 1.25, "one"], [4.76, 4.99, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.7,\"end\":1.25},{\"word\":\"two\",\"start\":4.76,\"end\":4.99}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "lexical_detection", "query": "american", "prompt": "After the audio ends, extract all occurrences of the word \"american\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.5, 15.11, "american"]], "response_mode": "offline", "answer": "[{\"word\":\"american\",\"start\":14.5,\"end\":15.11}]"} {"utterance_id": "7697-245712-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0014.flac", "task_type": "lexical_detection", "query": "primitive", "prompt": "When the utterance is complete, report all instances of the word \"primitive\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.63, "primitive"]], "response_mode": "offline", "answer": "[{\"word\":\"primitive\",\"start\":3.21,\"end\":3.63}]"} {"utterance_id": "4515-11057-0092", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0092.flac", "task_type": "lexical_detection", "query": "anything", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"anything\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.19, 1.89, "anything"]], "response_mode": "offline", "answer": "[{\"word\":\"anything\",\"start\":1.19,\"end\":1.89}]"} {"utterance_id": "700-122866-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0024.flac", "task_type": "lexical_detection", "query": "grew", "prompt": "When the utterance is complete, report all instances of the word \"grew\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.92, "grew"]], "response_mode": "offline", "answer": "[{\"word\":\"grew\",\"start\":0.63,\"end\":0.92}]"} {"utterance_id": "8254-115543-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0010.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.42, "birds"]], "response_mode": "offline", "answer": "[{\"word\":\"birds\",\"start\":1.04,\"end\":1.42}]"} {"utterance_id": "7641-96670-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0005.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.71, 2.09, "mist"]], "response_mode": "offline", "answer": "[{\"word\":\"mist\",\"start\":1.71,\"end\":2.09}]"} {"utterance_id": "6467-62797-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/62797/6467-62797-0005.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.87, 5.09, "eighty"]], "response_mode": "offline", "answer": "[{\"word\":\"eighty\",\"start\":4.87,\"end\":5.09}]"} {"utterance_id": "7641-96684-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0018.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.49, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.16,\"end\":0.49}]"} {"utterance_id": "700-122868-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0006.flac", "task_type": "lexical_detection", "query": "anne's", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"anne's\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.02, "anne's"]], "response_mode": "offline", "answer": "[{\"word\":\"anne's\",\"start\":0.77,\"end\":1.02}]"} {"utterance_id": "6467-56885-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0013.flac", "task_type": "lexical_detection", "query": "close", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"close\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.38, 0.64, "close"]], "response_mode": "offline", "answer": "[{\"word\":\"close\",\"start\":0.38,\"end\":0.64}]"} {"utterance_id": "2506-11267-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.65, 3.24, "snooks"]], "response_mode": "offline", "answer": "[{\"word\":\"snooks\",\"start\":2.65,\"end\":3.24}]"} {"utterance_id": "6599-38590-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0007.flac", "task_type": "lexical_detection", "query": "king", "prompt": "Once the audio has finished, identify every instance of the word \"king\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.7, 2.02, "king"]], "response_mode": "offline", "answer": "[{\"word\":\"king\",\"start\":1.7,\"end\":2.02}]"} {"utterance_id": "1701-141759-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0016.flac", "task_type": "lexical_detection", "query": "speaking", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"speaking\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.49, 1.86, "speaking"], [4.8, 5.26, "speaking"]], "response_mode": "offline", "answer": "[{\"word\":\"speaking\",\"start\":1.49,\"end\":1.86},{\"word\":\"speaking\",\"start\":4.8,\"end\":5.26}]"} {"utterance_id": "1255-90407-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0015.flac", "task_type": "lexical_detection", "query": "jocelyn", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"jocelyn\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.54, 0.96, "jocelyn"]], "response_mode": "offline", "answer": "[{\"word\":\"jocelyn\",\"start\":0.54,\"end\":0.96}]"} {"utterance_id": "6467-97061-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0000.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "After the audio ends, extract all occurrences of a time of day. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.84, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":0.48,\"end\":0.84}]"} {"utterance_id": "6467-97061-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0017.flac", "task_type": "lexical_detection", "query": "silver", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"silver\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.89, 2.26, "silver"]], "response_mode": "offline", "answer": "[{\"word\":\"silver\",\"start\":1.89,\"end\":2.26}]"} {"utterance_id": "4831-18525-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0016.flac", "task_type": "lexical_detection", "query": "dashing", "prompt": "When the utterance is complete, report all instances of the word \"dashing\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.85, 10.27, "dashing"]], "response_mode": "offline", "answer": "[{\"word\":\"dashing\",\"start\":9.85,\"end\":10.27}]"} {"utterance_id": "4572-64670-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0020.flac", "task_type": "lexical_detection", "query": "depend", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"depend\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.42, 2.93, "depend"]], "response_mode": "offline", "answer": "[{\"word\":\"depend\",\"start\":2.42,\"end\":2.93}]"} {"utterance_id": "3660-6517-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0027.flac", "task_type": "lexical_detection", "query": "centre", "prompt": "When the utterance is complete, report all instances of the word \"centre\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.04, 4.46, "centre"]], "response_mode": "offline", "answer": "[{\"word\":\"centre\",\"start\":4.04,\"end\":4.46}]"} {"utterance_id": "3660-172182-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0019.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.45, 0.73, "steward"]], "response_mode": "offline", "answer": "[{\"word\":\"steward\",\"start\":0.45,\"end\":0.73}]"} {"utterance_id": "1255-74899-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0007.flac", "task_type": "lexical_detection", "query": "refuse", "prompt": "After the audio ends, extract all occurrences of the word \"refuse\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.03, 3.34, "refuse"]], "response_mode": "offline", "answer": "[{\"word\":\"refuse\",\"start\":3.03,\"end\":3.34}]"} {"utterance_id": "6123-59150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0001.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.7, 1.35, "one"], [2.96, 3.55, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.7,\"end\":1.35},{\"word\":\"two\",\"start\":2.96,\"end\":3.55}]"} {"utterance_id": "4831-18525-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0027.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After listening to the whole recording, output timed count words for a person's name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.46, 1.76, "one"], [5.51, 5.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.46,\"end\":1.76},{\"word\":\"two\",\"start\":5.51,\"end\":5.86}]"} {"utterance_id": "4831-18525-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0003.flac", "task_type": "lexical_detection", "query": "tom", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"tom\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.25, "tom"], [5.12, 5.56, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":1.9,\"end\":2.25},{\"word\":\"tom\",\"start\":5.12,\"end\":5.56}]"} {"utterance_id": "1686-142278-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.77, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.34,\"end\":0.77}]"} {"utterance_id": "116-288045-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0029.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.56, 10.02, "golden"]], "response_mode": "offline", "answer": "[{\"word\":\"golden\",\"start\":9.56,\"end\":10.02}]"} {"utterance_id": "4572-112381-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0013.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.34, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":1.95,\"end\":2.34}]"} {"utterance_id": "8254-115543-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0036.flac", "task_type": "lexical_detection", "query": "beating", "prompt": "When the utterance is complete, report all instances of the word \"beating\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.75, 4.03, "beating"]], "response_mode": "offline", "answer": "[{\"word\":\"beating\",\"start\":3.75,\"end\":4.03}]"} {"utterance_id": "6841-88294-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0018.flac", "task_type": "lexical_detection", "query": "clothes", "prompt": "When the utterance is complete, report all instances of the word \"clothes\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.31, "clothes"]], "response_mode": "offline", "answer": "[{\"word\":\"clothes\",\"start\":0.83,\"end\":1.31}]"} {"utterance_id": "4515-11057-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0098.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.54, 1.27, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":0.54,\"end\":1.27}]"} {"utterance_id": "4570-14911-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0001.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a number (word, start, end). If none, reply [].", "labels": [[3.44, 4.02, "one"], [11.04, 11.4, "two"], [13.33, 13.79, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.44,\"end\":4.02},{\"word\":\"two\",\"start\":11.04,\"end\":11.4},{\"word\":\"three\",\"start\":13.33,\"end\":13.79}]"} {"utterance_id": "7697-105815-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0047.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.89, "charity"], [3.34, 3.92, "christian"]], "response_mode": "offline", "answer": "[{\"word\":\"charity\",\"start\":0.44,\"end\":0.89},{\"word\":\"christian\",\"start\":3.34,\"end\":3.92}]"} {"utterance_id": "4323-13259-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0023.flac", "task_type": "lexical_detection", "query": "railroad", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"railroad\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.36, 1.79, "railroad"]], "response_mode": "offline", "answer": "[{\"word\":\"railroad\",\"start\":1.36,\"end\":1.79}]"} {"utterance_id": "4515-11057-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0000.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Once the audio has finished, identify every instance of a money or currency term. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.45, 5.1, "fortune"]], "response_mode": "offline", "answer": "[{\"word\":\"fortune\",\"start\":4.45,\"end\":5.1}]"} {"utterance_id": "3663-172528-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0011.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[2.61, 3.22, "physicians"]], "response_mode": "offline", "answer": "[{\"word\":\"physicians\",\"start\":2.61,\"end\":3.22}]"} {"utterance_id": "116-288047-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0004.flac", "task_type": "lexical_detection", "query": "replaced", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"replaced\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.77, 3.29, "replaced"]], "response_mode": "offline", "answer": "[{\"word\":\"replaced\",\"start\":2.77,\"end\":3.29}]"} {"utterance_id": "3915-57461-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0004.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "You will hear a complete utterance. When it ends, list every instance of a weather term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.64, 7.93, "wind"]], "response_mode": "offline", "answer": "[{\"word\":\"wind\",\"start\":7.64,\"end\":7.93}]"} {"utterance_id": "7641-96252-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0006.flac", "task_type": "lexical_detection", "query": "distinction", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"distinction\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.11, 1.83, "distinction"]], "response_mode": "offline", "answer": "[{\"word\":\"distinction\",\"start\":1.11,\"end\":1.83}]"} {"utterance_id": "8173-294714-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0020.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[4.86, 5.31, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":4.86,\"end\":5.31}]"} {"utterance_id": "4831-18525-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0004.flac", "task_type": "lexical_detection", "query": "completed", "prompt": "When the utterance is complete, report all instances of the word \"completed\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.12, 6.61, "completed"]], "response_mode": "offline", "answer": "[{\"word\":\"completed\",\"start\":6.12,\"end\":6.61}]"} {"utterance_id": "1585-131718-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0010.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the full audio. Count every occurrence of a number. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.33, "one"], [8.92, 9.63, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.06,\"end\":1.33},{\"word\":\"two\",\"start\":8.92,\"end\":9.63}]"} {"utterance_id": "7697-105815-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0042.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After the audio ends, extract all occurrences of a body part. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.73, 1.99, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.73,\"end\":1.99}]"} {"utterance_id": "6267-53049-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0000.flac", "task_type": "lexical_detection", "query": "brains", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"brains\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.89, 3.38, "brains"]], "response_mode": "offline", "answer": "[{\"word\":\"brains\",\"start\":2.89,\"end\":3.38}]"} {"utterance_id": "3663-172528-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0037.flac", "task_type": "lexical_detection", "query": "marvelled", "prompt": "Process the full audio offline. Find every instance of the word \"marvelled\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.54, 3.16, "marvelled"]], "response_mode": "offline", "answer": "[{\"word\":\"marvelled\",\"start\":2.54,\"end\":3.16}]"} {"utterance_id": "1686-142278-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0039.flac", "task_type": "lexical_detection", "query": "humiliation", "prompt": "When the utterance is complete, report all instances of the word \"humiliation\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.78, 5.61, "humiliation"]], "response_mode": "offline", "answer": "[{\"word\":\"humiliation\",\"start\":4.78,\"end\":5.61}]"} {"utterance_id": "7697-245715-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0000.flac", "task_type": "lexical_detection", "query": "age", "prompt": "After listening to the whole recording, output every instance of the word \"age\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.82, 4.2, "age"]], "response_mode": "offline", "answer": "[{\"word\":\"age\",\"start\":3.82,\"end\":4.2}]"} {"utterance_id": "1255-138279-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0007.flac", "task_type": "lexical_detection", "query": "spoken", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"spoken\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 2.67, "spoken"]], "response_mode": "offline", "answer": "[{\"word\":\"spoken\",\"start\":2.3,\"end\":2.67}]"} {"utterance_id": "1686-142278-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0040.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.53, 5.05, "helstone"]], "response_mode": "offline", "answer": "[{\"word\":\"helstone\",\"start\":4.53,\"end\":5.05}]"} {"utterance_id": "116-288045-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0007.flac", "task_type": "lexical_detection", "query": "exist", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"exist\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.75, 2.49, "exist"]], "response_mode": "offline", "answer": "[{\"word\":\"exist\",\"start\":1.75,\"end\":2.49}]"} {"utterance_id": "3660-6517-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0024.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.09, 4.46, "ponies"]], "response_mode": "offline", "answer": "[{\"word\":\"ponies\",\"start\":4.09,\"end\":4.46}]"} {"utterance_id": "4515-11057-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0046.flac", "task_type": "lexical_detection", "query": "obtained", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"obtained\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.89, 3.49, "obtained"]], "response_mode": "offline", "answer": "[{\"word\":\"obtained\",\"start\":2.89,\"end\":3.49}]"} {"utterance_id": "8254-84205-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0066.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Do not answer until the audio ends. Then list each instance of an organization name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.6, 4.02, "redskins"]], "response_mode": "offline", "answer": "[{\"word\":\"redskins\",\"start\":3.6,\"end\":4.02}]"} {"utterance_id": "3660-172182-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0036.flac", "task_type": "lexical_counting", "query": "gifts", "prompt": "Process the full audio offline. For every occurrence of the word \"gifts\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.63, 4.14, "one"], [5.89, 6.29, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.63,\"end\":4.14},{\"word\":\"two\",\"start\":5.89,\"end\":6.29}]"} {"utterance_id": "6841-88291-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0015.flac", "task_type": "lexical_detection", "query": "swinging", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"swinging\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.61, 3.07, "swinging"]], "response_mode": "offline", "answer": "[{\"word\":\"swinging\",\"start\":2.61,\"end\":3.07}]"} {"utterance_id": "6841-88291-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0035.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of count events for an animal name (word, start, end). If none, reply [].", "labels": [[1.55, 1.87, "one"], [3.27, 3.57, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.55,\"end\":1.87},{\"word\":\"two\",\"start\":3.27,\"end\":3.57}]"} {"utterance_id": "700-122866-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0017.flac", "task_type": "lexical_detection", "query": "dreams", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"dreams\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.84, 4.49, "dreams"]], "response_mode": "offline", "answer": "[{\"word\":\"dreams\",\"start\":3.84,\"end\":4.49}]"} {"utterance_id": "6841-88291-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0032.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.16, 0.5, "one"], [1.69, 1.88, "few"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.16,\"end\":0.5},{\"word\":\"few\",\"start\":1.69,\"end\":1.88}]"} {"utterance_id": "7641-96252-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0019.flac", "task_type": "lexical_detection", "query": "therefore", "prompt": "When the utterance is complete, report all instances of the word \"therefore\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.89, "therefore"]], "response_mode": "offline", "answer": "[{\"word\":\"therefore\",\"start\":0.5,\"end\":0.89}]"} {"utterance_id": "1686-142278-0085", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0085.flac", "task_type": "lexical_detection", "query": "wanted", "prompt": "Process the full audio offline. Find every instance of the word \"wanted\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.24, 0.53, "wanted"]], "response_mode": "offline", "answer": "[{\"word\":\"wanted\",\"start\":0.24,\"end\":0.53}]"} {"utterance_id": "4831-25894-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 0.72, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":0.4,\"end\":0.72}]"} {"utterance_id": "700-122867-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.62, 3.09, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":2.62,\"end\":3.09}]"} {"utterance_id": "4323-18416-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0014.flac", "task_type": "lexical_detection", "query": "clasped", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"clasped\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.84, 8.2, "clasped"]], "response_mode": "offline", "answer": "[{\"word\":\"clasped\",\"start\":7.84,\"end\":8.2}]"} {"utterance_id": "4570-56594-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0007.flac", "task_type": "lexical_detection", "query": "accounts", "prompt": "After listening to the whole recording, output every instance of the word \"accounts\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 1.94, "accounts"]], "response_mode": "offline", "answer": "[{\"word\":\"accounts\",\"start\":1.3,\"end\":1.94}]"} {"utterance_id": "8254-115543-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0040.flac", "task_type": "lexical_detection", "query": "chola", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"chola\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.55, 0.86, "chola"]], "response_mode": "offline", "answer": "[{\"word\":\"chola\",\"start\":0.55,\"end\":0.86}]"} {"utterance_id": "4515-11057-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0028.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.97, 1.25, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":0.97,\"end\":1.25}]"} {"utterance_id": "6599-38591-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0002.flac", "task_type": "lexical_counting", "query": "fifteen", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"fifteen\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[2.61, 3.01, "one"], [4.69, 5.11, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.61,\"end\":3.01},{\"word\":\"two\",\"start\":4.69,\"end\":5.11}]"} {"utterance_id": "7641-96684-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0001.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.5, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":1.95,\"end\":2.5}]"} {"utterance_id": "8254-84205-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0007.flac", "task_type": "lexical_detection", "query": "leg", "prompt": "When the utterance is complete, report all instances of the word \"leg\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.81, "leg"]], "response_mode": "offline", "answer": "[{\"word\":\"leg\",\"start\":1.46,\"end\":1.81}]"} {"utterance_id": "7697-105815-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.91, 7.29, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":6.91,\"end\":7.29}]"} {"utterance_id": "1585-157660-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0003.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.14, 1.67, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":1.14,\"end\":1.67}]"} {"utterance_id": "8254-84205-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0017.flac", "task_type": "lexical_detection", "query": "griggs", "prompt": "After listening to the whole recording, output every instance of the word \"griggs\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.39, 3.69, "griggs"]], "response_mode": "offline", "answer": "[{\"word\":\"griggs\",\"start\":3.39,\"end\":3.69}]"} {"utterance_id": "3663-172528-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0051.flac", "task_type": "lexical_detection", "query": "credit", "prompt": "After listening to the whole recording, output every instance of the word \"credit\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.19, 8.62, "credit"]], "response_mode": "offline", "answer": "[{\"word\":\"credit\",\"start\":8.19,\"end\":8.62}]"} {"utterance_id": "700-122866-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0041.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.87, "missus lynde"]], "response_mode": "offline", "answer": "[{\"word\":\"missus lynde\",\"start\":0.21,\"end\":0.87}]"} {"utterance_id": "6123-59186-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0035.flac", "task_type": "lexical_detection", "query": "defeated", "prompt": "When the utterance is complete, report all instances of the word \"defeated\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.74, "defeated"]], "response_mode": "offline", "answer": "[{\"word\":\"defeated\",\"start\":3.21,\"end\":3.74}]"} {"utterance_id": "700-122867-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 2.65, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":2.34,\"end\":2.65}]"} {"utterance_id": "4572-64670-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0008.flac", "task_type": "lexical_detection", "query": "harbour", "prompt": "Process the full audio offline. Find every instance of the word \"harbour\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.84, "harbour"]], "response_mode": "offline", "answer": "[{\"word\":\"harbour\",\"start\":0.47,\"end\":0.84}]"} {"utterance_id": "4323-55228-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0026.flac", "task_type": "lexical_detection", "query": "slow", "prompt": "Process the full audio offline. Find every instance of the word \"slow\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.81, 1.02, "slow"]], "response_mode": "offline", "answer": "[{\"word\":\"slow\",\"start\":0.81,\"end\":1.02}]"} {"utterance_id": "8288-274162-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0010.flac", "task_type": "lexical_detection", "query": "something", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"something\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 2.05, "something"]], "response_mode": "offline", "answer": "[{\"word\":\"something\",\"start\":1.42,\"end\":2.05}]"} {"utterance_id": "1701-141759-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 1.2, "tongue"]], "response_mode": "offline", "answer": "[{\"word\":\"tongue\",\"start\":0.88,\"end\":1.2}]"} {"utterance_id": "6467-94831-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0029.flac", "task_type": "lexical_detection", "query": "collected", "prompt": "Process the full audio offline. Find every instance of the word \"collected\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.29, "collected"]], "response_mode": "offline", "answer": "[{\"word\":\"collected\",\"start\":1.84,\"end\":2.29}]"} {"utterance_id": "4515-11057-0059", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0059.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.09, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.79,\"end\":1.09}]"} {"utterance_id": "1630-102884-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0010.flac", "task_type": "lexical_detection", "query": "organic", "prompt": "Process the full audio offline. Find every instance of the word \"organic\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.3, "organic"]], "response_mode": "offline", "answer": "[{\"word\":\"organic\",\"start\":0.67,\"end\":1.3}]"} {"utterance_id": "7601-175351-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0010.flac", "task_type": "lexical_detection", "query": "cabin", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"cabin\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.97, 6.52, "cabin"]], "response_mode": "offline", "answer": "[{\"word\":\"cabin\",\"start\":5.97,\"end\":6.52}]"} {"utterance_id": "6267-65525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0013.flac", "task_type": "lexical_detection", "query": "ellen", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"ellen\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.55, 2.85, "ellen"]], "response_mode": "offline", "answer": "[{\"word\":\"ellen\",\"start\":2.55,\"end\":2.85}]"} {"utterance_id": "1630-141772-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.37, 2.97, "napoleon"]], "response_mode": "offline", "answer": "[{\"word\":\"napoleon\",\"start\":2.37,\"end\":2.97}]"} {"utterance_id": "5849-50873-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0036.flac", "task_type": "lexical_detection", "query": "youthful", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"youthful\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.4, "youthful"]], "response_mode": "offline", "answer": "[{\"word\":\"youthful\",\"start\":1.98,\"end\":2.4}]"} {"utterance_id": "5543-27761-0104", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0104.flac", "task_type": "lexical_detection", "query": "little", "prompt": "After listening to the whole recording, output every instance of the word \"little\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.61, 0.86, "little"]], "response_mode": "offline", "answer": "[{\"word\":\"little\",\"start\":0.61,\"end\":0.86}]"} {"utterance_id": "4572-112381-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.01, 4.68, "sugriva"]], "response_mode": "offline", "answer": "[{\"word\":\"sugriva\",\"start\":4.01,\"end\":4.68}]"} {"utterance_id": "4515-11057-0091", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0091.flac", "task_type": "lexical_detection", "query": "loved", "prompt": "After the audio ends, extract all occurrences of the word \"loved\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.14, "loved"]], "response_mode": "offline", "answer": "[{\"word\":\"loved\",\"start\":1.77,\"end\":2.14}]"} {"utterance_id": "1650-167613-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0037.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.74, 4.08, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":3.74,\"end\":4.08}]"} {"utterance_id": "1255-90407-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0016.flac", "task_type": "lexical_detection", "query": "lodging", "prompt": "Process the full audio offline. Find every instance of the word \"lodging\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.96, 2.27, "lodging"]], "response_mode": "offline", "answer": "[{\"word\":\"lodging\",\"start\":1.96,\"end\":2.27}]"} {"utterance_id": "7697-245712-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0018.flac", "task_type": "lexical_counting", "query": "man", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"man\" (word, start, end). If none, reply [].", "labels": [[2.06, 2.34, "one"], [3.56, 3.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.06,\"end\":2.34},{\"word\":\"two\",\"start\":3.56,\"end\":3.92}]"} {"utterance_id": "8288-274162-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0043.flac", "task_type": "semantic_detection", "query": "money", "prompt": "You will hear a complete utterance. When it ends, list every instance of a money or currency term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.72, 7.2, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":6.72,\"end\":7.2}]"} {"utterance_id": "3660-172182-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0026.flac", "task_type": "lexical_counting", "query": "maiden", "prompt": "Listen to the entire clip, then report the count sequence for the word \"maiden\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.76, "one"], [6.43, 6.83, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.47,\"end\":0.76},{\"word\":\"two\",\"start\":6.43,\"end\":6.83}]"} {"utterance_id": "7641-96252-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0017.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.55, 8.05, "priests"]], "response_mode": "offline", "answer": "[{\"word\":\"priests\",\"start\":7.55,\"end\":8.05}]"} {"utterance_id": "4515-11057-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.05, "amy"]], "response_mode": "offline", "answer": "[{\"word\":\"amy\",\"start\":0.72,\"end\":1.05}]"} {"utterance_id": "2506-11278-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0010.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.01, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":1.77,\"end\":2.01}]"} {"utterance_id": "3660-172182-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0030.flac", "task_type": "semantic_detection", "query": "state_or_province", "prompt": "Listen to the full audio. After it finishes, find every instance of a state or province name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.85, 5.64, "cornwall"]], "response_mode": "offline", "answer": "[{\"word\":\"cornwall\",\"start\":4.85,\"end\":5.64}]"} {"utterance_id": "6267-53049-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0011.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "When the utterance is complete, report how a person's name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 1.15, "one"], [11.32, 12.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.63,\"end\":1.15},{\"word\":\"two\",\"start\":11.32,\"end\":12.66}]"} {"utterance_id": "1686-142278-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0007.flac", "task_type": "lexical_detection", "query": "preparing", "prompt": "When the utterance is complete, report all instances of the word \"preparing\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.65, "preparing"]], "response_mode": "offline", "answer": "[{\"word\":\"preparing\",\"start\":1.17,\"end\":1.65}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "lexical_counting", "query": "year", "prompt": "When the utterance is complete, report how the word \"year\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.32, "one"], [5.47, 5.73, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.0,\"end\":1.32},{\"word\":\"two\",\"start\":5.47,\"end\":5.73}]"} {"utterance_id": "1650-167613-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0011.flac", "task_type": "lexical_detection", "query": "smolensk", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"smolensk\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.53, 1.42, "smolensk"]], "response_mode": "offline", "answer": "[{\"word\":\"smolensk\",\"start\":0.53,\"end\":1.42}]"} {"utterance_id": "4572-64670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0024.flac", "task_type": "semantic_counting", "query": "plant", "prompt": "After listening to the whole recording, output timed count words for a plant name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.02, 1.36, "one"], [3.36, 3.85, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.02,\"end\":1.36},{\"word\":\"two\",\"start\":3.36,\"end\":3.85}]"} {"utterance_id": "4515-11057-0085", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0085.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.85, 3.37, "frame"]], "response_mode": "offline", "answer": "[{\"word\":\"frame\",\"start\":2.85,\"end\":3.37}]"} {"utterance_id": "1255-74899-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0020.flac", "task_type": "lexical_detection", "query": "worsted", "prompt": "Process the full audio offline. Find every instance of the word \"worsted\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.89, 2.31, "worsted"]], "response_mode": "offline", "answer": "[{\"word\":\"worsted\",\"start\":1.89,\"end\":2.31}]"} {"utterance_id": "1701-141760-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0020.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.01, 6.53, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"voice\",\"start\":6.01,\"end\":6.53}]"} {"utterance_id": "1255-90407-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0024.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.38, "servant"]], "response_mode": "offline", "answer": "[{\"word\":\"servant\",\"start\":0.94,\"end\":1.38}]"} {"utterance_id": "4831-25894-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0017.flac", "task_type": "lexical_detection", "query": "gay", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"gay\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.7, "gay"]], "response_mode": "offline", "answer": "[{\"word\":\"gay\",\"start\":0.49,\"end\":0.7}]"} {"utterance_id": "1585-131718-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0046.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Once the audio has finished, identify every instance of a city name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.5, 3.05, "new york"]], "response_mode": "offline", "answer": "[{\"word\":\"new york\",\"start\":2.5,\"end\":3.05}]"} {"utterance_id": "1701-141760-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0028.flac", "task_type": "lexical_detection", "query": "intently", "prompt": "After listening to the whole recording, output every instance of the word \"intently\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.41, "intently"]], "response_mode": "offline", "answer": "[{\"word\":\"intently\",\"start\":0.82,\"end\":1.41}]"} {"utterance_id": "3660-172183-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0016.flac", "task_type": "lexical_detection", "query": "attempt", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"attempt\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.3, 2.61, "attempt"]], "response_mode": "offline", "answer": "[{\"word\":\"attempt\",\"start\":2.3,\"end\":2.61}]"} {"utterance_id": "1585-131718-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0024.flac", "task_type": "lexical_detection", "query": "cents", "prompt": "Once the audio has finished, identify every instance of the word \"cents\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.95, 3.35, "cents"]], "response_mode": "offline", "answer": "[{\"word\":\"cents\",\"start\":2.95,\"end\":3.35}]"} {"utterance_id": "3663-172528-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0040.flac", "task_type": "lexical_detection", "query": "vengeance", "prompt": "Process the full audio offline. Find every instance of the word \"vengeance\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.43, 9.12, "vengeance"]], "response_mode": "offline", "answer": "[{\"word\":\"vengeance\",\"start\":8.43,\"end\":9.12}]"} {"utterance_id": "7601-291468-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0003.flac", "task_type": "lexical_detection", "query": "flies", "prompt": "Process the full audio offline. Find every instance of the word \"flies\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.64, "flies"], [14.22, 14.92, "flies"]], "response_mode": "offline", "answer": "[{\"word\":\"flies\",\"start\":1.88,\"end\":2.64},{\"word\":\"flies\",\"start\":14.22,\"end\":14.92}]"} {"utterance_id": "4323-55228-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0039.flac", "task_type": "lexical_detection", "query": "wonder", "prompt": "When the utterance is complete, report all instances of the word \"wonder\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.28, 0.56, "wonder"]], "response_mode": "offline", "answer": "[{\"word\":\"wonder\",\"start\":0.28,\"end\":0.56}]"} {"utterance_id": "7601-101622-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0003.flac", "task_type": "lexical_detection", "query": "art", "prompt": "Once the audio has finished, identify every instance of the word \"art\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.55, 2.79, "art"]], "response_mode": "offline", "answer": "[{\"word\":\"art\",\"start\":2.55,\"end\":2.79}]"} {"utterance_id": "4153-186223-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.33, 1.45, "john merrick"]], "response_mode": "offline", "answer": "[{\"word\":\"john merrick\",\"start\":0.33,\"end\":1.45}]"} {"utterance_id": "8254-115543-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0024.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "When the utterance is complete, report how a profession accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.72, 2.06, "one"], [4.68, 5.09, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.72,\"end\":2.06},{\"word\":\"two\",\"start\":4.68,\"end\":5.09}]"} {"utterance_id": "1630-96099-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0019.flac", "task_type": "semantic_detection", "query": "country", "prompt": "When the utterance is complete, report all instances of a country name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.73, "italian"]], "response_mode": "offline", "answer": "[{\"word\":\"italian\",\"start\":1.11,\"end\":1.73}]"} {"utterance_id": "6455-66379-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0016.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.32, 0.73, "pilots"]], "response_mode": "offline", "answer": "[{\"word\":\"pilots\",\"start\":0.32,\"end\":0.73}]"} {"utterance_id": "6123-59186-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0020.flac", "task_type": "lexical_detection", "query": "grieve", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"grieve\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.73, 2.24, "grieve"]], "response_mode": "offline", "answer": "[{\"word\":\"grieve\",\"start\":1.73,\"end\":2.24}]"} {"utterance_id": "6455-67804-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0011.flac", "task_type": "lexical_detection", "query": "turning", "prompt": "Once the audio has finished, identify every instance of the word \"turning\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.28, 0.61, "turning"]], "response_mode": "offline", "answer": "[{\"word\":\"turning\",\"start\":0.28,\"end\":0.61}]"} {"utterance_id": "6467-97061-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0005.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After listening to the whole recording, output every instance of an organization name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.75, "army"]], "response_mode": "offline", "answer": "[{\"word\":\"army\",\"start\":2.27,\"end\":2.75}]"} {"utterance_id": "3915-57461-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0025.flac", "task_type": "lexical_detection", "query": "cheer", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"cheer\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.11, "cheer"]], "response_mode": "offline", "answer": "[{\"word\":\"cheer\",\"start\":0.77,\"end\":1.11}]"} {"utterance_id": "4153-186222-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0027.flac", "task_type": "lexical_detection", "query": "charming", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"charming\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.64, "charming"]], "response_mode": "offline", "answer": "[{\"word\":\"charming\",\"start\":1.25,\"end\":1.64}]"} {"utterance_id": "6841-88291-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0022.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.18, 2.56, "flank"]], "response_mode": "offline", "answer": "[{\"word\":\"flank\",\"start\":2.18,\"end\":2.56}]"} {"utterance_id": "116-288045-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0028.flac", "task_type": "lexical_detection", "query": "speak", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"speak\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 0.97, "speak"]], "response_mode": "offline", "answer": "[{\"word\":\"speak\",\"start\":0.63,\"end\":0.97}]"} {"utterance_id": "700-122868-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0038.flac", "task_type": "lexical_detection", "query": "sense", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"sense\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.5, 1.91, "sense"]], "response_mode": "offline", "answer": "[{\"word\":\"sense\",\"start\":1.5,\"end\":1.91}]"} {"utterance_id": "6841-88291-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0038.flac", "task_type": "lexical_detection", "query": "fails", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fails\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.6, "fails"]], "response_mode": "offline", "answer": "[{\"word\":\"fails\",\"start\":2.02,\"end\":2.6}]"} {"utterance_id": "6841-88291-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0035.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Do not answer until the audio ends. Then list each instance of an animal name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.55, 1.87, "calf"], [3.27, 3.57, "beast"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":1.55,\"end\":1.87},{\"word\":\"beast\",\"start\":3.27,\"end\":3.57}]"} {"utterance_id": "8254-115543-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0042.flac", "task_type": "lexical_detection", "query": "bundle", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"bundle\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.29, 3.55, "bundle"]], "response_mode": "offline", "answer": "[{\"word\":\"bundle\",\"start\":3.29,\"end\":3.55}]"} {"utterance_id": "116-288047-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0010.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.23, 4.69, "spirits fairies gods"]], "response_mode": "offline", "answer": "[{\"word\":\"spirits fairies gods\",\"start\":2.23,\"end\":4.69}]"} {"utterance_id": "4323-13259-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0019.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.36, 4.72, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":4.36,\"end\":4.72}]"} {"utterance_id": "116-288045-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0017.flac", "task_type": "lexical_detection", "query": "homer", "prompt": "After listening to the whole recording, output every instance of the word \"homer\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.03, 1.57, "homer"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":1.03,\"end\":1.57}]"} {"utterance_id": "7697-105815-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0024.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[4.81, 5.01, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":4.81,\"end\":5.01}]"} {"utterance_id": "7697-105815-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0039.flac", "task_type": "lexical_detection", "query": "dwelt", "prompt": "Once the audio has finished, identify every instance of the word \"dwelt\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[13.47, 13.99, "dwelt"]], "response_mode": "offline", "answer": "[{\"word\":\"dwelt\",\"start\":13.47,\"end\":13.99}]"} {"utterance_id": "1686-142278-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.5, 0.79, "god"]], "response_mode": "offline", "answer": "[{\"word\":\"god\",\"start\":0.5,\"end\":0.79}]"} {"utterance_id": "3660-172182-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.62, 1.92, "arthur"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur\",\"start\":1.62,\"end\":1.92}]"} {"utterance_id": "4515-11057-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0038.flac", "task_type": "lexical_detection", "query": "judge", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"judge\" with fields word, start, end (seconds). If none, reply [].", "labels": [[6.91, 7.27, "judge"]], "response_mode": "offline", "answer": "[{\"word\":\"judge\",\"start\":6.91,\"end\":7.27}]"} {"utterance_id": "6455-67803-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0013.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.99, 1.09, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.99,\"end\":1.09}]"} {"utterance_id": "4570-14911-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0007.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an organization name with fields word, start, end (seconds). If none, reply [].", "labels": [[7.93, 8.59, "association"]], "response_mode": "offline", "answer": "[{\"word\":\"association\",\"start\":7.93,\"end\":8.59}]"} {"utterance_id": "4570-14911-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0001.flac", "task_type": "lexical_detection", "query": "sixteen", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"sixteen\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.44, 4.02, "sixteen"]], "response_mode": "offline", "answer": "[{\"word\":\"sixteen\",\"start\":3.44,\"end\":4.02}]"} {"utterance_id": "6267-53049-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0021.flac", "task_type": "lexical_detection", "query": "altogether", "prompt": "When the utterance is complete, report all instances of the word \"altogether\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 2.46, "altogether"]], "response_mode": "offline", "answer": "[{\"word\":\"altogether\",\"start\":1.58,\"end\":2.46}]"} {"utterance_id": "1585-131718-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0031.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.96, 8.28, "fifth"], [9.68, 10.06, "fourteenth"]], "response_mode": "offline", "answer": "[{\"word\":\"fifth\",\"start\":7.96,\"end\":8.28},{\"word\":\"fourteenth\",\"start\":9.68,\"end\":10.06}]"} {"utterance_id": "7641-96684-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 2.04, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":1.5,\"end\":2.04}]"} {"utterance_id": "1701-141760-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0017.flac", "task_type": "lexical_detection", "query": "guards", "prompt": "After the audio ends, extract all occurrences of the word \"guards\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.54, "guards"]], "response_mode": "offline", "answer": "[{\"word\":\"guards\",\"start\":1.12,\"end\":1.54}]"} {"utterance_id": "2506-11278-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0013.flac", "task_type": "lexical_detection", "query": "greedy", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"greedy\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.3, 4.79, "greedy"]], "response_mode": "offline", "answer": "[{\"word\":\"greedy\",\"start\":4.3,\"end\":4.79}]"} {"utterance_id": "1701-141760-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0036.flac", "task_type": "lexical_detection", "query": "smoke", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"smoke\" with fields word, start, end (seconds). If none, reply [].", "labels": [[4.02, 4.53, "smoke"]], "response_mode": "offline", "answer": "[{\"word\":\"smoke\",\"start\":4.02,\"end\":4.53}]"} {"utterance_id": "7641-96670-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0008.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "After listening to the whole recording, output timed count words for an animal name (one, two, three, ...) as JSON. If none, reply [].", "labels": [[0.81, 1.1, "one"], [3.53, 3.95, "two"], [5.21, 5.59, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.81,\"end\":1.1},{\"word\":\"two\",\"start\":3.53,\"end\":3.95},{\"word\":\"three\",\"start\":5.21,\"end\":5.59}]"} {"utterance_id": "4831-18525-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0008.flac", "task_type": "lexical_detection", "query": "fun", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"fun\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.11, "fun"]], "response_mode": "offline", "answer": "[{\"word\":\"fun\",\"start\":1.82,\"end\":2.11}]"} {"utterance_id": "3660-172182-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0000.flac", "task_type": "lexical_detection", "query": "arthur", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"arthur\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.11, 1.5, "arthur"]], "response_mode": "offline", "answer": "[{\"word\":\"arthur\",\"start\":1.11,\"end\":1.5}]"} {"utterance_id": "1585-131718-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0002.flac", "task_type": "lexical_detection", "query": "raise", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"raise\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.03, 1.23, "raise"]], "response_mode": "offline", "answer": "[{\"word\":\"raise\",\"start\":1.03,\"end\":1.23}]"} {"utterance_id": "5849-50873-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0040.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.0, 13.52, "governor"]], "response_mode": "offline", "answer": "[{\"word\":\"governor\",\"start\":13.0,\"end\":13.52}]"} {"utterance_id": "7601-175351-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0010.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 4.07, "hatteras"]], "response_mode": "offline", "answer": "[{\"word\":\"hatteras\",\"start\":3.36,\"end\":4.07}]"} {"utterance_id": "5849-50873-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0035.flac", "task_type": "lexical_counting", "query": "live", "prompt": "After the audio ends, tally every instance of the word \"live\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.84, 1.97, "one"], [4.21, 4.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.84,\"end\":1.97},{\"word\":\"two\",\"start\":4.21,\"end\":4.56}]"} {"utterance_id": "2506-11278-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.23, 5.78, "journalist"]], "response_mode": "offline", "answer": "[{\"word\":\"journalist\",\"start\":5.23,\"end\":5.78}]"} {"utterance_id": "6841-88294-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0047.flac", "task_type": "lexical_detection", "query": "want", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"want\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.66, "want"]], "response_mode": "offline", "answer": "[{\"word\":\"want\",\"start\":0.49,\"end\":0.66}]"} {"utterance_id": "5849-50962-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0010.flac", "task_type": "lexical_detection", "query": "sorts", "prompt": "After the audio ends, extract all occurrences of the word \"sorts\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 2.31, "sorts"]], "response_mode": "offline", "answer": "[{\"word\":\"sorts\",\"start\":1.95,\"end\":2.31}]"} {"utterance_id": "7697-105815-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0050.flac", "task_type": "lexical_detection", "query": "break", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"break\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.21, 9.57, "break"]], "response_mode": "offline", "answer": "[{\"word\":\"break\",\"start\":9.21,\"end\":9.57}]"} {"utterance_id": "1650-167613-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0000.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen carefully to the entire clip, then report every instance of a food name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.92, 9.35, "tea"]], "response_mode": "offline", "answer": "[{\"word\":\"tea\",\"start\":8.92,\"end\":9.35}]"} {"utterance_id": "5543-27761-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0019.flac", "task_type": "lexical_detection", "query": "mother", "prompt": "After listening to the whole recording, output every instance of the word \"mother\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.94, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":0.68,\"end\":0.94}]"} {"utterance_id": "8254-84205-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0046.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.88, 2.31, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":1.88,\"end\":2.31}]"} {"utterance_id": "1686-142278-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0027.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.33, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.07,\"end\":2.33}]"} {"utterance_id": "1650-167613-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0027.flac", "task_type": "lexical_detection", "query": "awaits", "prompt": "When the utterance is complete, report all instances of the word \"awaits\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.33, "awaits"]], "response_mode": "offline", "answer": "[{\"word\":\"awaits\",\"start\":0.83,\"end\":1.33}]"} {"utterance_id": "1686-142278-0061", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0061.flac", "task_type": "lexical_detection", "query": "gratitude", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"gratitude\" with fields word, start, end (seconds). If none, reply [].", "labels": [[5.06, 5.67, "gratitude"]], "response_mode": "offline", "answer": "[{\"word\":\"gratitude\",\"start\":5.06,\"end\":5.67}]"} {"utterance_id": "2506-11278-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0023.flac", "task_type": "lexical_detection", "query": "sorry", "prompt": "After the audio ends, extract all occurrences of the word \"sorry\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 0.95, "sorry"]], "response_mode": "offline", "answer": "[{\"word\":\"sorry\",\"start\":0.67,\"end\":0.95}]"} {"utterance_id": "4572-112381-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0013.flac", "task_type": "lexical_detection", "query": "vain", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"vain\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.09, 6.81, "vain"]], "response_mode": "offline", "answer": "[{\"word\":\"vain\",\"start\":6.09,\"end\":6.81}]"} {"utterance_id": "8288-274162-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0012.flac", "task_type": "lexical_detection", "query": "father", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"father\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 2.04, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.48,\"end\":2.04}]"} {"utterance_id": "8288-274162-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0030.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "You will hear a complete utterance. When it ends, list every instance of a number you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.31, "hundred"]], "response_mode": "offline", "answer": "[{\"word\":\"hundred\",\"start\":2.93,\"end\":3.31}]"} {"utterance_id": "6455-67804-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0014.flac", "task_type": "lexical_detection", "query": "per", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"per\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.79, 1.0, "per"]], "response_mode": "offline", "answer": "[{\"word\":\"per\",\"start\":0.79,\"end\":1.0}]"} {"utterance_id": "3663-172528-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0014.flac", "task_type": "lexical_detection", "query": "answered", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"answered\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.56, 1.03, "answered"]], "response_mode": "offline", "answer": "[{\"word\":\"answered\",\"start\":0.56,\"end\":1.03}]"} {"utterance_id": "6455-67804-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0010.flac", "task_type": "lexical_detection", "query": "beg", "prompt": "Once the audio has finished, identify every instance of the word \"beg\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.64, 0.97, "beg"]], "response_mode": "offline", "answer": "[{\"word\":\"beg\",\"start\":0.64,\"end\":0.97}]"} {"utterance_id": "8254-84205-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.43, "forty nine"]], "response_mode": "offline", "answer": "[{\"word\":\"forty nine\",\"start\":0.7,\"end\":1.43}]"} {"utterance_id": "1255-138279-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0012.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 2.06, "snow"]], "response_mode": "offline", "answer": "[{\"word\":\"snow\",\"start\":1.51,\"end\":2.06}]"} {"utterance_id": "6455-67803-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0033.flac", "task_type": "lexical_detection", "query": "promise", "prompt": "After the audio ends, extract all occurrences of the word \"promise\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.93, "promise"]], "response_mode": "offline", "answer": "[{\"word\":\"promise\",\"start\":2.39,\"end\":2.93}]"} {"utterance_id": "8173-294714-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0033.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.34, 2.85, "doctor's"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor's\",\"start\":2.34,\"end\":2.85}]"} {"utterance_id": "6123-59186-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0033.flac", "task_type": "lexical_counting", "query": "poems", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"poems\" with word/start/end. If none, reply [].", "labels": [[2.85, 3.24, "one"], [7.15, 7.53, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.85,\"end\":3.24},{\"word\":\"two\",\"start\":7.15,\"end\":7.53}]"} {"utterance_id": "5849-50964-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0003.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a person's name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.3, 0.63, "one"], [2.76, 3.38, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.3,\"end\":0.63},{\"word\":\"two\",\"start\":2.76,\"end\":3.38}]"} {"utterance_id": "1701-141759-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.98, 2.38, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":1.98,\"end\":2.38}]"} {"utterance_id": "8173-294714-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0007.flac", "task_type": "lexical_detection", "query": "speak", "prompt": "Once the audio has finished, identify every instance of the word \"speak\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.76, 1.11, "speak"]], "response_mode": "offline", "answer": "[{\"word\":\"speak\",\"start\":0.76,\"end\":1.11}]"} {"utterance_id": "3660-172182-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0036.flac", "task_type": "lexical_detection", "query": "giving", "prompt": "After the audio ends, extract all occurrences of the word \"giving\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.22, 1.62, "giving"]], "response_mode": "offline", "answer": "[{\"word\":\"giving\",\"start\":1.22,\"end\":1.62}]"} {"utterance_id": "1686-142278-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0023.flac", "task_type": "lexical_detection", "query": "miserable", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"miserable\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.06, "miserable"]], "response_mode": "offline", "answer": "[{\"word\":\"miserable\",\"start\":2.61,\"end\":3.06}]"} {"utterance_id": "7697-245712-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.64, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.14,\"end\":1.64}]"} {"utterance_id": "1650-167613-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0021.flac", "task_type": "lexical_detection", "query": "equipment", "prompt": "When the utterance is complete, report all instances of the word \"equipment\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.22, 5.89, "equipment"]], "response_mode": "offline", "answer": "[{\"word\":\"equipment\",\"start\":5.22,\"end\":5.89}]"} {"utterance_id": "6455-67804-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0000.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.99, 1.53, "granddaughter"]], "response_mode": "offline", "answer": "[{\"word\":\"granddaughter\",\"start\":0.99,\"end\":1.53}]"} {"utterance_id": "6467-94831-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0000.flac", "task_type": "semantic_counting", "query": "city", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a city name with word/start/end. If none, reply [].", "labels": [[6.61, 7.33, "one"], [9.39, 10.11, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.61,\"end\":7.33},{\"word\":\"two\",\"start\":9.39,\"end\":10.11}]"} {"utterance_id": "1701-141760-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.56, 2.91, "nicholas"]], "response_mode": "offline", "answer": "[{\"word\":\"nicholas\",\"start\":2.56,\"end\":2.91}]"} {"utterance_id": "700-122866-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.97, "anne's"], [6.95, 7.49, "miss stacy"]], "response_mode": "offline", "answer": "[{\"word\":\"anne's\",\"start\":0.68,\"end\":0.97},{\"word\":\"miss stacy\",\"start\":6.95,\"end\":7.49}]"} {"utterance_id": "8173-294714-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0013.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Once the audio has finished, identify every instance of an ordinal number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.74, 5.12, "second"]], "response_mode": "offline", "answer": "[{\"word\":\"second\",\"start\":4.74,\"end\":5.12}]"} {"utterance_id": "4515-11057-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0065.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.71, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":1.18,\"end\":1.71}]"} {"utterance_id": "6267-65525-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0000.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.78, 4.11, "aunt"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt\",\"start\":3.78,\"end\":4.11}]"} {"utterance_id": "2506-13150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.09, 11.84, "brain"]], "response_mode": "offline", "answer": "[{\"word\":\"brain\",\"start\":11.09,\"end\":11.84}]"} {"utterance_id": "6467-56885-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0017.flac", "task_type": "lexical_detection", "query": "intense", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"intense\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.9, 1.32, "intense"]], "response_mode": "offline", "answer": "[{\"word\":\"intense\",\"start\":0.9,\"end\":1.32}]"} {"utterance_id": "4515-11057-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0021.flac", "task_type": "lexical_detection", "query": "especially", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"especially\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.93, "especially"]], "response_mode": "offline", "answer": "[{\"word\":\"especially\",\"start\":0.34,\"end\":0.93}]"} {"utterance_id": "4515-11057-0114", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0114.flac", "task_type": "lexical_counting", "query": "tell", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"tell\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[6.13, 6.39, "one"], [10.08, 10.28, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.13,\"end\":6.39},{\"word\":\"two\",\"start\":10.08,\"end\":10.28}]"} {"utterance_id": "6123-59150-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0011.flac", "task_type": "lexical_detection", "query": "deeds", "prompt": "Once the audio has finished, identify every instance of the word \"deeds\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.45, 1.81, "deeds"]], "response_mode": "offline", "answer": "[{\"word\":\"deeds\",\"start\":1.45,\"end\":1.81}]"} {"utterance_id": "6455-66379-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0008.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a plant name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.46, 0.68, "wood"]], "response_mode": "offline", "answer": "[{\"word\":\"wood\",\"start\":0.46,\"end\":0.68}]"} {"utterance_id": "4515-11057-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0049.flac", "task_type": "lexical_detection", "query": "jasper", "prompt": "After listening to the whole recording, output every instance of the word \"jasper\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.19, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":2.7,\"end\":3.19}]"} {"utterance_id": "8254-84205-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0037.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.33, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":0.99,\"end\":1.33}]"} {"utterance_id": "8173-294714-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0047.flac", "task_type": "lexical_detection", "query": "proposal", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"proposal\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.18, 6.71, "proposal"]], "response_mode": "offline", "answer": "[{\"word\":\"proposal\",\"start\":6.18,\"end\":6.71}]"} {"utterance_id": "700-122866-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0004.flac", "task_type": "lexical_detection", "query": "awake", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"awake\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.11, "awake"]], "response_mode": "offline", "answer": "[{\"word\":\"awake\",\"start\":0.79,\"end\":1.11}]"} {"utterance_id": "5849-50962-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0024.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.63, 0.9, "one"], [2.27, 2.66, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.63,\"end\":0.9},{\"word\":\"two\",\"start\":2.27,\"end\":2.66}]"} {"utterance_id": "7601-291468-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0003.flac", "task_type": "lexical_counting", "query": "spiders", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"spiders\" (word, start, end). If none, reply [].", "labels": [[8.74, 9.49, "one"], [10.49, 11.19, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":8.74,\"end\":9.49},{\"word\":\"two\",\"start\":10.49,\"end\":11.19}]"} {"utterance_id": "1255-90413-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0009.flac", "task_type": "lexical_detection", "query": "commonplace", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"commonplace\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[13.18, 13.89, "commonplace"]], "response_mode": "offline", "answer": "[{\"word\":\"commonplace\",\"start\":13.18,\"end\":13.89}]"} {"utterance_id": "5849-50964-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0013.flac", "task_type": "lexical_detection", "query": "spent", "prompt": "Process the full audio offline. Find every instance of the word \"spent\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.9, 2.29, "spent"]], "response_mode": "offline", "answer": "[{\"word\":\"spent\",\"start\":1.9,\"end\":2.29}]"} {"utterance_id": "700-122868-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.29, 2.76, "hundreds"]], "response_mode": "offline", "answer": "[{\"word\":\"hundreds\",\"start\":2.29,\"end\":2.76}]"} {"utterance_id": "700-122866-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.21, 2.86, "nose"]], "response_mode": "offline", "answer": "[{\"word\":\"nose\",\"start\":2.21,\"end\":2.86}]"} {"utterance_id": "4153-186223-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0018.flac", "task_type": "lexical_detection", "query": "man", "prompt": "After the audio ends, extract all occurrences of the word \"man\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 1.98, "man"]], "response_mode": "offline", "answer": "[{\"word\":\"man\",\"start\":1.68,\"end\":1.98}]"} {"utterance_id": "1255-90413-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0027.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.08, 2.56, "wife"]], "response_mode": "offline", "answer": "[{\"word\":\"wife\",\"start\":2.08,\"end\":2.56}]"} {"utterance_id": "6455-67804-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0030.flac", "task_type": "lexical_detection", "query": "surprised", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"surprised\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.57, 3.04, "surprised"]], "response_mode": "offline", "answer": "[{\"word\":\"surprised\",\"start\":2.57,\"end\":3.04}]"} {"utterance_id": "6455-67803-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0036.flac", "task_type": "lexical_detection", "query": "way", "prompt": "Once the audio has finished, identify every instance of the word \"way\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.04, 3.34, "way"]], "response_mode": "offline", "answer": "[{\"word\":\"way\",\"start\":3.04,\"end\":3.34}]"} {"utterance_id": "1686-142278-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0009.flac", "task_type": "lexical_detection", "query": "serious", "prompt": "When the utterance is complete, report all instances of the word \"serious\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.53, "serious"]], "response_mode": "offline", "answer": "[{\"word\":\"serious\",\"start\":2.1,\"end\":2.53}]"} {"utterance_id": "7697-105817-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0004.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Once the audio has finished, identify every instance of a plant name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.62, 4.1, "grass"]], "response_mode": "offline", "answer": "[{\"word\":\"grass\",\"start\":3.62,\"end\":4.1}]"} {"utterance_id": "1255-90413-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0004.flac", "task_type": "lexical_detection", "query": "able", "prompt": "After the audio ends, extract all occurrences of the word \"able\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.98, 1.44, "able"]], "response_mode": "offline", "answer": "[{\"word\":\"able\",\"start\":0.98,\"end\":1.44}]"} {"utterance_id": "700-122867-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0001.flac", "task_type": "lexical_detection", "query": "aged", "prompt": "Once the audio has finished, identify every instance of the word \"aged\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.91, 4.07, "aged"]], "response_mode": "offline", "answer": "[{\"word\":\"aged\",\"start\":3.91,\"end\":4.07}]"} {"utterance_id": "700-122868-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0005.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.6, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":3.33,\"end\":3.6}]"} {"utterance_id": "7641-96684-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.75, 4.17, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.75,\"end\":4.17}]"} {"utterance_id": "5849-50964-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0006.flac", "task_type": "lexical_detection", "query": "nellie", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"nellie\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.15, 2.44, "nellie"]], "response_mode": "offline", "answer": "[{\"word\":\"nellie\",\"start\":2.15,\"end\":2.44}]"} {"utterance_id": "4570-24733-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0001.flac", "task_type": "lexical_counting", "query": "attribute", "prompt": "Listen to the entire clip, then report the count sequence for the word \"attribute\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.83, 11.38, "one"], [15.82, 16.42, "two"], [29.33, 30.33, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":10.83,\"end\":11.38},{\"word\":\"two\",\"start\":15.82,\"end\":16.42},{\"word\":\"three\",\"start\":29.33,\"end\":30.33}]"} {"utterance_id": "5849-50873-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0040.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.45, 8.73, "five"]], "response_mode": "offline", "answer": "[{\"word\":\"five\",\"start\":8.45,\"end\":8.73}]"} {"utterance_id": "1255-90413-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0023.flac", "task_type": "lexical_detection", "query": "arrival", "prompt": "After the audio ends, extract all occurrences of the word \"arrival\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.43, 1.89, "arrival"]], "response_mode": "offline", "answer": "[{\"word\":\"arrival\",\"start\":1.43,\"end\":1.89}]"} {"utterance_id": "6123-59150-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0041.flac", "task_type": "lexical_detection", "query": "novels", "prompt": "Once the audio has finished, identify every instance of the word \"novels\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.59, 0.96, "novels"]], "response_mode": "offline", "answer": "[{\"word\":\"novels\",\"start\":0.59,\"end\":0.96}]"} {"utterance_id": "6267-65525-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0014.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a number (word, start, end). If none, reply [].", "labels": [[5.77, 6.35, "one"], [7.74, 8.31, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.77,\"end\":6.35},{\"word\":\"two\",\"start\":7.74,\"end\":8.31}]"} {"utterance_id": "4153-61735-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0002.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[13.15, 13.54, "horse"]], "response_mode": "offline", "answer": "[{\"word\":\"horse\",\"start\":13.15,\"end\":13.54}]"} {"utterance_id": "6455-67804-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0037.flac", "task_type": "lexical_detection", "query": "whatever", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"whatever\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.83, 5.21, "whatever"]], "response_mode": "offline", "answer": "[{\"word\":\"whatever\",\"start\":4.83,\"end\":5.21}]"} {"utterance_id": "4572-112381-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.49, 6.23, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":5.49,\"end\":6.23}]"} {"utterance_id": "7697-245712-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0007.flac", "task_type": "lexical_detection", "query": "excepted", "prompt": "When the utterance is complete, report all instances of the word \"excepted\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.39, 2.89, "excepted"], [6.74, 7.17, "excepted"]], "response_mode": "offline", "answer": "[{\"word\":\"excepted\",\"start\":2.39,\"end\":2.89},{\"word\":\"excepted\",\"start\":6.74,\"end\":7.17}]"} {"utterance_id": "6455-66379-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0007.flac", "task_type": "lexical_detection", "query": "mate", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mate\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.83, 2.15, "mate"]], "response_mode": "offline", "answer": "[{\"word\":\"mate\",\"start\":1.83,\"end\":2.15}]"} {"utterance_id": "3660-172182-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0025.flac", "task_type": "lexical_detection", "query": "became", "prompt": "After listening to the whole recording, output every instance of the word \"became\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.69, "became"]], "response_mode": "offline", "answer": "[{\"word\":\"became\",\"start\":1.34,\"end\":1.69}]"} {"utterance_id": "1630-141772-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0001.flac", "task_type": "lexical_detection", "query": "burning", "prompt": "When the utterance is complete, report all instances of the word \"burning\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.91, "burning"]], "response_mode": "offline", "answer": "[{\"word\":\"burning\",\"start\":4.32,\"end\":4.91}]"} {"utterance_id": "3663-172528-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 4.51, "francesco da monte varchi"]], "response_mode": "offline", "answer": "[{\"word\":\"francesco da monte varchi\",\"start\":2.16,\"end\":4.51}]"} {"utterance_id": "3663-172528-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0018.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.19, 2.65, "quack"]], "response_mode": "offline", "answer": "[{\"word\":\"quack\",\"start\":2.19,\"end\":2.65}]"} {"utterance_id": "3915-98647-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0015.flac", "task_type": "lexical_detection", "query": "burial", "prompt": "Once the audio has finished, identify every instance of the word \"burial\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[8.02, 8.45, "burial"]], "response_mode": "offline", "answer": "[{\"word\":\"burial\",\"start\":8.02,\"end\":8.45}]"} {"utterance_id": "6123-59186-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0032.flac", "task_type": "lexical_detection", "query": "hideous", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"hideous\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.06, 6.65, "hideous"]], "response_mode": "offline", "answer": "[{\"word\":\"hideous\",\"start\":6.06,\"end\":6.65}]"} {"utterance_id": "4323-55228-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.38, 4.78, "bertram's"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram's\",\"start\":4.38,\"end\":4.78}]"} {"utterance_id": "116-288047-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0022.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.09, 5.44, "millions"]], "response_mode": "offline", "answer": "[{\"word\":\"millions\",\"start\":5.09,\"end\":5.44}]"} {"utterance_id": "6123-59186-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0012.flac", "task_type": "lexical_detection", "query": "lines", "prompt": "After listening to the whole recording, output every instance of the word \"lines\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.35, 3.86, "lines"]], "response_mode": "offline", "answer": "[{\"word\":\"lines\",\"start\":3.35,\"end\":3.86}]"} {"utterance_id": "1255-74899-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0019.flac", "task_type": "lexical_detection", "query": "pink", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"pink\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.19, "pink"]], "response_mode": "offline", "answer": "[{\"word\":\"pink\",\"start\":0.79,\"end\":1.19}]"} {"utterance_id": "3915-98647-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0033.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "When the utterance is complete, report all instances of a plant name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.0, 12.59, "flowers"]], "response_mode": "offline", "answer": "[{\"word\":\"flowers\",\"start\":12.0,\"end\":12.59}]"} {"utterance_id": "3660-6517-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0023.flac", "task_type": "lexical_detection", "query": "veered", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"veered\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.01, 2.27, "veered"]], "response_mode": "offline", "answer": "[{\"word\":\"veered\",\"start\":2.01,\"end\":2.27}]"} {"utterance_id": "2506-11278-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0022.flac", "task_type": "lexical_detection", "query": "knob", "prompt": "When the utterance is complete, report all instances of the word \"knob\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.52, 5.96, "knob"]], "response_mode": "offline", "answer": "[{\"word\":\"knob\",\"start\":5.52,\"end\":5.96}]"} {"utterance_id": "1701-141759-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0024.flac", "task_type": "lexical_detection", "query": "mikhaylovna", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"mikhaylovna\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.64, 3.21, "mikhaylovna"]], "response_mode": "offline", "answer": "[{\"word\":\"mikhaylovna\",\"start\":2.64,\"end\":3.21}]"} {"utterance_id": "6467-94831-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0043.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[1.77, 2.09, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":1.77,\"end\":2.09}]"} {"utterance_id": "8254-84205-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0014.flac", "task_type": "lexical_detection", "query": "lad", "prompt": "When the utterance is complete, report all instances of the word \"lad\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.56, "lad"]], "response_mode": "offline", "answer": "[{\"word\":\"lad\",\"start\":1.15,\"end\":1.56}]"} {"utterance_id": "8254-84205-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0037.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 2.25, "ponies"]], "response_mode": "offline", "answer": "[{\"word\":\"ponies\",\"start\":1.81,\"end\":2.25}]"} {"utterance_id": "3660-172183-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0012.flac", "task_type": "lexical_detection", "query": "shouldst", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"shouldst\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.1, 1.46, "shouldst"]], "response_mode": "offline", "answer": "[{\"word\":\"shouldst\",\"start\":1.1,\"end\":1.46}]"} {"utterance_id": "5849-50873-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0018.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.71, 2.19, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":1.71,\"end\":2.19}]"} {"utterance_id": "6467-94831-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0019.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a disease name with fields word, start, end (seconds). If none, reply [].", "labels": [[2.27, 2.83, "consumption"]], "response_mode": "offline", "answer": "[{\"word\":\"consumption\",\"start\":2.27,\"end\":2.83}]"} {"utterance_id": "8254-115543-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0001.flac", "task_type": "lexical_detection", "query": "lot", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"lot\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.28, "lot"]], "response_mode": "offline", "answer": "[{\"word\":\"lot\",\"start\":3.14,\"end\":3.28}]"} {"utterance_id": "1701-141759-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0007.flac", "task_type": "lexical_detection", "query": "rushed", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"rushed\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.67, 0.97, "rushed"]], "response_mode": "offline", "answer": "[{\"word\":\"rushed\",\"start\":0.67,\"end\":0.97}]"} {"utterance_id": "6467-97061-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0009.flac", "task_type": "lexical_detection", "query": "torture", "prompt": "Process the full audio offline. Find every instance of the word \"torture\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.92, "torture"]], "response_mode": "offline", "answer": "[{\"word\":\"torture\",\"start\":3.38,\"end\":3.92}]"} {"utterance_id": "1630-96099-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0019.flac", "task_type": "lexical_detection", "query": "doubt", "prompt": "When the utterance is complete, report all instances of the word \"doubt\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.82, 1.11, "doubt"]], "response_mode": "offline", "answer": "[{\"word\":\"doubt\",\"start\":0.82,\"end\":1.11}]"} {"utterance_id": "4831-29134-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0011.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Once the audio has finished, identify every instance of a date. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.96, 1.35, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":0.96,\"end\":1.35}]"} {"utterance_id": "1255-138279-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0019.flac", "task_type": "lexical_detection", "query": "frank", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"frank\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.98, 4.41, "frank"]], "response_mode": "offline", "answer": "[{\"word\":\"frank\",\"start\":3.98,\"end\":4.41}]"} {"utterance_id": "7697-105815-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0037.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.09, 2.75, "beasts"]], "response_mode": "offline", "answer": "[{\"word\":\"beasts\",\"start\":2.09,\"end\":2.75}]"} {"utterance_id": "4153-186222-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0004.flac", "task_type": "lexical_detection", "query": "judged", "prompt": "After listening to the whole recording, output every instance of the word \"judged\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.87, 3.15, "judged"]], "response_mode": "offline", "answer": "[{\"word\":\"judged\",\"start\":2.87,\"end\":3.15}]"} {"utterance_id": "4153-185072-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0013.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.03, 5.51, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":5.03,\"end\":5.51}]"} {"utterance_id": "7641-96670-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.08, 3.53, "jason"]], "response_mode": "offline", "answer": "[{\"word\":\"jason\",\"start\":3.08,\"end\":3.53}]"} {"utterance_id": "4323-55228-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0003.flac", "task_type": "lexical_detection", "query": "object", "prompt": "When the utterance is complete, report all instances of the word \"object\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 1.76, "object"]], "response_mode": "offline", "answer": "[{\"word\":\"object\",\"start\":1.41,\"end\":1.76}]"} {"utterance_id": "1255-138279-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0020.flac", "task_type": "semantic_detection", "query": "money", "prompt": "When the utterance is complete, report all instances of a money or currency term including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.2, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":0.77,\"end\":1.2}]"} {"utterance_id": "3660-172183-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0015.flac", "task_type": "lexical_detection", "query": "according", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"according\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.35, 3.86, "according"]], "response_mode": "offline", "answer": "[{\"word\":\"according\",\"start\":3.35,\"end\":3.86}]"} {"utterance_id": "1686-142278-0095", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0095.flac", "task_type": "lexical_detection", "query": "blessing", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"blessing\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.61, 1.09, "blessing"]], "response_mode": "offline", "answer": "[{\"word\":\"blessing\",\"start\":0.61,\"end\":1.09}]"} {"utterance_id": "1651-136854-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0031.flac", "task_type": "lexical_detection", "query": "echoed", "prompt": "Process the full audio offline. Find every instance of the word \"echoed\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.94, 2.38, "echoed"]], "response_mode": "offline", "answer": "[{\"word\":\"echoed\",\"start\":1.94,\"end\":2.38}]"} {"utterance_id": "3660-6517-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0003.flac", "task_type": "lexical_detection", "query": "condition", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"condition\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.13, 3.69, "condition"]], "response_mode": "offline", "answer": "[{\"word\":\"condition\",\"start\":3.13,\"end\":3.69}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "lexical_detection", "query": "disagreeable", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"disagreeable\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.73, 2.68, "disagreeable"]], "response_mode": "offline", "answer": "[{\"word\":\"disagreeable\",\"start\":1.73,\"end\":2.68}]"} {"utterance_id": "4323-55228-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.77, "billy"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":0.5,\"end\":0.77}]"} {"utterance_id": "8254-115543-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0035.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.77, 3.14, "supper"]], "response_mode": "offline", "answer": "[{\"word\":\"supper\",\"start\":2.77,\"end\":3.14}]"} {"utterance_id": "4515-11057-0110", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0110.flac", "task_type": "semantic_detection", "query": "date", "prompt": "You will hear a complete utterance. When it ends, list every instance of a date you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 1.03, "yesterday"]], "response_mode": "offline", "answer": "[{\"word\":\"yesterday\",\"start\":0.46,\"end\":1.03}]"} {"utterance_id": "6123-59186-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0023.flac", "task_type": "lexical_counting", "query": "mood", "prompt": "Listen to the full audio. Count every occurrence of the word \"mood\". After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.64, 0.86, "one"], [6.23, 6.62, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.64,\"end\":0.86},{\"word\":\"two\",\"start\":6.23,\"end\":6.62}]"} {"utterance_id": "6123-59186-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0037.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After the audio ends, extract all occurrences of an ordinal number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.18, 1.58, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":1.18,\"end\":1.58}]"} {"utterance_id": "4572-112375-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0000.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the full audio. After it finishes, find every instance of a color name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.86, 6.48, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":5.86,\"end\":6.48}]"} {"utterance_id": "7697-105817-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0011.flac", "task_type": "lexical_detection", "query": "act", "prompt": "After the audio ends, extract all occurrences of the word \"act\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.04, "act"]], "response_mode": "offline", "answer": "[{\"word\":\"act\",\"start\":0.76,\"end\":1.04}]"} {"utterance_id": "1701-141760-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.59, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":3.05,\"end\":3.59}]"} {"utterance_id": "4570-56594-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0010.flac", "task_type": "lexical_detection", "query": "practice", "prompt": "Process the full audio offline. Find every instance of the word \"practice\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.59, 5.35, "practice"]], "response_mode": "offline", "answer": "[{\"word\":\"practice\",\"start\":4.59,\"end\":5.35}]"} {"utterance_id": "3660-172183-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0013.flac", "task_type": "lexical_detection", "query": "received", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"received\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 1.06, "received"]], "response_mode": "offline", "answer": "[{\"word\":\"received\",\"start\":0.57,\"end\":1.06}]"} {"utterance_id": "1686-142278-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0040.flac", "task_type": "lexical_detection", "query": "burden", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"burden\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.73, 3.17, "burden"]], "response_mode": "offline", "answer": "[{\"word\":\"burden\",\"start\":2.73,\"end\":3.17}]"} {"utterance_id": "700-122867-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0035.flac", "task_type": "lexical_detection", "query": "energies", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"energies\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.77, 2.24, "energies"]], "response_mode": "offline", "answer": "[{\"word\":\"energies\",\"start\":1.77,\"end\":2.24}]"} {"utterance_id": "7601-291468-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0003.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.88, 2.64, "flies"], [8.74, 9.49, "spiders"], [10.49, 11.19, "spiders"], [14.22, 14.92, "flies"]], "response_mode": "offline", "answer": "[{\"word\":\"flies\",\"start\":1.88,\"end\":2.64},{\"word\":\"spiders\",\"start\":8.74,\"end\":9.49},{\"word\":\"spiders\",\"start\":10.49,\"end\":11.19},{\"word\":\"flies\",\"start\":14.22,\"end\":14.92}]"} {"utterance_id": "1630-141772-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0012.flac", "task_type": "lexical_detection", "query": "reply", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"reply\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.69, "reply"]], "response_mode": "offline", "answer": "[{\"word\":\"reply\",\"start\":1.16,\"end\":1.69}]"} {"utterance_id": "4570-56594-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0010.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.22, 2.64, "preacher"]], "response_mode": "offline", "answer": "[{\"word\":\"preacher\",\"start\":2.22,\"end\":2.64}]"} {"utterance_id": "2506-11267-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0010.flac", "task_type": "lexical_detection", "query": "noble", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"noble\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.68, 1.22, "noble"]], "response_mode": "offline", "answer": "[{\"word\":\"noble\",\"start\":0.68,\"end\":1.22}]"} {"utterance_id": "8254-115543-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0002.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Process the full audio offline. Find every instance of a food name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.76, 7.13, "food"]], "response_mode": "offline", "answer": "[{\"word\":\"food\",\"start\":6.76,\"end\":7.13}]"} {"utterance_id": "1701-141760-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.02, 3.34, "berg"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":3.02,\"end\":3.34}]"} {"utterance_id": "8173-294714-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0033.flac", "task_type": "lexical_detection", "query": "doctor's", "prompt": "After the audio ends, extract all occurrences of the word \"doctor's\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 2.85, "doctor's"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor's\",\"start\":2.34,\"end\":2.85}]"} {"utterance_id": "1701-141759-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0013.flac", "task_type": "lexical_detection", "query": "nicholas", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"nicholas\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.9, 3.5, "nicholas"]], "response_mode": "offline", "answer": "[{\"word\":\"nicholas\",\"start\":2.9,\"end\":3.5}]"} {"utterance_id": "700-122866-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0037.flac", "task_type": "lexical_detection", "query": "parts", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"parts\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.11, 3.64, "parts"]], "response_mode": "offline", "answer": "[{\"word\":\"parts\",\"start\":3.11,\"end\":3.64}]"} {"utterance_id": "4570-14911-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0004.flac", "task_type": "lexical_detection", "query": "hound", "prompt": "After the audio ends, extract all occurrences of the word \"hound\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.22, 5.64, "hound"]], "response_mode": "offline", "answer": "[{\"word\":\"hound\",\"start\":5.22,\"end\":5.64}]"} {"utterance_id": "1701-141759-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0025.flac", "task_type": "lexical_counting", "query": "eyes", "prompt": "When the utterance is complete, report how the word \"eyes\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.11, 4.52, "one"], [10.7, 11.18, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.11,\"end\":4.52},{\"word\":\"two\",\"start\":10.7,\"end\":11.18}]"} {"utterance_id": "3663-172528-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0016.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.96, 5.55, "water"]], "response_mode": "offline", "answer": "[{\"word\":\"water\",\"start\":4.96,\"end\":5.55}]"} {"utterance_id": "700-122868-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.86, "gilbert"]], "response_mode": "offline", "answer": "[{\"word\":\"gilbert\",\"start\":0.48,\"end\":0.86}]"} {"utterance_id": "7641-96684-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0034.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.73, 9.91, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.73,\"end\":9.91}]"} {"utterance_id": "1701-141759-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0027.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.24, 5.22, "anna mikhaylovna s"]], "response_mode": "offline", "answer": "[{\"word\":\"anna mikhaylovna s\",\"start\":4.24,\"end\":5.22}]"} {"utterance_id": "3663-172528-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0033.flac", "task_type": "lexical_detection", "query": "maestro", "prompt": "Once the audio has finished, identify every instance of the word \"maestro\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.34, 0.99, "maestro"]], "response_mode": "offline", "answer": "[{\"word\":\"maestro\",\"start\":0.34,\"end\":0.99}]"} {"utterance_id": "1686-142278-0096", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0096.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Process the full audio offline. Find every instance of an ordinal number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.93, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":0.6,\"end\":0.93}]"} {"utterance_id": "700-122868-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0030.flac", "task_type": "lexical_detection", "query": "quick", "prompt": "When the utterance is complete, report all instances of the word \"quick\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.63, "quick"]], "response_mode": "offline", "answer": "[{\"word\":\"quick\",\"start\":1.15,\"end\":1.63}]"} {"utterance_id": "6455-67803-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.37, 3.18, "lady theobald"]], "response_mode": "offline", "answer": "[{\"word\":\"lady theobald\",\"start\":2.37,\"end\":3.18}]"} {"utterance_id": "4572-64670-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.0, 6.42, "rama's"], [8.86, 9.29, "sita's"]], "response_mode": "offline", "answer": "[{\"word\":\"rama's\",\"start\":6.0,\"end\":6.42},{\"word\":\"sita's\",\"start\":8.86,\"end\":9.29}]"} {"utterance_id": "4570-24733-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/24733/4570-24733-0000.flac", "task_type": "lexical_counting", "query": "attribute", "prompt": "After listening to the whole recording, output timed count words for the word \"attribute\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[4.22, 4.95, "one"], [16.55, 17.12, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.22,\"end\":4.95},{\"word\":\"two\",\"start\":16.55,\"end\":17.12}]"} {"utterance_id": "6267-65525-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0011.flac", "task_type": "lexical_detection", "query": "double", "prompt": "Once the audio has finished, identify every instance of the word \"double\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[14.36, 14.82, "double"]], "response_mode": "offline", "answer": "[{\"word\":\"double\",\"start\":14.36,\"end\":14.82}]"} {"utterance_id": "1701-141760-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0044.flac", "task_type": "semantic_detection", "query": "money", "prompt": "You will hear a complete utterance. When it ends, list every instance of a money or currency term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.52, 6.83, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":6.52,\"end\":6.83}]"} {"utterance_id": "7641-96684-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0004.flac", "task_type": "lexical_detection", "query": "radiant", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"radiant\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.46, 0.88, "radiant"]], "response_mode": "offline", "answer": "[{\"word\":\"radiant\",\"start\":0.46,\"end\":0.88}]"} {"utterance_id": "4153-186223-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 3.31, "kenneth forbes"]], "response_mode": "offline", "answer": "[{\"word\":\"kenneth forbes\",\"start\":2.46,\"end\":3.31}]"} {"utterance_id": "6267-53049-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0013.flac", "task_type": "lexical_detection", "query": "prettier", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"prettier\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.82, 3.54, "prettier"]], "response_mode": "offline", "answer": "[{\"word\":\"prettier\",\"start\":2.82,\"end\":3.54}]"} {"utterance_id": "4153-185072-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0001.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.33, 9.73, "cattle"]], "response_mode": "offline", "answer": "[{\"word\":\"cattle\",\"start\":9.33,\"end\":9.73}]"} {"utterance_id": "700-122866-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0023.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.48, 2.93, "twelve"]], "response_mode": "offline", "answer": "[{\"word\":\"twelve\",\"start\":2.48,\"end\":2.93}]"} {"utterance_id": "5849-50962-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0001.flac", "task_type": "lexical_detection", "query": "schooner", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"schooner\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 1.17, "schooner"]], "response_mode": "offline", "answer": "[{\"word\":\"schooner\",\"start\":0.63,\"end\":1.17}]"} {"utterance_id": "1650-167613-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0012.flac", "task_type": "lexical_detection", "query": "efforts", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"efforts\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.23, 3.75, "efforts"]], "response_mode": "offline", "answer": "[{\"word\":\"efforts\",\"start\":3.23,\"end\":3.75}]"} {"utterance_id": "1701-141759-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 3.21, "anna mikhaylovna"]], "response_mode": "offline", "answer": "[{\"word\":\"anna mikhaylovna\",\"start\":2.4,\"end\":3.21}]"} {"utterance_id": "8173-294714-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0057.flac", "task_type": "lexical_detection", "query": "saving", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"saving\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.3, 1.65, "saving"]], "response_mode": "offline", "answer": "[{\"word\":\"saving\",\"start\":1.3,\"end\":1.65}]"} {"utterance_id": "4831-18525-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0028.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.95, 2.28, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":1.95,\"end\":2.28}]"} {"utterance_id": "1701-141760-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0041.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.12, 10.63, "adjutants"]], "response_mode": "offline", "answer": "[{\"word\":\"adjutants\",\"start\":10.12,\"end\":10.63}]"} {"utterance_id": "3915-98647-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0030.flac", "task_type": "lexical_detection", "query": "release", "prompt": "When the utterance is complete, report all instances of the word \"release\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.47, 5.87, "release"]], "response_mode": "offline", "answer": "[{\"word\":\"release\",\"start\":5.47,\"end\":5.87}]"} {"utterance_id": "4323-55228-0047", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0047.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.41, 1.84, "william"]], "response_mode": "offline", "answer": "[{\"word\":\"william\",\"start\":1.41,\"end\":1.84}]"} {"utterance_id": "7601-175351-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0026.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "You will hear a complete utterance. When it ends, list every instance of an organization name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.47, "forward"], [9.76, 10.4, "hatteras"], [12.03, 12.68, "yankees"]], "response_mode": "offline", "answer": "[{\"word\":\"forward\",\"start\":2.95,\"end\":3.47},{\"word\":\"hatteras\",\"start\":9.76,\"end\":10.4},{\"word\":\"yankees\",\"start\":12.03,\"end\":12.68}]"} {"utterance_id": "1255-74899-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.46, 9.59, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":9.46,\"end\":9.59}]"} {"utterance_id": "5849-50962-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0023.flac", "task_type": "lexical_detection", "query": "bingham", "prompt": "After the audio ends, extract all occurrences of the word \"bingham\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.24, 1.57, "bingham"]], "response_mode": "offline", "answer": "[{\"word\":\"bingham\",\"start\":1.24,\"end\":1.57}]"} {"utterance_id": "4323-18416-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.15, 1.67, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.15,\"end\":1.67}]"} {"utterance_id": "1585-157660-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0013.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "When the utterance is complete, report how a body part accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.8, "one"], [9.24, 9.62, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.24,\"end\":2.8},{\"word\":\"two\",\"start\":9.24,\"end\":9.62}]"} {"utterance_id": "4323-55228-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0032.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.39, "finger"]], "response_mode": "offline", "answer": "[{\"word\":\"finger\",\"start\":1.09,\"end\":1.39}]"} {"utterance_id": "6841-88291-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0052.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.47, 2.02, "heels"], [3.2, 3.42, "ears"]], "response_mode": "offline", "answer": "[{\"word\":\"heels\",\"start\":1.47,\"end\":2.02},{\"word\":\"ears\",\"start\":3.2,\"end\":3.42}]"} {"utterance_id": "4153-186222-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0019.flac", "task_type": "lexical_detection", "query": "corner", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"corner\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.19, "corner"]], "response_mode": "offline", "answer": "[{\"word\":\"corner\",\"start\":1.84,\"end\":2.19}]"} {"utterance_id": "4515-11057-0106", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0106.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.83, 3.25, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":2.83,\"end\":3.25}]"} {"utterance_id": "4831-25894-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0035.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.87, 7.16, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":6.87,\"end\":7.16}]"} {"utterance_id": "4831-25894-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.74, 5.15, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":4.74,\"end\":5.15}]"} {"utterance_id": "4831-25894-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.16, 2.86, "bella monica"], [5.44, 5.77, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"bella monica\",\"start\":2.16,\"end\":2.86},{\"word\":\"tommo\",\"start\":5.44,\"end\":5.77}]"} {"utterance_id": "8173-294714-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0003.flac", "task_type": "lexical_detection", "query": "cruel", "prompt": "Process the full audio offline. Find every instance of the word \"cruel\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.41, 0.86, "cruel"]], "response_mode": "offline", "answer": "[{\"word\":\"cruel\",\"start\":0.41,\"end\":0.86}]"} {"utterance_id": "2506-11267-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0007.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After listening to the whole recording, output every instance of a plant name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.79, "strawberry leaves"]], "response_mode": "offline", "answer": "[{\"word\":\"strawberry leaves\",\"start\":0.78,\"end\":1.79}]"} {"utterance_id": "5849-50964-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0003.flac", "task_type": "lexical_detection", "query": "pictures", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"pictures\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.22, 2.67, "pictures"]], "response_mode": "offline", "answer": "[{\"word\":\"pictures\",\"start\":2.22,\"end\":2.67}]"} {"utterance_id": "8254-84205-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0023.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[2.42, 2.74, "tongue"]], "response_mode": "offline", "answer": "[{\"word\":\"tongue\",\"start\":2.42,\"end\":2.74}]"} {"utterance_id": "4831-18525-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.76, 10.22, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":9.76,\"end\":10.22}]"} {"utterance_id": "8173-294714-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0021.flac", "task_type": "lexical_detection", "query": "situation", "prompt": "After listening to the whole recording, output every instance of the word \"situation\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.69, "situation"]], "response_mode": "offline", "answer": "[{\"word\":\"situation\",\"start\":0.96,\"end\":1.69}]"} {"utterance_id": "700-122867-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0020.flac", "task_type": "lexical_detection", "query": "house", "prompt": "Process the full audio offline. Find every instance of the word \"house\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.7, 2.28, "house"]], "response_mode": "offline", "answer": "[{\"word\":\"house\",\"start\":1.7,\"end\":2.28}]"} {"utterance_id": "4153-61735-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0004.flac", "task_type": "lexical_detection", "query": "hacienda", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"hacienda\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.57, 4.41, "hacienda"]], "response_mode": "offline", "answer": "[{\"word\":\"hacienda\",\"start\":3.57,\"end\":4.41}]"} {"utterance_id": "1255-90413-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0011.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Once the audio has finished, identify every instance of a color name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[6.05, 6.73, "greenish blue"]], "response_mode": "offline", "answer": "[{\"word\":\"greenish blue\",\"start\":6.05,\"end\":6.73}]"} {"utterance_id": "116-288048-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0027.flac", "task_type": "lexical_detection", "query": "produced", "prompt": "Process the full audio offline. Find every instance of the word \"produced\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 3.99, "produced"]], "response_mode": "offline", "answer": "[{\"word\":\"produced\",\"start\":3.36,\"end\":3.99}]"} {"utterance_id": "7641-96252-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0009.flac", "task_type": "lexical_detection", "query": "opposite", "prompt": "After the audio ends, extract all occurrences of the word \"opposite\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 1.11, "opposite"]], "response_mode": "offline", "answer": "[{\"word\":\"opposite\",\"start\":0.66,\"end\":1.11}]"} {"utterance_id": "6455-67804-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0039.flac", "task_type": "lexical_detection", "query": "move", "prompt": "When the utterance is complete, report all instances of the word \"move\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 0.91, "move"]], "response_mode": "offline", "answer": "[{\"word\":\"move\",\"start\":0.7,\"end\":0.91}]"} {"utterance_id": "6841-88291-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0049.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.1, "homer"]], "response_mode": "offline", "answer": "[{\"word\":\"homer\",\"start\":0.77,\"end\":1.1}]"} {"utterance_id": "3663-172528-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0002.flac", "task_type": "lexical_detection", "query": "felice", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"felice\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.77, "felice"]], "response_mode": "offline", "answer": "[{\"word\":\"felice\",\"start\":0.34,\"end\":0.77}]"} {"utterance_id": "3660-172183-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0022.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.46, "meal"]], "response_mode": "offline", "answer": "[{\"word\":\"meal\",\"start\":1.1,\"end\":1.46}]"} {"utterance_id": "1650-167613-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0013.flac", "task_type": "lexical_detection", "query": "barclay", "prompt": "When the utterance is complete, report all instances of the word \"barclay\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.15, "barclay"]], "response_mode": "offline", "answer": "[{\"word\":\"barclay\",\"start\":1.57,\"end\":2.15}]"} {"utterance_id": "700-122867-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0030.flac", "task_type": "semantic_detection", "query": "color", "prompt": "After listening to the whole recording, output every instance of a color name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.6, 5.93, "red"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":5.6,\"end\":5.93}]"} {"utterance_id": "7697-105817-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0006.flac", "task_type": "lexical_detection", "query": "fears", "prompt": "Process the full audio offline. Find every instance of the word \"fears\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.71, "fears"]], "response_mode": "offline", "answer": "[{\"word\":\"fears\",\"start\":2.27,\"end\":2.71}]"} {"utterance_id": "6599-38591-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0005.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of a country name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.28, 2.68, "piedmont"]], "response_mode": "offline", "answer": "[{\"word\":\"piedmont\",\"start\":2.28,\"end\":2.68}]"} {"utterance_id": "4570-56594-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0004.flac", "task_type": "lexical_detection", "query": "wore", "prompt": "When the utterance is complete, report all instances of the word \"wore\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.4, 4.66, "wore"]], "response_mode": "offline", "answer": "[{\"word\":\"wore\",\"start\":4.4,\"end\":4.66}]"} {"utterance_id": "1650-173552-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0003.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Do not answer until the audio ends. Then list each instance of a color name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.51, 3.92, "golden"]], "response_mode": "offline", "answer": "[{\"word\":\"golden\",\"start\":3.51,\"end\":3.92}]"} {"utterance_id": "6841-88294-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0031.flac", "task_type": "lexical_detection", "query": "asleep", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"asleep\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.93, 2.37, "asleep"]], "response_mode": "offline", "answer": "[{\"word\":\"asleep\",\"start\":1.93,\"end\":2.37}]"} {"utterance_id": "5849-50873-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0040.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After listening to the whole recording, output every instance of an organization name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.92, 2.5, "assembly"], [10.18, 11.38, "legislative council"]], "response_mode": "offline", "answer": "[{\"word\":\"assembly\",\"start\":1.92,\"end\":2.5},{\"word\":\"legislative council\",\"start\":10.18,\"end\":11.38}]"} {"utterance_id": "4153-185072-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0014.flac", "task_type": "lexical_detection", "query": "escape", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"escape\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[13.65, 14.11, "escape"]], "response_mode": "offline", "answer": "[{\"word\":\"escape\",\"start\":13.65,\"end\":14.11}]"} {"utterance_id": "5849-50962-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0019.flac", "task_type": "lexical_detection", "query": "welcome", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"welcome\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.39, "welcome"]], "response_mode": "offline", "answer": "[{\"word\":\"welcome\",\"start\":2.02,\"end\":2.39}]"} {"utterance_id": "4515-11057-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0058.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.88, 5.33, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":4.88,\"end\":5.33}]"} {"utterance_id": "116-288048-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0017.flac", "task_type": "lexical_detection", "query": "paul", "prompt": "Once the audio has finished, identify every instance of the word \"paul\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.11, 1.53, "paul"]], "response_mode": "offline", "answer": "[{\"word\":\"paul\",\"start\":1.11,\"end\":1.53}]"} {"utterance_id": "1686-142278-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0004.flac", "task_type": "lexical_detection", "query": "thoughts", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"thoughts\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.64, 2.91, "thoughts"]], "response_mode": "offline", "answer": "[{\"word\":\"thoughts\",\"start\":2.64,\"end\":2.91}]"} {"utterance_id": "4831-29134-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0013.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Process the full audio offline. Find every instance of a date and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.05, 4.42, "sunday"]], "response_mode": "offline", "answer": "[{\"word\":\"sunday\",\"start\":4.05,\"end\":4.42}]"} {"utterance_id": "7697-245715-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245715/7697-245715-0001.flac", "task_type": "lexical_detection", "query": "imperfect", "prompt": "When the utterance is complete, report all instances of the word \"imperfect\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.64, 4.53, "imperfect"]], "response_mode": "offline", "answer": "[{\"word\":\"imperfect\",\"start\":3.64,\"end\":4.53}]"} {"utterance_id": "8254-84205-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0023.flac", "task_type": "lexical_detection", "query": "hold", "prompt": "Process the full audio offline. Find every instance of the word \"hold\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.11, 2.31, "hold"]], "response_mode": "offline", "answer": "[{\"word\":\"hold\",\"start\":2.11,\"end\":2.31}]"} {"utterance_id": "1630-141772-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0003.flac", "task_type": "lexical_detection", "query": "feebly", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"feebly\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.4, 0.88, "feebly"]], "response_mode": "offline", "answer": "[{\"word\":\"feebly\",\"start\":0.4,\"end\":0.88}]"} {"utterance_id": "116-288046-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0002.flac", "task_type": "lexical_counting", "query": "faith", "prompt": "Process the full audio offline. For every occurrence of the word \"faith\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.84, 7.29, "one"], [8.69, 9.2, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.84,\"end\":7.29},{\"word\":\"two\",\"start\":8.69,\"end\":9.2}]"} {"utterance_id": "6467-94831-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0004.flac", "task_type": "lexical_detection", "query": "thomas", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"thomas\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.86, 1.38, "thomas"]], "response_mode": "offline", "answer": "[{\"word\":\"thomas\",\"start\":0.86,\"end\":1.38}]"} {"utterance_id": "1585-131718-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0038.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen to the full audio. After it finishes, find every instance of an ordinal number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.36, 3.64, "eighth"]], "response_mode": "offline", "answer": "[{\"word\":\"eighth\",\"start\":3.36,\"end\":3.64}]"} {"utterance_id": "4323-55228-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0016.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.92, 4.23, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":3.92,\"end\":4.23}]"} {"utterance_id": "3663-172528-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0009.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "When the utterance is complete, report all instances of a disease name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.18, 7.78, "illness"]], "response_mode": "offline", "answer": "[{\"word\":\"illness\",\"start\":7.18,\"end\":7.78}]"} {"utterance_id": "4515-11057-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0012.flac", "task_type": "semantic_counting", "query": "city", "prompt": "Listen to the full audio. Count every occurrence of a city name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 2.38, "one"], [3.52, 3.89, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.4,\"end\":2.38},{\"word\":\"two\",\"start\":3.52,\"end\":3.89}]"} {"utterance_id": "1585-157660-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0007.flac", "task_type": "lexical_detection", "query": "fame", "prompt": "When the utterance is complete, report all instances of the word \"fame\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.42, 4.84, "fame"]], "response_mode": "offline", "answer": "[{\"word\":\"fame\",\"start\":4.42,\"end\":4.84}]"} {"utterance_id": "2506-169427-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.68, 2.29, "yorick"]], "response_mode": "offline", "answer": "[{\"word\":\"yorick\",\"start\":1.68,\"end\":2.29}]"} {"utterance_id": "4323-18416-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0000.flac", "task_type": "lexical_detection", "query": "ineffectual", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"ineffectual\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.93, "ineffectual"]], "response_mode": "offline", "answer": "[{\"word\":\"ineffectual\",\"start\":2.24,\"end\":2.93}]"} {"utterance_id": "1585-131718-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.28, "inventor"]], "response_mode": "offline", "answer": "[{\"word\":\"inventor\",\"start\":0.87,\"end\":1.28}]"} {"utterance_id": "700-122867-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0009.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.69, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":0.52,\"end\":0.69}]"} {"utterance_id": "7697-105815-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0003.flac", "task_type": "lexical_detection", "query": "stands", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"stands\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.58, 1.95, "stands"]], "response_mode": "offline", "answer": "[{\"word\":\"stands\",\"start\":1.58,\"end\":1.95}]"} {"utterance_id": "5849-50873-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0029.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[4.4, 5.35, "captain wadsworth"]], "response_mode": "offline", "answer": "[{\"word\":\"captain wadsworth\",\"start\":4.4,\"end\":5.35}]"} {"utterance_id": "116-288048-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0014.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.7, 4.23, "historian"]], "response_mode": "offline", "answer": "[{\"word\":\"historian\",\"start\":3.7,\"end\":4.23}]"} {"utterance_id": "1701-141760-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0003.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.51, 0.89, "one"], [3.97, 4.41, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.51,\"end\":0.89},{\"word\":\"two\",\"start\":3.97,\"end\":4.41}]"} {"utterance_id": "6599-38591-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0001.flac", "task_type": "semantic_detection", "query": "food", "prompt": "When the utterance is complete, report all instances of a food name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.35, 5.84, "medicine"]], "response_mode": "offline", "answer": "[{\"word\":\"medicine\",\"start\":5.35,\"end\":5.84}]"} {"utterance_id": "7601-175351-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0002.flac", "task_type": "lexical_detection", "query": "fires", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"fires\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.95, 1.68, "fires"]], "response_mode": "offline", "answer": "[{\"word\":\"fires\",\"start\":0.95,\"end\":1.68}]"} {"utterance_id": "4323-18416-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.58, 1.97, "phronsie"]], "response_mode": "offline", "answer": "[{\"word\":\"phronsie\",\"start\":1.58,\"end\":1.97}]"} {"utterance_id": "8254-84205-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0030.flac", "task_type": "lexical_detection", "query": "anxiously", "prompt": "Once the audio has finished, identify every instance of the word \"anxiously\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.91, 2.57, "anxiously"]], "response_mode": "offline", "answer": "[{\"word\":\"anxiously\",\"start\":1.91,\"end\":2.57}]"} {"utterance_id": "7601-101622-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0002.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[2.79, 3.75, "one"], [5.57, 6.08, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.79,\"end\":3.75},{\"word\":\"two\",\"start\":5.57,\"end\":6.08}]"} {"utterance_id": "3663-172528-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.71, 5.75, "pier landi"]], "response_mode": "offline", "answer": "[{\"word\":\"pier landi\",\"start\":4.71,\"end\":5.75}]"} {"utterance_id": "3915-57461-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0003.flac", "task_type": "lexical_counting", "query": "boats", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"boats\" with word/start/end. If none, reply [].", "labels": [[3.29, 3.72, "one"], [8.38, 8.7, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.29,\"end\":3.72},{\"word\":\"two\",\"start\":8.38,\"end\":8.7}]"} {"utterance_id": "700-122868-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0010.flac", "task_type": "lexical_detection", "query": "born", "prompt": "After listening to the whole recording, output every instance of the word \"born\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.15, 4.47, "born"]], "response_mode": "offline", "answer": "[{\"word\":\"born\",\"start\":4.15,\"end\":4.47}]"} {"utterance_id": "3663-172528-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0046.flac", "task_type": "lexical_detection", "query": "responded", "prompt": "After the audio ends, extract all occurrences of the word \"responded\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 1.24, "responded"]], "response_mode": "offline", "answer": "[{\"word\":\"responded\",\"start\":0.59,\"end\":1.24}]"} {"utterance_id": "4831-18525-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0001.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[3.63, 4.01, "one"], [9.13, 9.57, "two"], [10.7, 11.46, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.63,\"end\":4.01},{\"word\":\"two\",\"start\":9.13,\"end\":9.57},{\"word\":\"three\",\"start\":10.7,\"end\":11.46}]"} {"utterance_id": "1255-138279-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0001.flac", "task_type": "lexical_detection", "query": "familiar", "prompt": "Once the audio has finished, identify every instance of the word \"familiar\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.94, 10.46, "familiar"]], "response_mode": "offline", "answer": "[{\"word\":\"familiar\",\"start\":9.94,\"end\":10.46}]"} {"utterance_id": "6467-94831-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0031.flac", "task_type": "lexical_detection", "query": "finished", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"finished\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.1, 2.56, "finished"]], "response_mode": "offline", "answer": "[{\"word\":\"finished\",\"start\":2.1,\"end\":2.56}]"} {"utterance_id": "6455-67803-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0029.flac", "task_type": "lexical_detection", "query": "made", "prompt": "Once the audio has finished, identify every instance of the word \"made\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.78, 0.98, "made"]], "response_mode": "offline", "answer": "[{\"word\":\"made\",\"start\":0.78,\"end\":0.98}]"} {"utterance_id": "4570-102353-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0002.flac", "task_type": "lexical_detection", "query": "thorough", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"thorough\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.7, 5.06, "thorough"]], "response_mode": "offline", "answer": "[{\"word\":\"thorough\",\"start\":4.7,\"end\":5.06}]"} {"utterance_id": "1686-142278-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0016.flac", "task_type": "semantic_detection", "query": "country", "prompt": "When the utterance is complete, report all instances of a country name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.94, 3.44, "england"]], "response_mode": "offline", "answer": "[{\"word\":\"england\",\"start\":2.94,\"end\":3.44}]"} {"utterance_id": "4831-25894-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0002.flac", "task_type": "lexical_detection", "query": "harp", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"harp\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.88, 3.18, "harp"]], "response_mode": "offline", "answer": "[{\"word\":\"harp\",\"start\":2.88,\"end\":3.18}]"} {"utterance_id": "116-288047-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0017.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.41, 2.07, "child"]], "response_mode": "offline", "answer": "[{\"word\":\"child\",\"start\":1.41,\"end\":2.07}]"} {"utterance_id": "4515-11057-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0029.flac", "task_type": "lexical_detection", "query": "recall", "prompt": "When the utterance is complete, report all instances of the word \"recall\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.83, "recall"]], "response_mode": "offline", "answer": "[{\"word\":\"recall\",\"start\":0.34,\"end\":0.83}]"} {"utterance_id": "1701-141760-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0025.flac", "task_type": "lexical_counting", "query": "letter", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for the word \"letter\" with word/start/end. If none, reply [].", "labels": [[0.75, 1.05, "one"], [2.39, 2.69, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.75,\"end\":1.05},{\"word\":\"two\",\"start\":2.39,\"end\":2.69}]"} {"utterance_id": "8288-274150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.94, 2.62, "philosopher"]], "response_mode": "offline", "answer": "[{\"word\":\"philosopher\",\"start\":1.94,\"end\":2.62}]"} {"utterance_id": "1585-131718-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0045.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.29, 0.62, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.29,\"end\":0.62}]"} {"utterance_id": "1686-142278-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.88, 3.31, "frederick"]], "response_mode": "offline", "answer": "[{\"word\":\"frederick\",\"start\":2.88,\"end\":3.31}]"} {"utterance_id": "1650-167613-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.4, 2.83, "pierre's"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre's\",\"start\":2.4,\"end\":2.83}]"} {"utterance_id": "2506-11267-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0004.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After the audio ends, extract all occurrences of a city name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.05, 6.58, "rome"]], "response_mode": "offline", "answer": "[{\"word\":\"rome\",\"start\":6.05,\"end\":6.58}]"} {"utterance_id": "5849-50873-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0017.flac", "task_type": "lexical_detection", "query": "late", "prompt": "Once the audio has finished, identify every instance of the word \"late\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.44, 4.75, "late"]], "response_mode": "offline", "answer": "[{\"word\":\"late\",\"start\":4.44,\"end\":4.75}]"} {"utterance_id": "3663-172528-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0004.flac", "task_type": "lexical_detection", "query": "earnest", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"earnest\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.98, 6.56, "earnest"]], "response_mode": "offline", "answer": "[{\"word\":\"earnest\",\"start\":5.98,\"end\":6.56}]"} {"utterance_id": "4153-186222-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0009.flac", "task_type": "lexical_detection", "query": "mighty", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"mighty\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.12, 2.43, "mighty"]], "response_mode": "offline", "answer": "[{\"word\":\"mighty\",\"start\":2.12,\"end\":2.43}]"} {"utterance_id": "1585-131718-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0008.flac", "task_type": "lexical_detection", "query": "local", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"local\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.8, 6.16, "local"]], "response_mode": "offline", "answer": "[{\"word\":\"local\",\"start\":5.8,\"end\":6.16}]"} {"utterance_id": "6455-66379-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0013.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.15, 1.38, "voice"]], "response_mode": "offline", "answer": "[{\"word\":\"voice\",\"start\":1.15,\"end\":1.38}]"} {"utterance_id": "1701-141760-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.76, "berg"], [10.68, 11.82, "emperor alexander"], [14.26, 14.73, "rostov"]], "response_mode": "offline", "answer": "[{\"word\":\"berg\",\"start\":0.51,\"end\":0.76},{\"word\":\"emperor alexander\",\"start\":10.68,\"end\":11.82},{\"word\":\"rostov\",\"start\":14.26,\"end\":14.73}]"} {"utterance_id": "6455-66379-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0017.flac", "task_type": "lexical_detection", "query": "stream", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"stream\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.18, 3.56, "stream"]], "response_mode": "offline", "answer": "[{\"word\":\"stream\",\"start\":3.18,\"end\":3.56}]"} {"utterance_id": "4831-25894-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.27, 3.67, "tessa"]], "response_mode": "offline", "answer": "[{\"word\":\"tessa\",\"start\":3.27,\"end\":3.67}]"} {"utterance_id": "2506-169427-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/169427/2506-169427-0002.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[16.07, 16.25, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":16.07,\"end\":16.25}]"} {"utterance_id": "4570-102353-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0001.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Once the audio has finished, identify every instance of a country name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[10.02, 11.2, "united states"]], "response_mode": "offline", "answer": "[{\"word\":\"united states\",\"start\":10.02,\"end\":11.2}]"} {"utterance_id": "1255-138279-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/138279/1255-138279-0008.flac", "task_type": "lexical_detection", "query": "three", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"three\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 2.0, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":1.46,\"end\":2.0}]"} {"utterance_id": "7697-105815-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0048.flac", "task_type": "lexical_detection", "query": "bring", "prompt": "After the audio ends, extract all occurrences of the word \"bring\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.15, 2.34, "bring"]], "response_mode": "offline", "answer": "[{\"word\":\"bring\",\"start\":2.15,\"end\":2.34}]"} {"utterance_id": "4831-25894-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0006.flac", "task_type": "lexical_detection", "query": "ranza", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"ranza\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.21, 1.53, "ranza"]], "response_mode": "offline", "answer": "[{\"word\":\"ranza\",\"start\":1.21,\"end\":1.53}]"} {"utterance_id": "8254-84205-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0039.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.27, 2.59, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":2.27,\"end\":2.59}]"} {"utterance_id": "3660-6517-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0012.flac", "task_type": "semantic_detection", "query": "city", "prompt": "When the utterance is complete, report all instances of a city name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.73, 5.2, "melbourne"]], "response_mode": "offline", "answer": "[{\"word\":\"melbourne\",\"start\":4.73,\"end\":5.2}]"} {"utterance_id": "4831-25894-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0015.flac", "task_type": "lexical_detection", "query": "share", "prompt": "Once the audio has finished, identify every instance of the word \"share\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.71, 5.11, "share"]], "response_mode": "offline", "answer": "[{\"word\":\"share\",\"start\":4.71,\"end\":5.11}]"} {"utterance_id": "5543-27761-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0042.flac", "task_type": "lexical_detection", "query": "sign", "prompt": "After the audio ends, extract all occurrences of the word \"sign\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.88, "sign"]], "response_mode": "offline", "answer": "[{\"word\":\"sign\",\"start\":1.19,\"end\":1.88}]"} {"utterance_id": "6267-65525-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0002.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.89, 8.77, "aunt jane"]], "response_mode": "offline", "answer": "[{\"word\":\"aunt jane\",\"start\":7.89,\"end\":8.77}]"} {"utterance_id": "6123-59186-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0034.flac", "task_type": "lexical_detection", "query": "futility", "prompt": "Process the full audio offline. Find every instance of the word \"futility\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 1.21, "futility"]], "response_mode": "offline", "answer": "[{\"word\":\"futility\",\"start\":0.49,\"end\":1.21}]"} {"utterance_id": "1686-142278-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0020.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.42, "bishop"]], "response_mode": "offline", "answer": "[{\"word\":\"bishop\",\"start\":1.09,\"end\":1.42}]"} {"utterance_id": "8254-84205-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0075.flac", "task_type": "lexical_detection", "query": "lookout", "prompt": "After the audio ends, extract all occurrences of the word \"lookout\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.7, "lookout"]], "response_mode": "offline", "answer": "[{\"word\":\"lookout\",\"start\":1.37,\"end\":1.7}]"} {"utterance_id": "7641-96684-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0012.flac", "task_type": "lexical_counting", "query": "stones", "prompt": "Process the full audio offline. For every occurrence of the word \"stones\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.18, "one"], [5.83, 6.42, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.69,\"end\":3.18},{\"word\":\"two\",\"start\":5.83,\"end\":6.42}]"} {"utterance_id": "8254-115543-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0032.flac", "task_type": "lexical_detection", "query": "taking", "prompt": "Once the audio has finished, identify every instance of the word \"taking\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.25, 5.55, "taking"]], "response_mode": "offline", "answer": "[{\"word\":\"taking\",\"start\":5.25,\"end\":5.55}]"} {"utterance_id": "6267-65525-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0023.flac", "task_type": "lexical_detection", "query": "hum", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"hum\" with fields word, start, end (seconds). If none, reply [].", "labels": [[7.85, 8.22, "hum"]], "response_mode": "offline", "answer": "[{\"word\":\"hum\",\"start\":7.85,\"end\":8.22}]"} {"utterance_id": "4323-55228-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0002.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.24, 4.56, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":4.24,\"end\":4.56}]"} {"utterance_id": "1630-73710-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0001.flac", "task_type": "lexical_counting", "query": "hope", "prompt": "After listening to the whole recording, output timed count words for the word \"hope\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.79, 2.25, "one"], [3.76, 4.29, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.79,\"end\":2.25},{\"word\":\"two\",\"start\":3.76,\"end\":4.29}]"} {"utterance_id": "4153-186223-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.4, "jane"]], "response_mode": "offline", "answer": "[{\"word\":\"jane\",\"start\":1.19,\"end\":1.4}]"} {"utterance_id": "4153-61735-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0001.flac", "task_type": "semantic_detection", "query": "country", "prompt": "Listen to the full audio. After it finishes, find every instance of a country name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.02, 2.36, "earth"]], "response_mode": "offline", "answer": "[{\"word\":\"earth\",\"start\":2.02,\"end\":2.36}]"} {"utterance_id": "700-122866-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0039.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.91, 1.7, "missus allan"]], "response_mode": "offline", "answer": "[{\"word\":\"missus allan\",\"start\":0.91,\"end\":1.7}]"} {"utterance_id": "7641-96252-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0011.flac", "task_type": "lexical_counting", "query": "disposition", "prompt": "Listen to the entire clip, then report the count sequence for the word \"disposition\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.46, 5.26, "one"], [13.09, 13.87, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.46,\"end\":5.26},{\"word\":\"two\",\"start\":13.09,\"end\":13.87}]"} {"utterance_id": "3663-172005-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0000.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "You will hear a complete utterance. When it ends, list every instance of a plant name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.5, 6.08, "moss"]], "response_mode": "offline", "answer": "[{\"word\":\"moss\",\"start\":5.5,\"end\":6.08}]"} {"utterance_id": "7641-96670-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0026.flac", "task_type": "lexical_detection", "query": "fear", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"fear\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.69, 4.13, "fear"]], "response_mode": "offline", "answer": "[{\"word\":\"fear\",\"start\":3.69,\"end\":4.13}]"} {"utterance_id": "5849-50964-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0012.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a person's name with word/start/end. If none, reply [].", "labels": [[3.61, 4.14, "one"], [5.21, 5.54, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.61,\"end\":4.14},{\"word\":\"two\",\"start\":5.21,\"end\":5.54}]"} {"utterance_id": "8254-84205-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0066.flac", "task_type": "lexical_detection", "query": "month", "prompt": "When the utterance is complete, report all instances of the word \"month\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.51, 5.87, "month"]], "response_mode": "offline", "answer": "[{\"word\":\"month\",\"start\":5.51,\"end\":5.87}]"} {"utterance_id": "3663-172005-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172005/3663-172005-0006.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 0.93, "doe"]], "response_mode": "offline", "answer": "[{\"word\":\"doe\",\"start\":0.6,\"end\":0.93}]"} {"utterance_id": "7641-96252-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0000.flac", "task_type": "lexical_detection", "query": "dogmas", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"dogmas\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.07, 1.62, "dogmas"]], "response_mode": "offline", "answer": "[{\"word\":\"dogmas\",\"start\":1.07,\"end\":1.62}]"} {"utterance_id": "6599-38590-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0004.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen carefully to the entire clip, then report every instance of a color name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 0.89, "burgundy"]], "response_mode": "offline", "answer": "[{\"word\":\"burgundy\",\"start\":0.32,\"end\":0.89}]"} {"utterance_id": "1701-141759-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0012.flac", "task_type": "lexical_detection", "query": "petya", "prompt": "When the utterance is complete, report all instances of the word \"petya\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.2, 1.7, "petya"]], "response_mode": "offline", "answer": "[{\"word\":\"petya\",\"start\":1.2,\"end\":1.7}]"} {"utterance_id": "1686-142278-0089", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0089.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.19, "papa"]], "response_mode": "offline", "answer": "[{\"word\":\"papa\",\"start\":0.77,\"end\":1.19}]"} {"utterance_id": "6267-65525-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0043.flac", "task_type": "lexical_detection", "query": "deceive", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"deceive\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.91, 1.64, "deceive"]], "response_mode": "offline", "answer": "[{\"word\":\"deceive\",\"start\":0.91,\"end\":1.64}]"} {"utterance_id": "6841-88291-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0054.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.09, "wooden"]], "response_mode": "offline", "answer": "[{\"word\":\"wooden\",\"start\":0.8,\"end\":1.09}]"} {"utterance_id": "6123-59150-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0013.flac", "task_type": "lexical_detection", "query": "cautiously", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"cautiously\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.55, 4.14, "cautiously"]], "response_mode": "offline", "answer": "[{\"word\":\"cautiously\",\"start\":3.55,\"end\":4.14}]"} {"utterance_id": "1650-167613-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.59, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":2.69,\"end\":3.59}]"} {"utterance_id": "4323-18416-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0021.flac", "task_type": "lexical_detection", "query": "shouldn't", "prompt": "When the utterance is complete, report all instances of the word \"shouldn't\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.07, 3.35, "shouldn't"]], "response_mode": "offline", "answer": "[{\"word\":\"shouldn't\",\"start\":3.07,\"end\":3.35}]"} {"utterance_id": "6123-59150-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0017.flac", "task_type": "lexical_detection", "query": "passion", "prompt": "Process the full audio offline. Find every instance of the word \"passion\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.12, 4.66, "passion"]], "response_mode": "offline", "answer": "[{\"word\":\"passion\",\"start\":4.12,\"end\":4.66}]"} {"utterance_id": "8288-274162-0065", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0065.flac", "task_type": "lexical_detection", "query": "looking", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"looking\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.77, 3.12, "looking"]], "response_mode": "offline", "answer": "[{\"word\":\"looking\",\"start\":2.77,\"end\":3.12}]"} {"utterance_id": "1585-131718-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0025.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "Process the full audio offline. For every occurrence of a number, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.8, "one"], [5.96, 6.27, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":0.8},{\"word\":\"two\",\"start\":5.96,\"end\":6.27}]"} {"utterance_id": "7641-96684-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0000.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.99, 1.4, "apollo"]], "response_mode": "offline", "answer": "[{\"word\":\"apollo\",\"start\":0.99,\"end\":1.4}]"} {"utterance_id": "6467-97061-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0005.flac", "task_type": "lexical_detection", "query": "advancing", "prompt": "When the utterance is complete, report all instances of the word \"advancing\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.19, 0.77, "advancing"]], "response_mode": "offline", "answer": "[{\"word\":\"advancing\",\"start\":0.19,\"end\":0.77}]"} {"utterance_id": "1650-157641-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/157641/1650-157641-0006.flac", "task_type": "semantic_detection", "query": "food", "prompt": "After listening to the whole recording, output every instance of a food name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.46, 8.89, "cigar"]], "response_mode": "offline", "answer": "[{\"word\":\"cigar\",\"start\":8.46,\"end\":8.89}]"} {"utterance_id": "4831-18525-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0012.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.99, 1.4, "tom"]], "response_mode": "offline", "answer": "[{\"word\":\"tom\",\"start\":0.99,\"end\":1.4}]"} {"utterance_id": "1686-142278-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0098.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.74, 2.33, "missus hale"]], "response_mode": "offline", "answer": "[{\"word\":\"missus hale\",\"start\":1.74,\"end\":2.33}]"} {"utterance_id": "1701-141760-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0027.flac", "task_type": "lexical_detection", "query": "devil", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"devil\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.35, 2.64, "devil"]], "response_mode": "offline", "answer": "[{\"word\":\"devil\",\"start\":2.35,\"end\":2.64}]"} {"utterance_id": "5849-50964-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[0.82, 1.37, "fireman"]], "response_mode": "offline", "answer": "[{\"word\":\"fireman\",\"start\":0.82,\"end\":1.37}]"} {"utterance_id": "4515-11057-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0043.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.79, "marian"]], "response_mode": "offline", "answer": "[{\"word\":\"marian\",\"start\":0.42,\"end\":0.79}]"} {"utterance_id": "6267-65525-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0030.flac", "task_type": "lexical_counting", "query": "came", "prompt": "After the audio ends, tally every instance of the word \"came\". Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.99, 1.3, "one"], [15.49, 15.86, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.99,\"end\":1.3},{\"word\":\"two\",\"start\":15.49,\"end\":15.86}]"} {"utterance_id": "3915-98647-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0014.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.66, "body"]], "response_mode": "offline", "answer": "[{\"word\":\"body\",\"start\":3.38,\"end\":3.66}]"} {"utterance_id": "1650-167613-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0018.flac", "task_type": "lexical_detection", "query": "surprise", "prompt": "When the utterance is complete, report all instances of the word \"surprise\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.63, 2.49, "surprise"]], "response_mode": "offline", "answer": "[{\"word\":\"surprise\",\"start\":1.63,\"end\":2.49}]"} {"utterance_id": "8254-115543-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0036.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.22, 3.61, "servants"]], "response_mode": "offline", "answer": "[{\"word\":\"servants\",\"start\":3.22,\"end\":3.61}]"} {"utterance_id": "7641-96684-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0011.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.88, 6.46, "admetus"]], "response_mode": "offline", "answer": "[{\"word\":\"admetus\",\"start\":5.88,\"end\":6.46}]"} {"utterance_id": "7697-105815-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0052.flac", "task_type": "lexical_detection", "query": "name", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"name\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 1.07, "name"]], "response_mode": "offline", "answer": "[{\"word\":\"name\",\"start\":0.8,\"end\":1.07}]"} {"utterance_id": "6455-67803-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0009.flac", "task_type": "lexical_detection", "query": "theobald", "prompt": "After the audio ends, extract all occurrences of the word \"theobald\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.51, "theobald"]], "response_mode": "offline", "answer": "[{\"word\":\"theobald\",\"start\":2.01,\"end\":2.51}]"} {"utterance_id": "5849-50964-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0002.flac", "task_type": "lexical_detection", "query": "quite", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"quite\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.71, 1.0, "quite"]], "response_mode": "offline", "answer": "[{\"word\":\"quite\",\"start\":0.71,\"end\":1.0}]"} {"utterance_id": "1585-131718-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0030.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.57, 2.02, "sigmund bergmann"], [3.55, 4.1, "edison"], [8.5, 8.93, "edison"]], "response_mode": "offline", "answer": "[{\"word\":\"sigmund bergmann\",\"start\":0.57,\"end\":2.02},{\"word\":\"edison\",\"start\":3.55,\"end\":4.1},{\"word\":\"edison\",\"start\":8.5,\"end\":8.93}]"} {"utterance_id": "1701-141760-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0045.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the full audio. After it finishes, find every instance of a profession. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.14, 4.94, "aide de camp"]], "response_mode": "offline", "answer": "[{\"word\":\"aide de camp\",\"start\":4.14,\"end\":4.94}]"} {"utterance_id": "8254-84205-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0018.flac", "task_type": "lexical_detection", "query": "they're", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"they're\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.22, 0.51, "they're"]], "response_mode": "offline", "answer": "[{\"word\":\"they're\",\"start\":0.22,\"end\":0.51}]"} {"utterance_id": "6841-88291-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0055.flac", "task_type": "lexical_detection", "query": "heavy", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"heavy\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.47, 1.7, "heavy"]], "response_mode": "offline", "answer": "[{\"word\":\"heavy\",\"start\":1.47,\"end\":1.7}]"} {"utterance_id": "4572-112383-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112383/4572-112383-0009.flac", "task_type": "lexical_detection", "query": "bent", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"bent\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.54, "bent"]], "response_mode": "offline", "answer": "[{\"word\":\"bent\",\"start\":1.09,\"end\":1.54}]"} {"utterance_id": "1701-141760-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0034.flac", "task_type": "lexical_detection", "query": "remained", "prompt": "Once the audio has finished, identify every instance of the word \"remained\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.67, 1.03, "remained"]], "response_mode": "offline", "answer": "[{\"word\":\"remained\",\"start\":0.67,\"end\":1.03}]"} {"utterance_id": "7641-96670-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0006.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "When the utterance is complete, report all instances of a time of day including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.71, 5.17, "morning"]], "response_mode": "offline", "answer": "[{\"word\":\"morning\",\"start\":4.71,\"end\":5.17}]"} {"utterance_id": "700-122868-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0019.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.91, 7.27, "eyes"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":6.91,\"end\":7.27}]"} {"utterance_id": "4572-112375-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0004.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[4.81, 5.27, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":4.81,\"end\":5.27}]"} {"utterance_id": "1255-90413-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0022.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.4, 1.01, "lady iris"], [8.89, 9.93, "nichola pine avon"]], "response_mode": "offline", "answer": "[{\"word\":\"lady iris\",\"start\":0.4,\"end\":1.01},{\"word\":\"nichola pine avon\",\"start\":8.89,\"end\":9.93}]"} {"utterance_id": "700-122866-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0039.flac", "task_type": "lexical_detection", "query": "silly", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"silly\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.36, 2.76, "silly"]], "response_mode": "offline", "answer": "[{\"word\":\"silly\",\"start\":2.36,\"end\":2.76}]"} {"utterance_id": "4570-102353-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/102353/4570-102353-0001.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.39, 0.91, "several"]], "response_mode": "offline", "answer": "[{\"word\":\"several\",\"start\":0.39,\"end\":0.91}]"} {"utterance_id": "1585-131718-0043", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0043.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Do not answer until the audio ends. Then list each instance of a number as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[5.0, 5.17, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.0,\"end\":5.17}]"} {"utterance_id": "5849-50962-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.81, 2.16, "bert"]], "response_mode": "offline", "answer": "[{\"word\":\"bert\",\"start\":1.81,\"end\":2.16}]"} {"utterance_id": "116-288047-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0005.flac", "task_type": "lexical_detection", "query": "use", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"use\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.46, 1.83, "use"]], "response_mode": "offline", "answer": "[{\"word\":\"use\",\"start\":1.46,\"end\":1.83}]"} {"utterance_id": "7697-245712-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0017.flac", "task_type": "lexical_counting", "query": "man", "prompt": "Once the audio has finished, list each occurrence of the word \"man\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[3.96, 4.21, "one"], [5.54, 5.82, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.96,\"end\":4.21},{\"word\":\"two\",\"start\":5.54,\"end\":5.82}]"} {"utterance_id": "3915-98647-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0034.flac", "task_type": "lexical_detection", "query": "favorite", "prompt": "After listening to the whole recording, output every instance of the word \"favorite\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.17, "favorite"]], "response_mode": "offline", "answer": "[{\"word\":\"favorite\",\"start\":0.83,\"end\":1.17}]"} {"utterance_id": "1686-142278-0052", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0052.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[14.04, 14.48, "mind"]], "response_mode": "offline", "answer": "[{\"word\":\"mind\",\"start\":14.04,\"end\":14.48}]"} {"utterance_id": "6267-53049-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0009.flac", "task_type": "lexical_detection", "query": "beauty", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"beauty\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.17, 1.73, "beauty"]], "response_mode": "offline", "answer": "[{\"word\":\"beauty\",\"start\":1.17,\"end\":1.73}]"} {"utterance_id": "6267-65525-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0041.flac", "task_type": "lexical_detection", "query": "everything", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"everything\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.2, 0.75, "everything"]], "response_mode": "offline", "answer": "[{\"word\":\"everything\",\"start\":0.2,\"end\":0.75}]"} {"utterance_id": "4831-25894-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0015.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Process the full audio offline. Find every instance of a number and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.89, 1.13, "half"]], "response_mode": "offline", "answer": "[{\"word\":\"half\",\"start\":0.89,\"end\":1.13}]"} {"utterance_id": "4831-18525-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0006.flac", "task_type": "lexical_detection", "query": "picked", "prompt": "Process the full audio offline. Find every instance of the word \"picked\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 1.92, "picked"]], "response_mode": "offline", "answer": "[{\"word\":\"picked\",\"start\":1.67,\"end\":1.92}]"} {"utterance_id": "1585-131718-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0049.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "After listening to the whole recording, output every instance of an organization name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.48, 5.76, "harrison lamp factory"]], "response_mode": "offline", "answer": "[{\"word\":\"harrison lamp factory\",\"start\":4.48,\"end\":5.76}]"} {"utterance_id": "1630-96099-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0015.flac", "task_type": "lexical_detection", "query": "italian", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"italian\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.39, "italian"]], "response_mode": "offline", "answer": "[{\"word\":\"italian\",\"start\":2.95,\"end\":3.39}]"} {"utterance_id": "8288-274162-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0012.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.48, 2.04, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.48,\"end\":2.04}]"} {"utterance_id": "8254-115543-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0014.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.48, 0.75, "many"], [4.63, 4.87, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"many\",\"start\":0.48,\"end\":0.75},{\"word\":\"one\",\"start\":4.63,\"end\":4.87}]"} {"utterance_id": "1255-90407-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0029.flac", "task_type": "lexical_detection", "query": "moving", "prompt": "When the utterance is complete, report all instances of the word \"moving\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.11, "moving"]], "response_mode": "offline", "answer": "[{\"word\":\"moving\",\"start\":1.84,\"end\":2.11}]"} {"utterance_id": "8254-115543-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0016.flac", "task_type": "lexical_counting", "query": "dogs", "prompt": "When the utterance is complete, report how the word \"dogs\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.52, "one"], [4.72, 5.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.07,\"end\":1.52},{\"word\":\"two\",\"start\":4.72,\"end\":5.02}]"} {"utterance_id": "7697-105815-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0053.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.3, "immanuel's"]], "response_mode": "offline", "answer": "[{\"word\":\"immanuel's\",\"start\":1.74,\"end\":2.3}]"} {"utterance_id": "1650-167613-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0048.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After listening to the whole recording, output every instance of a plant name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.07, "tree"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":2.69,\"end\":3.07}]"} {"utterance_id": "7641-96684-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0013.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Do not answer until the audio ends. Then list each instance of a plant name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[8.08, 8.46, "corn"]], "response_mode": "offline", "answer": "[{\"word\":\"corn\",\"start\":8.08,\"end\":8.46}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the full audio. After it finishes, find every instance of a time of day. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.74, 9.18, "afternoon"]], "response_mode": "offline", "answer": "[{\"word\":\"afternoon\",\"start\":8.74,\"end\":9.18}]"} {"utterance_id": "3660-172182-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[2.77, 3.13, "knight"]], "response_mode": "offline", "answer": "[{\"word\":\"knight\",\"start\":2.77,\"end\":3.13}]"} {"utterance_id": "6599-38591-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0000.flac", "task_type": "semantic_detection", "query": "city", "prompt": "After listening to the whole recording, output every instance of a city name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.36, "dreux"]], "response_mode": "offline", "answer": "[{\"word\":\"dreux\",\"start\":2.01,\"end\":2.36}]"} {"utterance_id": "8254-115543-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0034.flac", "task_type": "lexical_detection", "query": "neck", "prompt": "After the audio ends, extract all occurrences of the word \"neck\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.63, "neck"]], "response_mode": "offline", "answer": "[{\"word\":\"neck\",\"start\":1.38,\"end\":1.63}]"} {"utterance_id": "1630-96099-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0010.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "When the utterance is complete, report all instances of an ordinal number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 2.98, "last"]], "response_mode": "offline", "answer": "[{\"word\":\"last\",\"start\":2.58,\"end\":2.98}]"} {"utterance_id": "3915-57461-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0029.flac", "task_type": "lexical_detection", "query": "managed", "prompt": "After listening to the whole recording, output every instance of the word \"managed\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.2, "managed"]], "response_mode": "offline", "answer": "[{\"word\":\"managed\",\"start\":0.77,\"end\":1.2}]"} {"utterance_id": "7641-96670-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.97, "ares"]], "response_mode": "offline", "answer": "[{\"word\":\"ares\",\"start\":1.59,\"end\":1.97}]"} {"utterance_id": "6123-59150-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0019.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.56, 2.99, "dickens"]], "response_mode": "offline", "answer": "[{\"word\":\"dickens\",\"start\":2.56,\"end\":2.99}]"} {"utterance_id": "7641-96670-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0023.flac", "task_type": "lexical_counting", "query": "brave", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"brave\" (word, start, end). If none, reply [].", "labels": [[6.7, 6.99, "one"], [9.95, 10.38, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.7,\"end\":6.99},{\"word\":\"two\",\"start\":9.95,\"end\":10.38}]"} {"utterance_id": "1650-167613-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.79, 2.57, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":1.79,\"end\":2.57}]"} {"utterance_id": "8254-115543-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0024.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen carefully to the entire clip, then report every instance of a profession as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.72, 2.06, "servant"], [4.68, 5.09, "rajah's"]], "response_mode": "offline", "answer": "[{\"word\":\"servant\",\"start\":1.72,\"end\":2.06},{\"word\":\"rajah's\",\"start\":4.68,\"end\":5.09}]"} {"utterance_id": "4323-18416-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0003.flac", "task_type": "lexical_detection", "query": "i'm", "prompt": "When the utterance is complete, report all instances of the word \"i'm\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.39, "i'm"]], "response_mode": "offline", "answer": "[{\"word\":\"i'm\",\"start\":0.26,\"end\":0.39}]"} {"utterance_id": "4153-186222-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.88, 2.07, "patricia doyle"]], "response_mode": "offline", "answer": "[{\"word\":\"patricia doyle\",\"start\":0.88,\"end\":2.07}]"} {"utterance_id": "4323-13259-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.77, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":1.37,\"end\":1.77}]"} {"utterance_id": "7641-96684-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0038.flac", "task_type": "lexical_detection", "query": "across", "prompt": "When the utterance is complete, report all instances of the word \"across\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.21, 3.73, "across"]], "response_mode": "offline", "answer": "[{\"word\":\"across\",\"start\":3.21,\"end\":3.73}]"} {"utterance_id": "8254-84205-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0045.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.32, "doctor"]], "response_mode": "offline", "answer": "[{\"word\":\"doctor\",\"start\":1.02,\"end\":1.32}]"} {"utterance_id": "4323-18416-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0032.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the full audio. After it finishes, find every instance of a family relation. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.33, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":1.05,\"end\":1.33}]"} {"utterance_id": "1701-141759-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0026.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.63, 1.05, "count"]], "response_mode": "offline", "answer": "[{\"word\":\"count\",\"start\":0.63,\"end\":1.05}]"} {"utterance_id": "2506-11267-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0004.flac", "task_type": "semantic_detection", "query": "color", "prompt": "You will hear a complete utterance. When it ends, list every instance of a color name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.48, "pink"]], "response_mode": "offline", "answer": "[{\"word\":\"pink\",\"start\":2.14,\"end\":2.48}]"} {"utterance_id": "4831-25894-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0035.flac", "task_type": "lexical_detection", "query": "seal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"seal\" with fields word, start, end (seconds). If none, reply [].", "labels": [[8.2, 8.56, "seal"]], "response_mode": "offline", "answer": "[{\"word\":\"seal\",\"start\":8.2,\"end\":8.56}]"} {"utterance_id": "6467-56885-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0012.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.97, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.68,\"end\":0.97}]"} {"utterance_id": "1255-90413-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0000.flac", "task_type": "lexical_detection", "query": "draws", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"draws\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.45, 0.79, "draws"]], "response_mode": "offline", "answer": "[{\"word\":\"draws\",\"start\":0.45,\"end\":0.79}]"} {"utterance_id": "1650-167613-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0042.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.89, 4.27, "eyes"], [6.69, 7.07, "lips"]], "response_mode": "offline", "answer": "[{\"word\":\"eyes\",\"start\":3.89,\"end\":4.27},{\"word\":\"lips\",\"start\":6.69,\"end\":7.07}]"} {"utterance_id": "1701-141760-0053", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0053.flac", "task_type": "lexical_detection", "query": "headquarters", "prompt": "When the utterance is complete, report all instances of the word \"headquarters\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.12, 1.81, "headquarters"]], "response_mode": "offline", "answer": "[{\"word\":\"headquarters\",\"start\":1.12,\"end\":1.81}]"} {"utterance_id": "7641-96684-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0000.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Once the audio has finished, identify every instance of a food name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.22, 2.47, "bread"]], "response_mode": "offline", "answer": "[{\"word\":\"bread\",\"start\":2.22,\"end\":2.47}]"} {"utterance_id": "7641-96670-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0024.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.93, 1.47, "jason"]], "response_mode": "offline", "answer": "[{\"word\":\"jason\",\"start\":0.93,\"end\":1.47}]"} {"utterance_id": "7601-291468-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0000.flac", "task_type": "semantic_counting", "query": "disease", "prompt": "When the utterance is complete, report how a disease name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.81, 5.32, "one"], [27.75, 28.21, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.81,\"end\":5.32},{\"word\":\"two\",\"start\":27.75,\"end\":28.21}]"} {"utterance_id": "6841-88294-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0045.flac", "task_type": "lexical_detection", "query": "holding", "prompt": "When the utterance is complete, report all instances of the word \"holding\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.42, 4.74, "holding"]], "response_mode": "offline", "answer": "[{\"word\":\"holding\",\"start\":4.42,\"end\":4.74}]"} {"utterance_id": "7601-175351-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0021.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.34, 1.54, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.34,\"end\":1.54}]"} {"utterance_id": "4323-55228-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0032.flac", "task_type": "lexical_detection", "query": "bertram", "prompt": "After the audio ends, extract all occurrences of the word \"bertram\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.35, 0.68, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"bertram\",\"start\":0.35,\"end\":0.68}]"} {"utterance_id": "116-288048-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0028.flac", "task_type": "lexical_detection", "query": "promises", "prompt": "After the audio ends, extract all occurrences of the word \"promises\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.56, 4.69, "promises"]], "response_mode": "offline", "answer": "[{\"word\":\"promises\",\"start\":3.56,\"end\":4.69}]"} {"utterance_id": "4570-56594-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0000.flac", "task_type": "lexical_detection", "query": "see", "prompt": "Process the full audio offline. Find every instance of the word \"see\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.61, "see"]], "response_mode": "offline", "answer": "[{\"word\":\"see\",\"start\":4.32,\"end\":4.61}]"} {"utterance_id": "3915-57461-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0016.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.69, 5.91, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":5.69,\"end\":5.91}]"} {"utterance_id": "3915-98647-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0013.flac", "task_type": "lexical_detection", "query": "provisions", "prompt": "After the audio ends, extract all occurrences of the word \"provisions\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.5, 4.35, "provisions"]], "response_mode": "offline", "answer": "[{\"word\":\"provisions\",\"start\":3.5,\"end\":4.35}]"} {"utterance_id": "4515-11057-0098", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0098.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[5.11, 5.57, "six"]], "response_mode": "offline", "answer": "[{\"word\":\"six\",\"start\":5.11,\"end\":5.57}]"} {"utterance_id": "116-288045-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0030.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, produce a running count of a body part: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.4, 1.78, "one"], [2.81, 3.2, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.4,\"end\":1.78},{\"word\":\"two\",\"start\":2.81,\"end\":3.2}]"} {"utterance_id": "4515-11057-0103", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0103.flac", "task_type": "lexical_detection", "query": "home", "prompt": "After the audio ends, extract all occurrences of the word \"home\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.77, 1.11, "home"]], "response_mode": "offline", "answer": "[{\"word\":\"home\",\"start\":0.77,\"end\":1.11}]"} {"utterance_id": "1701-141759-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0013.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the entire clip, then report the count sequence for a person's name as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.48, 0.96, "one"], [2.9, 3.5, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.48,\"end\":0.96},{\"word\":\"two\",\"start\":2.9,\"end\":3.5}]"} {"utterance_id": "700-122866-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0010.flac", "task_type": "lexical_detection", "query": "comfort", "prompt": "When the utterance is complete, report all instances of the word \"comfort\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.94, 1.34, "comfort"]], "response_mode": "offline", "answer": "[{\"word\":\"comfort\",\"start\":0.94,\"end\":1.34}]"} {"utterance_id": "116-288048-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0004.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.33, 2.86, "jesus"]], "response_mode": "offline", "answer": "[{\"word\":\"jesus\",\"start\":2.33,\"end\":2.86}]"} {"utterance_id": "7697-105817-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0001.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.67, "judge"]], "response_mode": "offline", "answer": "[{\"word\":\"judge\",\"start\":0.39,\"end\":0.67}]"} {"utterance_id": "4515-11057-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0014.flac", "task_type": "lexical_detection", "query": "amy", "prompt": "After the audio ends, extract all occurrences of the word \"amy\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.05, "amy"]], "response_mode": "offline", "answer": "[{\"word\":\"amy\",\"start\":0.72,\"end\":1.05}]"} {"utterance_id": "1255-90407-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0012.flac", "task_type": "lexical_detection", "query": "scolding", "prompt": "Once the audio has finished, identify every instance of the word \"scolding\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.02, 4.51, "scolding"]], "response_mode": "offline", "answer": "[{\"word\":\"scolding\",\"start\":4.02,\"end\":4.51}]"} {"utterance_id": "116-288045-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0001.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.99, 6.24, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":5.99,\"end\":6.24}]"} {"utterance_id": "6123-59150-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0005.flac", "task_type": "lexical_detection", "query": "conflict", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"conflict\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.14, 3.88, "conflict"]], "response_mode": "offline", "answer": "[{\"word\":\"conflict\",\"start\":3.14,\"end\":3.88}]"} {"utterance_id": "6455-67803-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0018.flac", "task_type": "lexical_detection", "query": "afraid", "prompt": "After the audio ends, extract all occurrences of the word \"afraid\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.56, 1.84, "afraid"]], "response_mode": "offline", "answer": "[{\"word\":\"afraid\",\"start\":1.56,\"end\":1.84}]"} {"utterance_id": "6455-67804-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0004.flac", "task_type": "lexical_detection", "query": "fool", "prompt": "Once the audio has finished, identify every instance of the word \"fool\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.97, 2.35, "fool"]], "response_mode": "offline", "answer": "[{\"word\":\"fool\",\"start\":1.97,\"end\":2.35}]"} {"utterance_id": "4831-29134-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.84, 2.35, "governor"]], "response_mode": "offline", "answer": "[{\"word\":\"governor\",\"start\":1.84,\"end\":2.35}]"} {"utterance_id": "4323-13259-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0022.flac", "task_type": "lexical_detection", "query": "maxwell", "prompt": "After the audio ends, extract all occurrences of the word \"maxwell\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.72, 1.11, "maxwell"]], "response_mode": "offline", "answer": "[{\"word\":\"maxwell\",\"start\":0.72,\"end\":1.11}]"} {"utterance_id": "7601-175351-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0009.flac", "task_type": "lexical_counting", "query": "shan't", "prompt": "Listen to the entire clip, then report the count sequence for the word \"shan't\" as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.43, 0.79, "one"], [4.04, 4.54, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.43,\"end\":0.79},{\"word\":\"two\",\"start\":4.04,\"end\":4.54}]"} {"utterance_id": "5849-50873-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0022.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Once the audio has finished, identify every instance of a profession. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.77, 4.65, "secretaries"]], "response_mode": "offline", "answer": "[{\"word\":\"secretaries\",\"start\":3.77,\"end\":4.65}]"} {"utterance_id": "3915-57461-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0014.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.2, 2.55, "ghost"]], "response_mode": "offline", "answer": "[{\"word\":\"ghost\",\"start\":2.2,\"end\":2.55}]"} {"utterance_id": "7601-101619-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101619/7601-101619-0004.flac", "task_type": "lexical_detection", "query": "successful", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"successful\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.97, 8.87, "successful"]], "response_mode": "offline", "answer": "[{\"word\":\"successful\",\"start\":7.97,\"end\":8.87}]"} {"utterance_id": "7641-96252-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0021.flac", "task_type": "lexical_counting", "query": "ourselves", "prompt": "Once the audio has finished, list each occurrence of the word \"ourselves\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.46, 1.94, "one"], [12.7, 13.13, "two"], [14.82, 15.36, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.46,\"end\":1.94},{\"word\":\"two\",\"start\":12.7,\"end\":13.13},{\"word\":\"three\",\"start\":14.82,\"end\":15.36}]"} {"utterance_id": "4153-186223-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0000.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[0.68, 1.01, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.68,\"end\":1.01}]"} {"utterance_id": "8173-294714-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0018.flac", "task_type": "lexical_detection", "query": "tie", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"tie\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.4, 2.67, "tie"]], "response_mode": "offline", "answer": "[{\"word\":\"tie\",\"start\":2.4,\"end\":2.67}]"} {"utterance_id": "4572-112381-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0010.flac", "task_type": "lexical_detection", "query": "roar", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"roar\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.62, 2.03, "roar"]], "response_mode": "offline", "answer": "[{\"word\":\"roar\",\"start\":1.62,\"end\":2.03}]"} {"utterance_id": "6267-65525-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0024.flac", "task_type": "lexical_detection", "query": "talking", "prompt": "When the utterance is complete, report all instances of the word \"talking\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.38, 1.82, "talking"]], "response_mode": "offline", "answer": "[{\"word\":\"talking\",\"start\":1.38,\"end\":1.82}]"} {"utterance_id": "6123-59186-0029", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0029.flac", "task_type": "lexical_detection", "query": "loves", "prompt": "When the utterance is complete, report all instances of the word \"loves\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.93, 3.45, "loves"]], "response_mode": "offline", "answer": "[{\"word\":\"loves\",\"start\":2.93,\"end\":3.45}]"} {"utterance_id": "5849-50873-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0008.flac", "task_type": "lexical_detection", "query": "turns", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"turns\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.26, 3.74, "turns"]], "response_mode": "offline", "answer": "[{\"word\":\"turns\",\"start\":3.26,\"end\":3.74}]"} {"utterance_id": "4572-112381-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0015.flac", "task_type": "lexical_detection", "query": "buy", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"buy\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.22, 11.93, "buy"]], "response_mode": "offline", "answer": "[{\"word\":\"buy\",\"start\":11.22,\"end\":11.93}]"} {"utterance_id": "7601-175351-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0009.flac", "task_type": "lexical_detection", "query": "shan't", "prompt": "When the utterance is complete, report all instances of the word \"shan't\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.43, 0.79, "shan't"], [4.04, 4.54, "shan't"]], "response_mode": "offline", "answer": "[{\"word\":\"shan't\",\"start\":0.43,\"end\":0.79},{\"word\":\"shan't\",\"start\":4.04,\"end\":4.54}]"} {"utterance_id": "4570-56594-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0001.flac", "task_type": "lexical_detection", "query": "wishing", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"wishing\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.74, 1.3, "wishing"]], "response_mode": "offline", "answer": "[{\"word\":\"wishing\",\"start\":0.74,\"end\":1.3}]"} {"utterance_id": "4831-25894-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0027.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "After listening to the whole recording, output every instance of a body part with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.32, 4.76, "hands"]], "response_mode": "offline", "answer": "[{\"word\":\"hands\",\"start\":4.32,\"end\":4.76}]"} {"utterance_id": "6455-67803-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0000.flac", "task_type": "lexical_detection", "query": "chapter", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"chapter\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.22, 0.62, "chapter"]], "response_mode": "offline", "answer": "[{\"word\":\"chapter\",\"start\":0.22,\"end\":0.62}]"} {"utterance_id": "7697-245712-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0008.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.48, 2.78, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":2.48,\"end\":2.78}]"} {"utterance_id": "700-122867-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0038.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the full audio. After it finishes, find every instance of a color name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.34, 3.66, "black"]], "response_mode": "offline", "answer": "[{\"word\":\"black\",\"start\":3.34,\"end\":3.66}]"} {"utterance_id": "2506-11267-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0011.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen carefully to the entire clip, then report every instance of a city name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.7, "houndsditch"]], "response_mode": "offline", "answer": "[{\"word\":\"houndsditch\",\"start\":1.09,\"end\":1.7}]"} {"utterance_id": "4831-18525-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0015.flac", "task_type": "lexical_detection", "query": "fingers", "prompt": "After the audio ends, extract all occurrences of the word \"fingers\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.64, 9.02, "fingers"]], "response_mode": "offline", "answer": "[{\"word\":\"fingers\",\"start\":8.64,\"end\":9.02}]"} {"utterance_id": "6841-88294-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0008.flac", "task_type": "lexical_detection", "query": "shape", "prompt": "Once the audio has finished, identify every instance of the word \"shape\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.13, 2.46, "shape"]], "response_mode": "offline", "answer": "[{\"word\":\"shape\",\"start\":2.13,\"end\":2.46}]"} {"utterance_id": "6123-59150-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.34, 0.96, "dostoevsky"]], "response_mode": "offline", "answer": "[{\"word\":\"dostoevsky\",\"start\":0.34,\"end\":0.96}]"} {"utterance_id": "7601-175351-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0008.flac", "task_type": "lexical_counting", "query": "say", "prompt": "Once the audio has finished, list each occurrence of the word \"say\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[0.75, 1.47, "one"], [4.45, 4.91, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.75,\"end\":1.47},{\"word\":\"two\",\"start\":4.45,\"end\":4.91}]"} {"utterance_id": "3660-172182-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0038.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.62, 4.03, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":3.62,\"end\":4.03}]"} {"utterance_id": "2506-13150-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0001.flac", "task_type": "lexical_counting", "query": "sick", "prompt": "Once the audio has finished, list each occurrence of the word \"sick\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.57, 2.79, "one"], [7.62, 7.9, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.57,\"end\":2.79},{\"word\":\"two\",\"start\":7.62,\"end\":7.9}]"} {"utterance_id": "1651-136854-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0016.flac", "task_type": "lexical_detection", "query": "hinders", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"hinders\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.55, 2.09, "hinders"]], "response_mode": "offline", "answer": "[{\"word\":\"hinders\",\"start\":1.55,\"end\":2.09}]"} {"utterance_id": "6267-65525-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0013.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Do not answer until the audio ends. Then list each instance of a profession as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[6.77, 7.37, "editor"]], "response_mode": "offline", "answer": "[{\"word\":\"editor\",\"start\":6.77,\"end\":7.37}]"} {"utterance_id": "6467-56885-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0000.flac", "task_type": "lexical_detection", "query": "aim", "prompt": "Process the full audio offline. Find every instance of the word \"aim\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.86, 6.12, "aim"]], "response_mode": "offline", "answer": "[{\"word\":\"aim\",\"start\":5.86,\"end\":6.12}]"} {"utterance_id": "4831-29134-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0005.flac", "task_type": "lexical_detection", "query": "yes", "prompt": "When the utterance is complete, report all instances of the word \"yes\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.63, "yes"]], "response_mode": "offline", "answer": "[{\"word\":\"yes\",\"start\":1.26,\"end\":1.63}]"} {"utterance_id": "7641-96684-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Process the full audio offline. Find every instance of a body part and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.0, 3.32, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"face\",\"start\":3.0,\"end\":3.32}]"} {"utterance_id": "1651-136854-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0026.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.26, 1.58, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.26,\"end\":1.58}]"} {"utterance_id": "6267-53049-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0002.flac", "task_type": "lexical_detection", "query": "doris", "prompt": "After the audio ends, extract all occurrences of the word \"doris\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.26, 0.78, "doris"]], "response_mode": "offline", "answer": "[{\"word\":\"doris\",\"start\":0.26,\"end\":0.78}]"} {"utterance_id": "7601-291468-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0005.flac", "task_type": "lexical_detection", "query": "passion", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"passion\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.5, 8.19, "passion"]], "response_mode": "offline", "answer": "[{\"word\":\"passion\",\"start\":7.5,\"end\":8.19}]"} {"utterance_id": "8254-84205-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0066.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.06, 2.4, "chris's"]], "response_mode": "offline", "answer": "[{\"word\":\"chris's\",\"start\":2.06,\"end\":2.4}]"} {"utterance_id": "3660-172182-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.79, 1.27, "geraint"]], "response_mode": "offline", "answer": "[{\"word\":\"geraint\",\"start\":0.79,\"end\":1.27}]"} {"utterance_id": "8288-274162-0054", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0054.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.56, 2.76, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":2.56,\"end\":2.76}]"} {"utterance_id": "7641-96670-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0021.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "After the audio ends, extract all occurrences of a plant name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.42, "seed"]], "response_mode": "offline", "answer": "[{\"word\":\"seed\",\"start\":2.07,\"end\":2.42}]"} {"utterance_id": "8288-274162-0058", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0058.flac", "task_type": "lexical_detection", "query": "evil", "prompt": "After the audio ends, extract all occurrences of the word \"evil\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.94, 5.26, "evil"]], "response_mode": "offline", "answer": "[{\"word\":\"evil\",\"start\":4.94,\"end\":5.26}]"} {"utterance_id": "700-122867-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 0.77, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":0.52,\"end\":0.77}]"} {"utterance_id": "3660-6517-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0023.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "After listening to the whole recording, output every instance of a weather term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.64, 1.83, "wind"]], "response_mode": "offline", "answer": "[{\"word\":\"wind\",\"start\":1.64,\"end\":1.83}]"} {"utterance_id": "3660-6517-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0008.flac", "task_type": "semantic_detection", "query": "weather", "prompt": "You will hear a complete utterance. When it ends, list every instance of a weather term you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.2, 6.59, "weather"]], "response_mode": "offline", "answer": "[{\"word\":\"weather\",\"start\":6.2,\"end\":6.59}]"} {"utterance_id": "5849-50962-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0016.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.76, 6.07, "arms"]], "response_mode": "offline", "answer": "[{\"word\":\"arms\",\"start\":5.76,\"end\":6.07}]"} {"utterance_id": "700-122867-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0030.flac", "task_type": "lexical_detection", "query": "curly", "prompt": "After listening to the whole recording, output every instance of the word \"curly\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.06, 8.48, "curly"]], "response_mode": "offline", "answer": "[{\"word\":\"curly\",\"start\":8.06,\"end\":8.48}]"} {"utterance_id": "4153-185072-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/185072/4153-185072-0013.flac", "task_type": "lexical_detection", "query": "unsatisfied", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"unsatisfied\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.89, 1.83, "unsatisfied"]], "response_mode": "offline", "answer": "[{\"word\":\"unsatisfied\",\"start\":0.89,\"end\":1.83}]"} {"utterance_id": "1650-173552-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173552/1650-173552-0008.flac", "task_type": "lexical_detection", "query": "gentlemen", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"gentlemen\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.58, 4.09, "gentlemen"], [5.54, 6.11, "gentlemen"]], "response_mode": "offline", "answer": "[{\"word\":\"gentlemen\",\"start\":3.58,\"end\":4.09},{\"word\":\"gentlemen\",\"start\":5.54,\"end\":6.11}]"} {"utterance_id": "1650-167613-0055", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0055.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen carefully to the entire clip, then report every instance of a number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.41, 0.7, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.41,\"end\":0.7}]"} {"utterance_id": "700-122866-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0016.flac", "task_type": "lexical_detection", "query": "diana", "prompt": "After the audio ends, extract all occurrences of the word \"diana\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.7, 1.28, "diana"]], "response_mode": "offline", "answer": "[{\"word\":\"diana\",\"start\":0.7,\"end\":1.28}]"} {"utterance_id": "3915-57461-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0020.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Listen carefully to the entire clip, then report every instance of a disease name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.69, 6.29, "ailment"]], "response_mode": "offline", "answer": "[{\"word\":\"ailment\",\"start\":5.69,\"end\":6.29}]"} {"utterance_id": "1630-96099-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0012.flac", "task_type": "lexical_detection", "query": "excuse", "prompt": "After listening to the whole recording, output every instance of the word \"excuse\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.28, 0.82, "excuse"]], "response_mode": "offline", "answer": "[{\"word\":\"excuse\",\"start\":0.28,\"end\":0.82}]"} {"utterance_id": "4515-11057-0111", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0111.flac", "task_type": "lexical_detection", "query": "unworthy", "prompt": "Process the full audio offline. Find every instance of the word \"unworthy\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.46, 6.1, "unworthy"]], "response_mode": "offline", "answer": "[{\"word\":\"unworthy\",\"start\":5.46,\"end\":6.1}]"} {"utterance_id": "1630-141772-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0020.flac", "task_type": "lexical_detection", "query": "moved", "prompt": "When the utterance is complete, report all instances of the word \"moved\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.29, 1.61, "moved"]], "response_mode": "offline", "answer": "[{\"word\":\"moved\",\"start\":1.29,\"end\":1.61}]"} {"utterance_id": "1255-90413-0020", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0020.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.53, 1.17, "missus pine avon"]], "response_mode": "offline", "answer": "[{\"word\":\"missus pine avon\",\"start\":0.53,\"end\":1.17}]"} {"utterance_id": "700-122867-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0017.flac", "task_type": "lexical_detection", "query": "proof", "prompt": "After the audio ends, extract all occurrences of the word \"proof\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.44, 5.8, "proof"]], "response_mode": "offline", "answer": "[{\"word\":\"proof\",\"start\":5.44,\"end\":5.8}]"} {"utterance_id": "4323-18416-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0034.flac", "task_type": "lexical_counting", "query": "i've", "prompt": "When the utterance is complete, report how the word \"i've\" accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.18, 0.37, "one"], [2.98, 3.09, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.18,\"end\":0.37},{\"word\":\"two\",\"start\":2.98,\"end\":3.09}]"} {"utterance_id": "8254-115543-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0006.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "When the utterance is complete, report all instances of a profession including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.47, 0.84, "cook"]], "response_mode": "offline", "answer": "[{\"word\":\"cook\",\"start\":0.47,\"end\":0.84}]"} {"utterance_id": "1585-157660-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0003.flac", "task_type": "lexical_detection", "query": "london", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"london\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.67, "london"]], "response_mode": "offline", "answer": "[{\"word\":\"london\",\"start\":1.14,\"end\":1.67}]"} {"utterance_id": "5849-50873-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0000.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Once the audio has finished, list each occurrence of a profession as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[1.47, 1.98, "one"], [7.62, 8.06, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.47,\"end\":1.98},{\"word\":\"two\",\"start\":7.62,\"end\":8.06}]"} {"utterance_id": "6267-53049-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0005.flac", "task_type": "lexical_detection", "query": "hunter", "prompt": "When the utterance is complete, report all instances of the word \"hunter\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.6, 1.24, "hunter"]], "response_mode": "offline", "answer": "[{\"word\":\"hunter\",\"start\":0.6,\"end\":1.24}]"} {"utterance_id": "7601-291468-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0004.flac", "task_type": "lexical_counting", "query": "ancient", "prompt": "You will hear a complete utterance. When it ends, produce a running count of the word \"ancient\": each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[0.53, 1.0, "one"], [15.91, 16.39, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.53,\"end\":1.0},{\"word\":\"two\",\"start\":15.91,\"end\":16.39}]"} {"utterance_id": "4831-29134-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0015.flac", "task_type": "lexical_detection", "query": "senator", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"senator\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.54, 1.03, "senator"]], "response_mode": "offline", "answer": "[{\"word\":\"senator\",\"start\":0.54,\"end\":1.03}]"} {"utterance_id": "1585-157660-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0009.flac", "task_type": "lexical_detection", "query": "won", "prompt": "After listening to the whole recording, output every instance of the word \"won\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.14, 1.4, "won"]], "response_mode": "offline", "answer": "[{\"word\":\"won\",\"start\":1.14,\"end\":1.4}]"} {"utterance_id": "6467-94831-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0016.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Do not answer until the audio ends. Then list each instance of a food name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.95, 2.43, "pie"]], "response_mode": "offline", "answer": "[{\"word\":\"pie\",\"start\":1.95,\"end\":2.43}]"} {"utterance_id": "1630-96099-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0021.flac", "task_type": "semantic_counting", "query": "cardinal", "prompt": "After the audio ends, tally every instance of a number. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[1.25, 1.91, "one"], [3.48, 4.11, "two"], [7.26, 7.83, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.25,\"end\":1.91},{\"word\":\"two\",\"start\":3.48,\"end\":4.11},{\"word\":\"three\",\"start\":7.26,\"end\":7.83}]"} {"utterance_id": "6841-88291-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0012.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.66, 5.14, "fish"]], "response_mode": "offline", "answer": "[{\"word\":\"fish\",\"start\":4.66,\"end\":5.14}]"} {"utterance_id": "3663-172528-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0044.flac", "task_type": "lexical_detection", "query": "pope", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"pope\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.86, 5.3, "pope"]], "response_mode": "offline", "answer": "[{\"word\":\"pope\",\"start\":4.86,\"end\":5.3}]"} {"utterance_id": "116-288047-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0015.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a body part with fields word, start, end (seconds). If none, reply [].", "labels": [[1.2, 1.57, "womb"]], "response_mode": "offline", "answer": "[{\"word\":\"womb\",\"start\":1.2,\"end\":1.57}]"} {"utterance_id": "1630-73710-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0000.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Do not answer until the audio ends. Then list each instance of a body part as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.05, 1.36, "foot"], [12.35, 12.91, "voice"], [16.61, 17.18, "brain"]], "response_mode": "offline", "answer": "[{\"word\":\"foot\",\"start\":1.05,\"end\":1.36},{\"word\":\"voice\",\"start\":12.35,\"end\":12.91},{\"word\":\"brain\",\"start\":16.61,\"end\":17.18}]"} {"utterance_id": "8173-294714-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0008.flac", "task_type": "lexical_counting", "query": "know", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"know\" (word, start, end). If none, reply [].", "labels": [[0.33, 0.66, "one"], [4.28, 4.51, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.33,\"end\":0.66},{\"word\":\"two\",\"start\":4.28,\"end\":4.51}]"} {"utterance_id": "3915-98647-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0009.flac", "task_type": "lexical_detection", "query": "depends", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"depends\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.66, 8.39, "depends"]], "response_mode": "offline", "answer": "[{\"word\":\"depends\",\"start\":7.66,\"end\":8.39}]"} {"utterance_id": "5543-27761-0068", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0068.flac", "task_type": "lexical_detection", "query": "suffered", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"suffered\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.57, 1.06, "suffered"]], "response_mode": "offline", "answer": "[{\"word\":\"suffered\",\"start\":0.57,\"end\":1.06}]"} {"utterance_id": "1585-131718-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0025.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.8, "one"], [5.96, 6.27, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.5,\"end\":0.8},{\"word\":\"one\",\"start\":5.96,\"end\":6.27}]"} {"utterance_id": "4323-18416-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0002.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Do not answer until the audio ends. Then list each instance of a color name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.45, 0.67, "red"], [2.07, 2.55, "pale blue"]], "response_mode": "offline", "answer": "[{\"word\":\"red\",\"start\":0.45,\"end\":0.67},{\"word\":\"pale blue\",\"start\":2.07,\"end\":2.55}]"} {"utterance_id": "7697-105815-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0035.flac", "task_type": "lexical_detection", "query": "prey", "prompt": "Once the audio has finished, identify every instance of the word \"prey\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.98, 4.41, "prey"]], "response_mode": "offline", "answer": "[{\"word\":\"prey\",\"start\":3.98,\"end\":4.41}]"} {"utterance_id": "8254-115543-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0025.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.48, 4.65, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.48,\"end\":4.65}]"} {"utterance_id": "700-122867-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0004.flac", "task_type": "lexical_detection", "query": "bringing", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"bringing\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.82, 3.07, "bringing"]], "response_mode": "offline", "answer": "[{\"word\":\"bringing\",\"start\":2.82,\"end\":3.07}]"} {"utterance_id": "6841-88291-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0008.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.27, 0.44, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.27,\"end\":0.44}]"} {"utterance_id": "5849-50963-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.49, 4.69, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.49,\"end\":4.69}]"} {"utterance_id": "7641-96670-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0025.flac", "task_type": "lexical_detection", "query": "dare", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"dare\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.05, "dare"]], "response_mode": "offline", "answer": "[{\"word\":\"dare\",\"start\":0.76,\"end\":1.05}]"} {"utterance_id": "1255-74899-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0018.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a color name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.77, 1.02, "blue"]], "response_mode": "offline", "answer": "[{\"word\":\"blue\",\"start\":0.77,\"end\":1.02}]"} {"utterance_id": "1650-173551-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/173551/1650-173551-0005.flac", "task_type": "lexical_detection", "query": "behaved", "prompt": "After listening to the whole recording, output every instance of the word \"behaved\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.94, 3.35, "behaved"]], "response_mode": "offline", "answer": "[{\"word\":\"behaved\",\"start\":2.94,\"end\":3.35}]"} {"utterance_id": "3660-6517-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0016.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a time of day with fields word, start, end (seconds). If none, reply [].", "labels": [[2.91, 3.3, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":2.91,\"end\":3.3}]"} {"utterance_id": "6599-38591-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0007.flac", "task_type": "lexical_detection", "query": "montmorency", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"montmorency\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.9, 3.64, "montmorency"]], "response_mode": "offline", "answer": "[{\"word\":\"montmorency\",\"start\":2.9,\"end\":3.64}]"} {"utterance_id": "1686-142278-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0000.flac", "task_type": "lexical_detection", "query": "gone", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"gone\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.85, 1.25, "gone"]], "response_mode": "offline", "answer": "[{\"word\":\"gone\",\"start\":0.85,\"end\":1.25}]"} {"utterance_id": "3663-172528-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0015.flac", "task_type": "lexical_detection", "query": "bleed", "prompt": "After the audio ends, extract all occurrences of the word \"bleed\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.89, 7.35, "bleed"]], "response_mode": "offline", "answer": "[{\"word\":\"bleed\",\"start\":6.89,\"end\":7.35}]"} {"utterance_id": "1255-74899-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0016.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.67, 5.15, "husband"]], "response_mode": "offline", "answer": "[{\"word\":\"husband\",\"start\":4.67,\"end\":5.15}]"} {"utterance_id": "3660-172182-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172182/3660-172182-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.7, 2.2, "guenever"]], "response_mode": "offline", "answer": "[{\"word\":\"guenever\",\"start\":1.7,\"end\":2.2}]"} {"utterance_id": "1686-142278-0086", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0086.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.88, 2.13, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.88,\"end\":2.13}]"} {"utterance_id": "1585-131718-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0002.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After listening to the whole recording, output every instance of a money or currency term with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.23, 1.68, "capital"]], "response_mode": "offline", "answer": "[{\"word\":\"capital\",\"start\":1.23,\"end\":1.68}]"} {"utterance_id": "1630-141772-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0004.flac", "task_type": "lexical_detection", "query": "dressing", "prompt": "After listening to the whole recording, output every instance of the word \"dressing\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.47, 2.85, "dressing"]], "response_mode": "offline", "answer": "[{\"word\":\"dressing\",\"start\":2.47,\"end\":2.85}]"} {"utterance_id": "700-122866-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0035.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.67, 2.29, "mister allan"]], "response_mode": "offline", "answer": "[{\"word\":\"mister allan\",\"start\":1.67,\"end\":2.29}]"} {"utterance_id": "1650-167613-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0051.flac", "task_type": "lexical_detection", "query": "one's", "prompt": "Process the full audio offline. Find every instance of the word \"one's\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.95, 3.38, "one's"]], "response_mode": "offline", "answer": "[{\"word\":\"one's\",\"start\":2.95,\"end\":3.38}]"} {"utterance_id": "8288-274162-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0044.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.88, "raoul"]], "response_mode": "offline", "answer": "[{\"word\":\"raoul\",\"start\":0.5,\"end\":0.88}]"} {"utterance_id": "4572-112381-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0001.flac", "task_type": "semantic_detection", "query": "food", "prompt": "Listen to the full audio. After it finishes, find every instance of a food name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.07, 5.62, "fare"]], "response_mode": "offline", "answer": "[{\"word\":\"fare\",\"start\":5.07,\"end\":5.62}]"} {"utterance_id": "6123-59186-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.99, 2.67, "mister hardy's"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy's\",\"start\":1.99,\"end\":2.67}]"} {"utterance_id": "4572-112381-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0004.flac", "task_type": "lexical_detection", "query": "works", "prompt": "Once the audio has finished, identify every instance of the word \"works\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.21, 1.67, "works"]], "response_mode": "offline", "answer": "[{\"word\":\"works\",\"start\":1.21,\"end\":1.67}]"} {"utterance_id": "4515-11057-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0007.flac", "task_type": "lexical_detection", "query": "let", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"let\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.8, 0.97, "let"]], "response_mode": "offline", "answer": "[{\"word\":\"let\",\"start\":0.8,\"end\":0.97}]"} {"utterance_id": "8173-294714-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0046.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Once the audio has finished, identify every instance of a number. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[3.36, 3.63, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.36,\"end\":3.63}]"} {"utterance_id": "1701-141760-0051", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0051.flac", "task_type": "lexical_detection", "query": "angry", "prompt": "Process the full audio offline. Find every instance of the word \"angry\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.16, 1.55, "angry"]], "response_mode": "offline", "answer": "[{\"word\":\"angry\",\"start\":1.16,\"end\":1.55}]"} {"utterance_id": "7601-175351-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0014.flac", "task_type": "lexical_counting", "query": "together", "prompt": "After listening to the whole recording, output timed count words for the word \"together\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[10.31, 10.85, "one"], [28.71, 29.41, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":10.31,\"end\":10.85},{\"word\":\"two\",\"start\":28.71,\"end\":29.41}]"} {"utterance_id": "6841-88291-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0034.flac", "task_type": "lexical_detection", "query": "gorgeous", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"gorgeous\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.14, 2.55, "gorgeous"]], "response_mode": "offline", "answer": "[{\"word\":\"gorgeous\",\"start\":2.14,\"end\":2.55}]"} {"utterance_id": "700-122866-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0038.flac", "task_type": "lexical_detection", "query": "barry", "prompt": "When the utterance is complete, report all instances of the word \"barry\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 1.78, "barry"]], "response_mode": "offline", "answer": "[{\"word\":\"barry\",\"start\":1.42,\"end\":1.78}]"} {"utterance_id": "3660-6517-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0010.flac", "task_type": "lexical_detection", "query": "carcases", "prompt": "When the utterance is complete, report all instances of the word \"carcases\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.0, 3.57, "carcases"], [4.58, 5.13, "carcases"]], "response_mode": "offline", "answer": "[{\"word\":\"carcases\",\"start\":3.0,\"end\":3.57},{\"word\":\"carcases\",\"start\":4.58,\"end\":5.13}]"} {"utterance_id": "5849-50962-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0024.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.19, 1.51, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.19,\"end\":1.51}]"} {"utterance_id": "700-122868-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0015.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Do not answer until the audio ends. Then list each instance of a person's name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.51, 0.74, "anne"]], "response_mode": "offline", "answer": "[{\"word\":\"anne\",\"start\":0.51,\"end\":0.74}]"} {"utterance_id": "1630-102884-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/102884/1630-102884-0000.flac", "task_type": "lexical_detection", "query": "travels", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"travels\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.42, 1.99, "travels"]], "response_mode": "offline", "answer": "[{\"word\":\"travels\",\"start\":1.42,\"end\":1.99}]"} {"utterance_id": "4572-112375-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0008.flac", "task_type": "lexical_detection", "query": "lasting", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"lasting\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[4.63, 5.17, "lasting"]], "response_mode": "offline", "answer": "[{\"word\":\"lasting\",\"start\":4.63,\"end\":5.17}]"} {"utterance_id": "4153-186222-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0031.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.44, 7.83, "jane"]], "response_mode": "offline", "answer": "[{\"word\":\"jane\",\"start\":7.44,\"end\":7.83}]"} {"utterance_id": "8254-115543-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0016.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, produce a running count of an animal name: each event is {\"word\": \"one\"|\"two\"|..., \"start\", \"end\"}. If none, reply [].", "labels": [[1.07, 1.52, "one"], [4.72, 5.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.07,\"end\":1.52},{\"word\":\"two\",\"start\":4.72,\"end\":5.02}]"} {"utterance_id": "5543-27761-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0031.flac", "task_type": "lexical_detection", "query": "smile", "prompt": "When the utterance is complete, report all instances of the word \"smile\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.82, 6.34, "smile"]], "response_mode": "offline", "answer": "[{\"word\":\"smile\",\"start\":5.82,\"end\":6.34}]"} {"utterance_id": "8254-115543-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0015.flac", "task_type": "semantic_detection", "query": "food", "prompt": "You will hear a complete utterance. When it ends, list every instance of a food name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.33, 1.78, "lunch"]], "response_mode": "offline", "answer": "[{\"word\":\"lunch\",\"start\":1.33,\"end\":1.78}]"} {"utterance_id": "116-288048-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0015.flac", "task_type": "lexical_detection", "query": "presence", "prompt": "When the utterance is complete, report all instances of the word \"presence\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.69, "presence"]], "response_mode": "offline", "answer": "[{\"word\":\"presence\",\"start\":1.27,\"end\":1.69}]"} {"utterance_id": "3660-172183-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0018.flac", "task_type": "lexical_detection", "query": "thereupon", "prompt": "Once the audio has finished, identify every instance of the word \"thereupon\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[5.17, 5.6, "thereupon"]], "response_mode": "offline", "answer": "[{\"word\":\"thereupon\",\"start\":5.17,\"end\":5.6}]"} {"utterance_id": "5849-50873-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0037.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Listen to the full audio. After it finishes, find every instance of a plant name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.62, 0.9, "tree"], [5.14, 5.83, "charter oak"]], "response_mode": "offline", "answer": "[{\"word\":\"tree\",\"start\":0.62,\"end\":0.9},{\"word\":\"charter oak\",\"start\":5.14,\"end\":5.83}]"} {"utterance_id": "6841-88294-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88294/6841-88294-0014.flac", "task_type": "lexical_detection", "query": "important", "prompt": "Process the full audio offline. Find every instance of the word \"important\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.54, 4.05, "important"]], "response_mode": "offline", "answer": "[{\"word\":\"important\",\"start\":3.54,\"end\":4.05}]"} {"utterance_id": "8254-84205-0038", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0038.flac", "task_type": "lexical_detection", "query": "trap", "prompt": "When the utterance is complete, report all instances of the word \"trap\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.39, 5.79, "trap"]], "response_mode": "offline", "answer": "[{\"word\":\"trap\",\"start\":5.39,\"end\":5.79}]"} {"utterance_id": "116-288045-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0002.flac", "task_type": "lexical_detection", "query": "realized", "prompt": "Once the audio has finished, identify every instance of the word \"realized\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.94, 1.4, "realized"]], "response_mode": "offline", "answer": "[{\"word\":\"realized\",\"start\":0.94,\"end\":1.4}]"} {"utterance_id": "6455-67804-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0013.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.97, 1.56, "miss belinda"]], "response_mode": "offline", "answer": "[{\"word\":\"miss belinda\",\"start\":0.97,\"end\":1.56}]"} {"utterance_id": "5543-27761-0077", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5543/27761/5543-27761-0077.flac", "task_type": "semantic_detection", "query": "disease", "prompt": "Do not answer until the audio ends. Then list each instance of a disease name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.36, 0.72, "fever"]], "response_mode": "offline", "answer": "[{\"word\":\"fever\",\"start\":0.36,\"end\":0.72}]"} {"utterance_id": "700-122868-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0028.flac", "task_type": "semantic_detection", "query": "city", "prompt": "Listen to the full audio. After it finishes, find every instance of a city name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.31, 7.81, "camelot"]], "response_mode": "offline", "answer": "[{\"word\":\"camelot\",\"start\":7.31,\"end\":7.81}]"} {"utterance_id": "4323-18416-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.27, 1.76, "charlotte"]], "response_mode": "offline", "answer": "[{\"word\":\"charlotte\",\"start\":1.27,\"end\":1.76}]"} {"utterance_id": "4515-11057-0071", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0071.flac", "task_type": "lexical_detection", "query": "drop", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"drop\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.74, 1.99, "drop"]], "response_mode": "offline", "answer": "[{\"word\":\"drop\",\"start\":1.74,\"end\":1.99}]"} {"utterance_id": "6455-67803-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0026.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "You will hear a complete utterance. When it ends, list every instance of a body part you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.83, 1.29, "hand"]], "response_mode": "offline", "answer": "[{\"word\":\"hand\",\"start\":0.83,\"end\":1.29}]"} {"utterance_id": "8254-84205-0076", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0076.flac", "task_type": "lexical_detection", "query": "showing", "prompt": "When the utterance is complete, report all instances of the word \"showing\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.33, 3.66, "showing"]], "response_mode": "offline", "answer": "[{\"word\":\"showing\",\"start\":3.33,\"end\":3.66}]"} {"utterance_id": "700-122867-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0028.flac", "task_type": "lexical_detection", "query": "glass", "prompt": "After the audio ends, extract all occurrences of the word \"glass\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.05, 4.5, "glass"]], "response_mode": "offline", "answer": "[{\"word\":\"glass\",\"start\":4.05,\"end\":4.5}]"} {"utterance_id": "1630-141772-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0015.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a family relation with fields word, start, end (seconds). If none, reply [].", "labels": [[9.32, 9.9, "brother's"]], "response_mode": "offline", "answer": "[{\"word\":\"brother's\",\"start\":9.32,\"end\":9.9}]"} {"utterance_id": "2506-11278-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0005.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Process the full audio offline. Find every instance of a profession and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.42, 6.85, "teller"]], "response_mode": "offline", "answer": "[{\"word\":\"teller\",\"start\":6.42,\"end\":6.85}]"} {"utterance_id": "7601-175351-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0025.flac", "task_type": "lexical_detection", "query": "he'll", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"he'll\" with fields word, start, end (seconds). If none, reply [].", "labels": [[7.3, 7.52, "he'll"]], "response_mode": "offline", "answer": "[{\"word\":\"he'll\",\"start\":7.3,\"end\":7.52}]"} {"utterance_id": "7697-245712-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/245712/7697-245712-0004.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Process the full audio offline. Find every instance of an animal name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 3.9, "animals"]], "response_mode": "offline", "answer": "[{\"word\":\"animals\",\"start\":3.38,\"end\":3.9}]"} {"utterance_id": "4323-55228-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0030.flac", "task_type": "lexical_detection", "query": "care", "prompt": "Once the audio has finished, identify every instance of the word \"care\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.32, 2.65, "care"]], "response_mode": "offline", "answer": "[{\"word\":\"care\",\"start\":2.32,\"end\":2.65}]"} {"utterance_id": "6467-94831-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0018.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After the audio ends, extract all occurrences of a person's name. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.32, 1.11, "missus green"]], "response_mode": "offline", "answer": "[{\"word\":\"missus green\",\"start\":0.32,\"end\":1.11}]"} {"utterance_id": "4572-112381-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.23, "hearts"]], "response_mode": "offline", "answer": "[{\"word\":\"hearts\",\"start\":0.54,\"end\":1.23}]"} {"utterance_id": "4831-25894-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0014.flac", "task_type": "lexical_detection", "query": "curly", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"curly\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.24, 1.61, "curly"]], "response_mode": "offline", "answer": "[{\"word\":\"curly\",\"start\":1.24,\"end\":1.61}]"} {"utterance_id": "700-122868-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0039.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.97, 2.36, "marilla"]], "response_mode": "offline", "answer": "[{\"word\":\"marilla\",\"start\":1.97,\"end\":2.36}]"} {"utterance_id": "3660-172183-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0022.flac", "task_type": "lexical_detection", "query": "meal", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"meal\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.46, "meal"]], "response_mode": "offline", "answer": "[{\"word\":\"meal\",\"start\":1.1,\"end\":1.46}]"} {"utterance_id": "7641-96670-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96670/7641-96670-0013.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.68, 5.12, "leopard's"]], "response_mode": "offline", "answer": "[{\"word\":\"leopard's\",\"start\":4.68,\"end\":5.12}]"} {"utterance_id": "3660-6517-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0009.flac", "task_type": "lexical_detection", "query": "fore", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"fore\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.89, 1.16, "fore"]], "response_mode": "offline", "answer": "[{\"word\":\"fore\",\"start\":0.89,\"end\":1.16}]"} {"utterance_id": "6841-88291-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0008.flac", "task_type": "lexical_detection", "query": "stood", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"stood\" with fields word, start, end (seconds). If none, reply [].", "labels": [[0.44, 0.69, "stood"]], "response_mode": "offline", "answer": "[{\"word\":\"stood\",\"start\":0.44,\"end\":0.69}]"} {"utterance_id": "700-122868-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0034.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.82, 2.57, "anne shirley"]], "response_mode": "offline", "answer": "[{\"word\":\"anne shirley\",\"start\":1.82,\"end\":2.57}]"} {"utterance_id": "4515-11057-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0066.flac", "task_type": "lexical_detection", "query": "think", "prompt": "After listening to the whole recording, output every instance of the word \"think\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.59, 0.87, "think"]], "response_mode": "offline", "answer": "[{\"word\":\"think\",\"start\":0.59,\"end\":0.87}]"} {"utterance_id": "7601-291468-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "You will hear a complete utterance. When it ends, list every instance of a profession you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.41, 10.93, "diver"]], "response_mode": "offline", "answer": "[{\"word\":\"diver\",\"start\":10.41,\"end\":10.93}]"} {"utterance_id": "6455-66379-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0002.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[5.62, 6.16, "captain"], [8.23, 8.52, "pilot"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":5.62,\"end\":6.16},{\"word\":\"pilot\",\"start\":8.23,\"end\":8.52}]"} {"utterance_id": "6599-38591-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38591/6599-38591-0006.flac", "task_type": "lexical_counting", "query": "emperor", "prompt": "Once the audio has finished, list each occurrence of the word \"emperor\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[5.02, 5.47, "one"], [12.99, 13.37, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.02,\"end\":5.47},{\"word\":\"two\",\"start\":12.99,\"end\":13.37}]"} {"utterance_id": "1255-90413-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0011.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.33, 8.68, "head"], [13.56, 14.06, "face"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":8.33,\"end\":8.68},{\"word\":\"face\",\"start\":13.56,\"end\":14.06}]"} {"utterance_id": "4323-13259-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0007.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Process the full audio offline. Find every instance of an organization name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.06, 1.3, "church"]], "response_mode": "offline", "answer": "[{\"word\":\"church\",\"start\":1.06,\"end\":1.3}]"} {"utterance_id": "1701-141759-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0012.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the full audio. After it finishes, find every instance of an animal name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.66, 3.01, "goose"]], "response_mode": "offline", "answer": "[{\"word\":\"goose\",\"start\":2.66,\"end\":3.01}]"} {"utterance_id": "4831-29134-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0007.flac", "task_type": "semantic_detection", "query": "country", "prompt": "After listening to the whole recording, output every instance of a country name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.42, 0.63, "earth"]], "response_mode": "offline", "answer": "[{\"word\":\"earth\",\"start\":0.42,\"end\":0.63}]"} {"utterance_id": "5849-50873-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0032.flac", "task_type": "lexical_detection", "query": "charter", "prompt": "When the utterance is complete, report all instances of the word \"charter\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.39, 1.93, "charter"]], "response_mode": "offline", "answer": "[{\"word\":\"charter\",\"start\":1.39,\"end\":1.93}]"} {"utterance_id": "116-288048-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0026.flac", "task_type": "lexical_detection", "query": "prayers", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"prayers\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.75, 1.15, "prayers"]], "response_mode": "offline", "answer": "[{\"word\":\"prayers\",\"start\":0.75,\"end\":1.15}]"} {"utterance_id": "3660-6517-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0022.flac", "task_type": "semantic_detection", "query": "plant", "prompt": "Process the full audio offline. Find every instance of a plant name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.76, 4.35, "squash"]], "response_mode": "offline", "answer": "[{\"word\":\"squash\",\"start\":3.76,\"end\":4.35}]"} {"utterance_id": "6455-66379-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0004.flac", "task_type": "lexical_detection", "query": "pilot's", "prompt": "Process the full audio offline. Find every instance of the word \"pilot's\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.39, 0.75, "pilot's"]], "response_mode": "offline", "answer": "[{\"word\":\"pilot's\",\"start\":0.39,\"end\":0.75}]"} {"utterance_id": "4831-25894-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.49, 0.83, "tommo"]], "response_mode": "offline", "answer": "[{\"word\":\"tommo\",\"start\":0.49,\"end\":0.83}]"} {"utterance_id": "7601-291468-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/291468/7601-291468-0006.flac", "task_type": "semantic_counting", "query": "animal", "prompt": "When the utterance is complete, report how an animal name accumulated over time using count words and spans. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[18.11, 18.52, "one"], [24.76, 25.37, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":18.11,\"end\":18.52},{\"word\":\"two\",\"start\":24.76,\"end\":25.37}]"} {"utterance_id": "4570-56594-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/56594/4570-56594-0008.flac", "task_type": "lexical_detection", "query": "signs", "prompt": "Once the audio has finished, identify every instance of the word \"signs\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.71, 3.4, "signs"]], "response_mode": "offline", "answer": "[{\"word\":\"signs\",\"start\":2.71,\"end\":3.4}]"} {"utterance_id": "4153-186222-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0006.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "When the utterance is complete, report all instances of a person's name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.46, 2.89, "louise"]], "response_mode": "offline", "answer": "[{\"word\":\"louise\",\"start\":2.46,\"end\":2.89}]"} {"utterance_id": "6841-88291-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.21, 0.41, "three"], [1.89, 2.24, "five"], [5.15, 5.37, "three"], [10.87, 11.12, "half"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":0.21,\"end\":0.41},{\"word\":\"five\",\"start\":1.89,\"end\":2.24},{\"word\":\"three\",\"start\":5.15,\"end\":5.37},{\"word\":\"half\",\"start\":10.87,\"end\":11.12}]"} {"utterance_id": "1686-142278-0067", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0067.flac", "task_type": "lexical_detection", "query": "foreign", "prompt": "Once the audio has finished, identify every instance of the word \"foreign\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.27, 4.68, "foreign"]], "response_mode": "offline", "answer": "[{\"word\":\"foreign\",\"start\":4.27,\"end\":4.68}]"} {"utterance_id": "1686-142278-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0002.flac", "task_type": "lexical_detection", "query": "cheerful", "prompt": "When the utterance is complete, report all instances of the word \"cheerful\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.62, 10.01, "cheerful"]], "response_mode": "offline", "answer": "[{\"word\":\"cheerful\",\"start\":9.62,\"end\":10.01}]"} {"utterance_id": "700-122868-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0005.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Once the audio has finished, identify every instance of an animal name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.55, 1.96, "trout"], [7.55, 7.78, "duck"]], "response_mode": "offline", "answer": "[{\"word\":\"trout\",\"start\":1.55,\"end\":1.96},{\"word\":\"duck\",\"start\":7.55,\"end\":7.78}]"} {"utterance_id": "1701-141760-0050", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0050.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.71, 1.43, "prince andrew"]], "response_mode": "offline", "answer": "[{\"word\":\"prince andrew\",\"start\":0.71,\"end\":1.43}]"} {"utterance_id": "8254-84205-0069", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0069.flac", "task_type": "lexical_detection", "query": "starve", "prompt": "When the utterance is complete, report all instances of the word \"starve\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.18, 2.72, "starve"]], "response_mode": "offline", "answer": "[{\"word\":\"starve\",\"start\":2.18,\"end\":2.72}]"} {"utterance_id": "4831-25894-0035", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/25894/4831-25894-0035.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Do not answer until the audio ends. Then list each instance of a family relation as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.93, 2.25, "mamma"]], "response_mode": "offline", "answer": "[{\"word\":\"mamma\",\"start\":1.93,\"end\":2.25}]"} {"utterance_id": "1585-157660-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0006.flac", "task_type": "lexical_detection", "query": "realm", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"realm\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.71, 5.98, "realm"]], "response_mode": "offline", "answer": "[{\"word\":\"realm\",\"start\":5.71,\"end\":5.98}]"} {"utterance_id": "1585-157660-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/157660/1585-157660-0007.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.68, 0.94, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":0.68,\"end\":0.94}]"} {"utterance_id": "8254-115543-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0028.flac", "task_type": "semantic_counting", "query": "food", "prompt": "Listen to the full audio. Count every occurrence of a food name. After it finishes, for each occurrence in order output a JSON object whose word is the count word (\"one\", \"two\", ...) with that occurrence's start/end times. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.56, 7.0, "one"], [8.05, 8.3, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":6.56,\"end\":7.0},{\"word\":\"two\",\"start\":8.05,\"end\":8.3}]"} {"utterance_id": "4831-18525-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0018.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Once the audio has finished, identify every instance of a body part. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.96, 3.14, "head"]], "response_mode": "offline", "answer": "[{\"word\":\"head\",\"start\":2.96,\"end\":3.14}]"} {"utterance_id": "8254-115543-0011", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0011.flac", "task_type": "lexical_detection", "query": "matted", "prompt": "After listening to the whole recording, output every instance of the word \"matted\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.78, 1.1, "matted"]], "response_mode": "offline", "answer": "[{\"word\":\"matted\",\"start\":0.78,\"end\":1.1}]"} {"utterance_id": "1650-167613-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0028.flac", "task_type": "lexical_detection", "query": "ridden", "prompt": "After the audio ends, extract all occurrences of the word \"ridden\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.46, 4.02, "ridden"]], "response_mode": "offline", "answer": "[{\"word\":\"ridden\",\"start\":3.46,\"end\":4.02}]"} {"utterance_id": "6841-88291-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0046.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "After listening to the whole recording, output every instance of an animal name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.4, 4.72, "calf"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":4.4,\"end\":4.72}]"} {"utterance_id": "2506-11267-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0009.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.42, 3.85, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":3.42,\"end\":3.85}]"} {"utterance_id": "4153-61735-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0012.flac", "task_type": "semantic_detection", "query": "money", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a money or currency term with fields word, start, end (seconds). If none, reply [].", "labels": [[5.66, 6.24, "cash"]], "response_mode": "offline", "answer": "[{\"word\":\"cash\",\"start\":5.66,\"end\":6.24}]"} {"utterance_id": "1650-167613-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0017.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.21, "pierre"]], "response_mode": "offline", "answer": "[{\"word\":\"pierre\",\"start\":2.7,\"end\":3.21}]"} {"utterance_id": "8288-274162-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274162/8288-274162-0023.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Once the audio has finished, identify every instance of a family relation. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[2.0, 2.6, "son"]], "response_mode": "offline", "answer": "[{\"word\":\"son\",\"start\":2.0,\"end\":2.6}]"} {"utterance_id": "2506-11267-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0001.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[9.7, 10.23, "brother"]], "response_mode": "offline", "answer": "[{\"word\":\"brother\",\"start\":9.7,\"end\":10.23}]"} {"utterance_id": "7697-105817-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105817/7697-105817-0003.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "When the utterance is complete, report all instances of a number including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.25, 1.38, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.25,\"end\":1.38}]"} {"utterance_id": "6467-97061-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/97061/6467-97061-0003.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Listen to the entire clip, then report the count sequence for a body part as timed JSON events (word is the count word, not the surface form). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.41, "one"], [7.32, 7.69, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.96,\"end\":1.41},{\"word\":\"two\",\"start\":7.32,\"end\":7.69}]"} {"utterance_id": "7641-96684-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96684/7641-96684-0034.flac", "task_type": "lexical_detection", "query": "knowing", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"knowing\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.0, 1.35, "knowing"]], "response_mode": "offline", "answer": "[{\"word\":\"knowing\",\"start\":1.0,\"end\":1.35}]"} {"utterance_id": "1255-90413-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0025.flac", "task_type": "lexical_detection", "query": "appointed", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"appointed\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.45, "appointed"]], "response_mode": "offline", "answer": "[{\"word\":\"appointed\",\"start\":1.05,\"end\":1.45}]"} {"utterance_id": "116-288046-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0006.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen carefully to the entire clip, then report every instance of a body part as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.74, 3.08, "mouth"]], "response_mode": "offline", "answer": "[{\"word\":\"mouth\",\"start\":2.74,\"end\":3.08}]"} {"utterance_id": "4153-186223-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186223/4153-186223-0007.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "After the audio ends, tally every instance of a person's name. Reply with JSON only — one object per occurrence, word=\"one\"/\"two\"/... plus start and end in seconds. If none, reply [].", "labels": [[0.4, 1.0, "one"], [2.4, 2.94, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.4,\"end\":1.0},{\"word\":\"two\",\"start\":2.4,\"end\":2.94}]"} {"utterance_id": "6123-59150-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59150/6123-59150-0000.flac", "task_type": "lexical_detection", "query": "characters", "prompt": "When the utterance is complete, report all instances of the word \"characters\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.52, 1.0, "characters"]], "response_mode": "offline", "answer": "[{\"word\":\"characters\",\"start\":0.52,\"end\":1.0}]"} {"utterance_id": "1686-142278-0048", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0048.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.44, 0.95, "margaret"]], "response_mode": "offline", "answer": "[{\"word\":\"margaret\",\"start\":0.44,\"end\":0.95}]"} {"utterance_id": "7697-105815-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7697/105815/7697-105815-0010.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a number with fields word, start, end (seconds). If none, reply [].", "labels": [[2.82, 3.06, "three"], [4.83, 5.03, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"three\",\"start\":2.82,\"end\":3.06},{\"word\":\"one\",\"start\":4.83,\"end\":5.03}]"} {"utterance_id": "4323-55228-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0033.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.73, 2.07, "kate"], [3.11, 3.46, "bertram"]], "response_mode": "offline", "answer": "[{\"word\":\"kate\",\"start\":1.73,\"end\":2.07},{\"word\":\"bertram\",\"start\":3.11,\"end\":3.46}]"} {"utterance_id": "2506-13150-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/13150/2506-13150-0006.flac", "task_type": "lexical_detection", "query": "full", "prompt": "After the audio ends, extract all occurrences of the word \"full\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.5, 2.75, "full"]], "response_mode": "offline", "answer": "[{\"word\":\"full\",\"start\":2.5,\"end\":2.75}]"} {"utterance_id": "116-288047-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0022.flac", "task_type": "semantic_detection", "query": "money", "prompt": "After the audio ends, extract all occurrences of a money or currency term. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.03, 6.46, "money"]], "response_mode": "offline", "answer": "[{\"word\":\"money\",\"start\":6.03,\"end\":6.46}]"} {"utterance_id": "4572-112381-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0014.flac", "task_type": "lexical_counting", "query": "leave", "prompt": "Listen to the audio in full, then emit a JSON array of count events for the word \"leave\" (word, start, end). If none, reply [].", "labels": [[3.65, 4.01, "one"], [6.97, 7.41, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.65,\"end\":4.01},{\"word\":\"two\",\"start\":6.97,\"end\":7.41}]"} {"utterance_id": "6267-65525-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/65525/6267-65525-0040.flac", "task_type": "lexical_counting", "query": "understand", "prompt": "After listening to the whole recording, output timed count words for the word \"understand\" (one, two, three, ...) as JSON. If none, reply [].", "labels": [[1.82, 2.56, "one"], [3.88, 4.56, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.82,\"end\":2.56},{\"word\":\"two\",\"start\":3.88,\"end\":4.56}]"} {"utterance_id": "8254-84205-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0013.flac", "task_type": "lexical_detection", "query": "chris", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"chris\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.34, 1.76, "chris"]], "response_mode": "offline", "answer": "[{\"word\":\"chris\",\"start\":1.34,\"end\":1.76}]"} {"utterance_id": "4153-186222-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0018.flac", "task_type": "lexical_detection", "query": "beth", "prompt": "When the utterance is complete, report all instances of the word \"beth\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.09, 1.39, "beth"]], "response_mode": "offline", "answer": "[{\"word\":\"beth\",\"start\":1.09,\"end\":1.39}]"} {"utterance_id": "1686-142278-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0022.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Do not answer until the audio ends. Then list each instance of a time of day as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[3.29, 3.59, "night"]], "response_mode": "offline", "answer": "[{\"word\":\"night\",\"start\":3.29,\"end\":3.59}]"} {"utterance_id": "3915-57461-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0030.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "You will hear a complete utterance. When it ends, list every instance of an animal name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.2, 6.65, "stork's"]], "response_mode": "offline", "answer": "[{\"word\":\"stork's\",\"start\":6.2,\"end\":6.65}]"} {"utterance_id": "1585-131718-0033", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0033.flac", "task_type": "lexical_detection", "query": "gas", "prompt": "After listening to the whole recording, output every instance of the word \"gas\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.9, "gas"]], "response_mode": "offline", "answer": "[{\"word\":\"gas\",\"start\":1.59,\"end\":1.9}]"} {"utterance_id": "4831-29134-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/29134/4831-29134-0014.flac", "task_type": "lexical_detection", "query": "grand", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"grand\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.08, 3.49, "grand"]], "response_mode": "offline", "answer": "[{\"word\":\"grand\",\"start\":3.08,\"end\":3.49}]"} {"utterance_id": "3663-172528-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0001.flac", "task_type": "lexical_detection", "query": "annoyance", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"annoyance\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.76, 3.54, "annoyance"]], "response_mode": "offline", "answer": "[{\"word\":\"annoyance\",\"start\":2.76,\"end\":3.54}]"} {"utterance_id": "4515-11057-0088", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0088.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.5, 2.0, "jasper"]], "response_mode": "offline", "answer": "[{\"word\":\"jasper\",\"start\":1.5,\"end\":2.0}]"} {"utterance_id": "8254-84205-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0040.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Listen carefully to the entire clip, then report every instance of an organization name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.76, 1.29, "redskins"]], "response_mode": "offline", "answer": "[{\"word\":\"redskins\",\"start\":0.76,\"end\":1.29}]"} {"utterance_id": "6599-38590-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6599/38590/6599-38590-0005.flac", "task_type": "lexical_detection", "query": "appointment", "prompt": "After listening to the whole recording, output every instance of the word \"appointment\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.53, 5.02, "appointment"]], "response_mode": "offline", "answer": "[{\"word\":\"appointment\",\"start\":4.53,\"end\":5.02}]"} {"utterance_id": "4323-13259-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0019.flac", "task_type": "lexical_detection", "query": "passing", "prompt": "After listening to the whole recording, output every instance of the word \"passing\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.85, 4.22, "passing"]], "response_mode": "offline", "answer": "[{\"word\":\"passing\",\"start\":3.85,\"end\":4.22}]"} {"utterance_id": "1686-142278-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0057.flac", "task_type": "lexical_detection", "query": "mother", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"mother\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.12, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.74,\"end\":2.12}]"} {"utterance_id": "700-122866-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0024.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.06, 3.85, "sixteen"]], "response_mode": "offline", "answer": "[{\"word\":\"sixteen\",\"start\":3.06,\"end\":3.85}]"} {"utterance_id": "7601-175351-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/175351/7601-175351-0007.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a profession with fields word, start, end (seconds). If none, reply [].", "labels": [[3.99, 4.6, "captain"]], "response_mode": "offline", "answer": "[{\"word\":\"captain\",\"start\":3.99,\"end\":4.6}]"} {"utterance_id": "1255-90413-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90413/1255-90413-0026.flac", "task_type": "lexical_detection", "query": "present", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"present\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.54, 1.88, "present"]], "response_mode": "offline", "answer": "[{\"word\":\"present\",\"start\":1.54,\"end\":1.88}]"} {"utterance_id": "8254-115543-0045", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0045.flac", "task_type": "lexical_detection", "query": "patted", "prompt": "Once the audio has finished, identify every instance of the word \"patted\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.2, 1.54, "patted"]], "response_mode": "offline", "answer": "[{\"word\":\"patted\",\"start\":1.2,\"end\":1.54}]"} {"utterance_id": "2506-11267-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11267/2506-11267-0014.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.3, 2.16, "captain bull"]], "response_mode": "offline", "answer": "[{\"word\":\"captain bull\",\"start\":1.3,\"end\":2.16}]"} {"utterance_id": "4572-112381-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112381/4572-112381-0008.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[9.6, 10.47, "king sugriva s"]], "response_mode": "offline", "answer": "[{\"word\":\"king sugriva s\",\"start\":9.6,\"end\":10.47}]"} {"utterance_id": "4570-14911-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0017.flac", "task_type": "lexical_detection", "query": "sixteen", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"sixteen\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.71, 3.21, "sixteen"]], "response_mode": "offline", "answer": "[{\"word\":\"sixteen\",\"start\":2.71,\"end\":3.21}]"} {"utterance_id": "5849-50963-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0004.flac", "task_type": "lexical_detection", "query": "delayed", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"delayed\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.92, 1.29, "delayed"]], "response_mode": "offline", "answer": "[{\"word\":\"delayed\",\"start\":0.92,\"end\":1.29}]"} {"utterance_id": "6467-94831-0040", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0040.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the full audio. After it finishes, find every instance of a person's name. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.82, "peleg"]], "response_mode": "offline", "answer": "[{\"word\":\"peleg\",\"start\":1.4,\"end\":1.82}]"} {"utterance_id": "3663-172528-0044", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0044.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.95, 3.23, "one"], [4.86, 5.83, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.95,\"end\":3.23},{\"word\":\"two\",\"start\":4.86,\"end\":5.83}]"} {"utterance_id": "700-122866-0025", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0025.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.87, 1.48, "ruby gillis"]], "response_mode": "offline", "answer": "[{\"word\":\"ruby gillis\",\"start\":0.87,\"end\":1.48}]"} {"utterance_id": "1650-167613-0041", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0041.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[12.74, 13.43, "throat"]], "response_mode": "offline", "answer": "[{\"word\":\"throat\",\"start\":12.74,\"end\":13.43}]"} {"utterance_id": "6467-56885-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/56885/6467-56885-0015.flac", "task_type": "lexical_detection", "query": "women", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"women\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.8, "women"]], "response_mode": "offline", "answer": "[{\"word\":\"women\",\"start\":0.51,\"end\":0.8}]"} {"utterance_id": "3663-172528-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3663/172528/3663-172528-0003.flac", "task_type": "lexical_detection", "query": "detected", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"detected\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[7.16, 7.83, "detected"]], "response_mode": "offline", "answer": "[{\"word\":\"detected\",\"start\":7.16,\"end\":7.83}]"} {"utterance_id": "1255-90407-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/90407/1255-90407-0006.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "You will hear a complete utterance. When it ends, list every instance of a family relation you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.56, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":1.04,\"end\":1.56}]"} {"utterance_id": "4153-61735-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/61735/4153-61735-0003.flac", "task_type": "lexical_detection", "query": "staying", "prompt": "Process the full audio offline. Find every instance of the word \"staying\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.99, 4.36, "staying"]], "response_mode": "offline", "answer": "[{\"word\":\"staying\",\"start\":3.99,\"end\":4.36}]"} {"utterance_id": "6455-67803-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67803/6455-67803-0000.flac", "task_type": "semantic_detection", "query": "date", "prompt": "Do not answer until the audio ends. Then list each instance of a date as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.9, 2.08, "may"]], "response_mode": "offline", "answer": "[{\"word\":\"may\",\"start\":1.9,\"end\":2.08}]"} {"utterance_id": "5849-50962-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50962/5849-50962-0014.flac", "task_type": "lexical_detection", "query": "belt", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"belt\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[5.71, 5.98, "belt"]], "response_mode": "offline", "answer": "[{\"word\":\"belt\",\"start\":5.71,\"end\":5.98}]"} {"utterance_id": "4323-13259-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0000.flac", "task_type": "lexical_detection", "query": "curious", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"curious\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.65, 4.23, "curious"]], "response_mode": "offline", "answer": "[{\"word\":\"curious\",\"start\":3.65,\"end\":4.23}]"} {"utterance_id": "1686-142278-0057", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0057.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Process the full audio offline. Find every instance of a family relation and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.12, "mother"]], "response_mode": "offline", "answer": "[{\"word\":\"mother\",\"start\":1.74,\"end\":2.12}]"} {"utterance_id": "116-288047-0012", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0012.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "Listen to the full audio. After it finishes, find every instance of a body part. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.51, 0.76, "feet"]], "response_mode": "offline", "answer": "[{\"word\":\"feet\",\"start\":0.51,\"end\":0.76}]"} {"utterance_id": "5849-50873-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0042.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Once the audio has finished, list each occurrence of a person's name as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[5.8, 6.49, "one"], [8.82, 9.51, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":5.8,\"end\":6.49},{\"word\":\"two\",\"start\":8.82,\"end\":9.51}]"} {"utterance_id": "4323-13259-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0005.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen carefully to the entire clip, then report every instance of a person's name as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.37, 0.99, "president marsh"]], "response_mode": "offline", "answer": "[{\"word\":\"president marsh\",\"start\":0.37,\"end\":0.99}]"} {"utterance_id": "3660-6517-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0019.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After listening to the whole recording, output every instance of a number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.05, 1.25, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"two\",\"start\":1.05,\"end\":1.25}]"} {"utterance_id": "3660-172183-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0001.flac", "task_type": "semantic_detection", "query": "color", "prompt": "When the utterance is complete, report all instances of a color name including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.98, 4.23, "white"]], "response_mode": "offline", "answer": "[{\"word\":\"white\",\"start\":3.98,\"end\":4.23}]"} {"utterance_id": "6467-94831-0026", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0026.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.48, 0.81, "sam"]], "response_mode": "offline", "answer": "[{\"word\":\"sam\",\"start\":0.48,\"end\":0.81}]"} {"utterance_id": "700-122868-0036", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0036.flac", "task_type": "lexical_detection", "query": "gets", "prompt": "Process the full audio offline. Find every instance of the word \"gets\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.08, 1.47, "gets"]], "response_mode": "offline", "answer": "[{\"word\":\"gets\",\"start\":1.08,\"end\":1.47}]"} {"utterance_id": "1630-73710-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/73710/1630-73710-0000.flac", "task_type": "lexical_detection", "query": "poet", "prompt": "After the audio ends, extract all occurrences of the word \"poet\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.05, 10.64, "poet"]], "response_mode": "offline", "answer": "[{\"word\":\"poet\",\"start\":10.05,\"end\":10.64}]"} {"utterance_id": "5849-50964-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0000.flac", "task_type": "lexical_detection", "query": "dorothy", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"dorothy\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.75, 3.2, "dorothy"]], "response_mode": "offline", "answer": "[{\"word\":\"dorothy\",\"start\":2.75,\"end\":3.2}]"} {"utterance_id": "4323-55228-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0003.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "After listening to the whole recording, output every instance of a person's name with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.24, 2.45, "billy"], [3.93, 4.22, "pete"]], "response_mode": "offline", "answer": "[{\"word\":\"billy\",\"start\":2.24,\"end\":2.45},{\"word\":\"pete\",\"start\":3.93,\"end\":4.22}]"} {"utterance_id": "1585-131718-0049", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0049.flac", "task_type": "lexical_detection", "query": "harrison", "prompt": "Once the audio has finished, identify every instance of the word \"harrison\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[4.48, 4.91, "harrison"]], "response_mode": "offline", "answer": "[{\"word\":\"harrison\",\"start\":4.48,\"end\":4.91}]"} {"utterance_id": "1701-141760-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0022.flac", "task_type": "lexical_detection", "query": "coats", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"coats\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.7, 2.27, "coats"]], "response_mode": "offline", "answer": "[{\"word\":\"coats\",\"start\":1.7,\"end\":2.27}]"} {"utterance_id": "1630-96099-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0021.flac", "task_type": "semantic_counting", "query": "money", "prompt": "Process the full audio offline. For every occurrence of a money or currency term, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.07, 2.67, "one"], [4.17, 4.99, "two"], [7.83, 8.48, "three"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.07,\"end\":2.67},{\"word\":\"two\",\"start\":4.17,\"end\":4.99},{\"word\":\"three\",\"start\":7.83,\"end\":8.48}]"} {"utterance_id": "4572-64670-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/64670/4572-64670-0016.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "After listening to the whole recording, output every instance of a family relation with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.59, 4.08, "father"]], "response_mode": "offline", "answer": "[{\"word\":\"father\",\"start\":3.59,\"end\":4.08}]"} {"utterance_id": "1255-74899-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1255/74899/1255-74899-0018.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "Listen to the full audio. After it finishes, find every instance of a number. For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.02, 1.24, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.02,\"end\":1.24}]"} {"utterance_id": "6467-94831-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6467/94831/6467-94831-0000.flac", "task_type": "semantic_detection", "query": "clock_time", "prompt": "Listen carefully to the entire clip, then report every instance of a time of day as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[11.08, 11.66, "afternoon"]], "response_mode": "offline", "answer": "[{\"word\":\"afternoon\",\"start\":11.08,\"end\":11.66}]"} {"utterance_id": "5849-50873-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0006.flac", "task_type": "lexical_detection", "query": "nigh", "prompt": "Once the audio has finished, identify every instance of the word \"nigh\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.43, 1.69, "nigh"]], "response_mode": "offline", "answer": "[{\"word\":\"nigh\",\"start\":1.43,\"end\":1.69}]"} {"utterance_id": "6841-88291-0037", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0037.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 1.96, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":1.51,\"end\":1.96}]"} {"utterance_id": "4570-14911-0015", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4570/14911/4570-14911-0015.flac", "task_type": "lexical_detection", "query": "strongly", "prompt": "After the audio ends, extract all occurrences of the word \"strongly\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.81, 4.36, "strongly"]], "response_mode": "offline", "answer": "[{\"word\":\"strongly\",\"start\":3.81,\"end\":4.36}]"} {"utterance_id": "6267-53049-0024", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0024.flac", "task_type": "semantic_counting", "query": "body_part", "prompt": "Process the full audio offline. For every occurrence of a body part, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.95, 5.35, "one"], [10.43, 10.83, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.95,\"end\":5.35},{\"word\":\"two\",\"start\":10.43,\"end\":10.83}]"} {"utterance_id": "4831-18525-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4831/18525/4831-18525-0028.flac", "task_type": "lexical_detection", "query": "thanks", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"thanks\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.7, 3.14, "thanks"]], "response_mode": "offline", "answer": "[{\"word\":\"thanks\",\"start\":2.7,\"end\":3.14}]"} {"utterance_id": "3915-57461-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/57461/3915-57461-0028.flac", "task_type": "semantic_detection", "query": "cardinal", "prompt": "After the audio ends, extract all occurrences of a number. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.81, 3.11, "one"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.81,\"end\":3.11}]"} {"utterance_id": "6841-88291-0022", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0022.flac", "task_type": "lexical_detection", "query": "iron", "prompt": "Process the full audio offline. Find every instance of the word \"iron\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.04, 1.33, "iron"]], "response_mode": "offline", "answer": "[{\"word\":\"iron\",\"start\":1.04,\"end\":1.33}]"} {"utterance_id": "116-288047-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288047/116-288047-0021.flac", "task_type": "lexical_detection", "query": "entirely", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"entirely\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.71, 2.23, "entirely"]], "response_mode": "offline", "answer": "[{\"word\":\"entirely\",\"start\":1.71,\"end\":2.23}]"} {"utterance_id": "1701-141760-0039", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0039.flac", "task_type": "lexical_detection", "query": "meaning", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"meaning\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.88, "meaning"]], "response_mode": "offline", "answer": "[{\"word\":\"meaning\",\"start\":1.59,\"end\":1.88}]"} {"utterance_id": "1650-167613-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1650/167613/1650-167613-0019.flac", "task_type": "lexical_detection", "query": "chess", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"chess\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.05, 3.6, "chess"]], "response_mode": "offline", "answer": "[{\"word\":\"chess\",\"start\":3.05,\"end\":3.6}]"} {"utterance_id": "4153-186222-0006", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0006.flac", "task_type": "lexical_detection", "query": "corrected", "prompt": "After listening to the whole recording, output every instance of the word \"corrected\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.01, 2.46, "corrected"]], "response_mode": "offline", "answer": "[{\"word\":\"corrected\",\"start\":2.01,\"end\":2.46}]"} {"utterance_id": "116-288046-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288046/116-288046-0002.flac", "task_type": "lexical_detection", "query": "directly", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"directly\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.22, 3.75, "directly"]], "response_mode": "offline", "answer": "[{\"word\":\"directly\",\"start\":3.22,\"end\":3.75}]"} {"utterance_id": "6455-66379-0003", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0003.flac", "task_type": "lexical_counting", "query": "it's", "prompt": "Once the audio has finished, list each occurrence of the word \"it's\" as {\"word\": , \"start\", \"end\"} in order. If none, reply [].", "labels": [[2.22, 2.4, "one"], [4.89, 5.02, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.22,\"end\":2.4},{\"word\":\"two\",\"start\":4.89,\"end\":5.02}]"} {"utterance_id": "1686-142278-0066", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0066.flac", "task_type": "semantic_detection", "query": "organization", "prompt": "Do not answer until the audio ends. Then list each instance of an organization name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[2.37, 3.23, "spanish army"]], "response_mode": "offline", "answer": "[{\"word\":\"spanish army\",\"start\":2.37,\"end\":3.23}]"} {"utterance_id": "4323-55228-0010", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0010.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[0.08, 0.46, "one"], [5.18, 5.49, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.08,\"end\":0.46},{\"word\":\"two\",\"start\":5.18,\"end\":5.49}]"} {"utterance_id": "4515-11057-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0002.flac", "task_type": "lexical_detection", "query": "duty", "prompt": "When the utterance is complete, report all instances of the word \"duty\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.57, 2.02, "duty"]], "response_mode": "offline", "answer": "[{\"word\":\"duty\",\"start\":1.57,\"end\":2.02}]"} {"utterance_id": "1651-136854-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1651/136854/1651-136854-0004.flac", "task_type": "lexical_detection", "query": "fragments", "prompt": "After listening to the whole recording, output every instance of the word \"fragments\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.37, 3.89, "fragments"]], "response_mode": "offline", "answer": "[{\"word\":\"fragments\",\"start\":3.37,\"end\":3.89}]"} {"utterance_id": "3660-172183-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0019.flac", "task_type": "lexical_counting", "query": "meadows", "prompt": "Process the full audio offline. For every occurrence of the word \"meadows\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.61, 3.0, "one"], [4.86, 5.4, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":2.61,\"end\":3.0},{\"word\":\"two\",\"start\":4.86,\"end\":5.4}]"} {"utterance_id": "116-288048-0030", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0030.flac", "task_type": "lexical_detection", "query": "ask", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"ask\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.51, 1.84, "ask"]], "response_mode": "offline", "answer": "[{\"word\":\"ask\",\"start\":1.51,\"end\":1.84}]"} {"utterance_id": "700-122868-0001", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122868/700-122868-0001.flac", "task_type": "lexical_detection", "query": "romantic", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"romantic\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.74, 2.47, "romantic"]], "response_mode": "offline", "answer": "[{\"word\":\"romantic\",\"start\":1.74,\"end\":2.47}]"} {"utterance_id": "1585-131718-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1585/131718/1585-131718-0027.flac", "task_type": "lexical_detection", "query": "nicely", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"nicely\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.69, 3.22, "nicely"]], "response_mode": "offline", "answer": "[{\"word\":\"nicely\",\"start\":2.69,\"end\":3.22}]"} {"utterance_id": "1701-141760-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141760/1701-141760-0042.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of a person's name with fields word, start, end (seconds). If none, reply [].", "labels": [[0.46, 0.82, "boris"]], "response_mode": "offline", "answer": "[{\"word\":\"boris\",\"start\":0.46,\"end\":0.82}]"} {"utterance_id": "116-288048-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288048/116-288048-0018.flac", "task_type": "lexical_detection", "query": "authors", "prompt": "After the audio ends, extract all occurrences of the word \"authors\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.94, 4.36, "authors"]], "response_mode": "offline", "answer": "[{\"word\":\"authors\",\"start\":3.94,\"end\":4.36}]"} {"utterance_id": "4572-112375-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4572/112375/4572-112375-0008.flac", "task_type": "semantic_detection", "query": "body_part", "prompt": "When the utterance is complete, report all instances of a body part including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.59, 1.96, "heart"]], "response_mode": "offline", "answer": "[{\"word\":\"heart\",\"start\":1.59,\"end\":1.96}]"} {"utterance_id": "3660-6517-0002", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0002.flac", "task_type": "lexical_detection", "query": "steam", "prompt": "After listening to the whole recording, output every instance of the word \"steam\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.67, 1.16, "steam"]], "response_mode": "offline", "answer": "[{\"word\":\"steam\",\"start\":0.67,\"end\":1.16}]"} {"utterance_id": "6841-88291-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6841/88291/6841-88291-0014.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of an animal name with fields word, start, end (seconds). If none, reply [].", "labels": [[1.71, 2.11, "calf"]], "response_mode": "offline", "answer": "[{\"word\":\"calf\",\"start\":1.71,\"end\":2.11}]"} {"utterance_id": "8254-115543-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0023.flac", "task_type": "semantic_detection", "query": "animal", "prompt": "Do not answer until the audio ends. Then list each instance of an animal name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.8, 1.25, "elephant's"]], "response_mode": "offline", "answer": "[{\"word\":\"elephant's\",\"start\":0.8,\"end\":1.25}]"} {"utterance_id": "4323-55228-0018", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0018.flac", "task_type": "lexical_detection", "query": "mystified", "prompt": "Process the full audio offline. Find every instance of the word \"mystified\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.45, 1.98, "mystified"]], "response_mode": "offline", "answer": "[{\"word\":\"mystified\",\"start\":1.45,\"end\":1.98}]"} {"utterance_id": "6455-67804-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0005.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "Listen carefully to the entire clip, then report every instance of an ordinal number as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.58, 2.86, "first"]], "response_mode": "offline", "answer": "[{\"word\":\"first\",\"start\":2.58,\"end\":2.86}]"} {"utterance_id": "8254-115543-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0008.flac", "task_type": "lexical_detection", "query": "wonderful", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"wonderful\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[10.74, 11.17, "wonderful"]], "response_mode": "offline", "answer": "[{\"word\":\"wonderful\",\"start\":10.74,\"end\":11.17}]"} {"utterance_id": "6455-67804-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/67804/6455-67804-0017.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Process the full audio offline. For every occurrence of a person's name, emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.54, 1.08, "one"], [5.06, 5.72, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":0.54,\"end\":1.08},{\"word\":\"two\",\"start\":5.06,\"end\":5.72}]"} {"utterance_id": "8254-84205-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0034.flac", "task_type": "lexical_detection", "query": "bit", "prompt": "After listening to the whole recording, output every instance of the word \"bit\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.07, 1.25, "bit"]], "response_mode": "offline", "answer": "[{\"word\":\"bit\",\"start\":1.07,\"end\":1.25}]"} {"utterance_id": "1686-142278-0013", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1686/142278/1686-142278-0013.flac", "task_type": "lexical_detection", "query": "minute", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"minute\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.9, 2.21, "minute"]], "response_mode": "offline", "answer": "[{\"word\":\"minute\",\"start\":1.9,\"end\":2.21}]"} {"utterance_id": "4515-11057-0042", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0042.flac", "task_type": "lexical_detection", "query": "reproached", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"reproached\" with fields word, start, end (seconds). If none, reply [].", "labels": [[2.1, 2.68, "reproached"]], "response_mode": "offline", "answer": "[{\"word\":\"reproached\",\"start\":2.1,\"end\":2.68}]"} {"utterance_id": "1701-141759-0021", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1701/141759/1701-141759-0021.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Once the audio has finished, identify every instance of a person's name. Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[0.5, 0.95, "sonya"]], "response_mode": "offline", "answer": "[{\"word\":\"sonya\",\"start\":0.5,\"end\":0.95}]"} {"utterance_id": "3660-172183-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/172183/3660-172183-0005.flac", "task_type": "lexical_detection", "query": "completely", "prompt": "When the utterance is complete, report all instances of the word \"completely\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.22, 6.82, "completely"]], "response_mode": "offline", "answer": "[{\"word\":\"completely\",\"start\":6.22,\"end\":6.82}]"} {"utterance_id": "1630-141772-0014", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/141772/1630-141772-0014.flac", "task_type": "lexical_detection", "query": "shone", "prompt": "After listening to the whole recording, output every instance of the word \"shone\" with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.28, 1.69, "shone"]], "response_mode": "offline", "answer": "[{\"word\":\"shone\",\"start\":1.28,\"end\":1.69}]"} {"utterance_id": "3660-6517-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0004.flac", "task_type": "semantic_detection", "query": "color", "prompt": "Do not answer until the audio ends. Then list each instance of a color name as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.74, 1.02, "green"]], "response_mode": "offline", "answer": "[{\"word\":\"green\",\"start\":0.74,\"end\":1.02}]"} {"utterance_id": "6123-59186-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0032.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.35, 2.15, "mister hardy"]], "response_mode": "offline", "answer": "[{\"word\":\"mister hardy\",\"start\":1.35,\"end\":2.15}]"} {"utterance_id": "4323-55228-0009", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0009.flac", "task_type": "lexical_detection", "query": "indignant", "prompt": "When the utterance is complete, report all instances of the word \"indignant\" including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.97, 5.42, "indignant"]], "response_mode": "offline", "answer": "[{\"word\":\"indignant\",\"start\":4.97,\"end\":5.42}]"} {"utterance_id": "700-122867-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122867/700-122867-0031.flac", "task_type": "semantic_detection", "query": "ordinal", "prompt": "After listening to the whole recording, output every instance of an ordinal number with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.62, 3.1, "next"]], "response_mode": "offline", "answer": "[{\"word\":\"next\",\"start\":2.62,\"end\":3.1}]"} {"utterance_id": "8288-274150-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8288/274150/8288-274150-0007.flac", "task_type": "lexical_detection", "query": "articulate", "prompt": "After the audio ends, extract all occurrences of the word \"articulate\". Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[7.25, 7.91, "articulate"]], "response_mode": "offline", "answer": "[{\"word\":\"articulate\",\"start\":7.25,\"end\":7.91}]"} {"utterance_id": "4515-11057-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4515/11057/4515-11057-0016.flac", "task_type": "lexical_detection", "query": "imagined", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"imagined\" with fields word, start, end (seconds). If none, reply [].", "labels": [[1.67, 2.24, "imagined"]], "response_mode": "offline", "answer": "[{\"word\":\"imagined\",\"start\":1.67,\"end\":2.24}]"} {"utterance_id": "4323-55228-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/55228/4323-55228-0017.flac", "task_type": "lexical_detection", "query": "stammered", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"stammered\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[2.34, 2.75, "stammered"]], "response_mode": "offline", "answer": "[{\"word\":\"stammered\",\"start\":2.34,\"end\":2.75}]"} {"utterance_id": "8173-294714-0046", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8173/294714/8173-294714-0046.flac", "task_type": "lexical_detection", "query": "reach", "prompt": "Process the full audio offline. Find every instance of the word \"reach\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[6.21, 6.63, "reach"]], "response_mode": "offline", "answer": "[{\"word\":\"reach\",\"start\":6.21,\"end\":6.63}]"} {"utterance_id": "700-122866-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/700/122866/700-122866-0005.flac", "task_type": "lexical_counting", "query": "old", "prompt": "Process the full audio offline. For every occurrence of the word \"old\", emit a count-word event with timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[4.05, 4.37, "one"], [5.68, 5.92, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":4.05,\"end\":4.37},{\"word\":\"two\",\"start\":5.68,\"end\":5.92}]"} {"utterance_id": "4153-186222-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4153/186222/4153-186222-0034.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "When the utterance is complete, report all instances of a family relation including the spoken word and time span. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.1, 1.45, "niece"]], "response_mode": "offline", "answer": "[{\"word\":\"niece\",\"start\":1.1,\"end\":1.45}]"} {"utterance_id": "8254-115543-0028", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/115543/8254-115543-0028.flac", "task_type": "lexical_detection", "query": "halt", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"halt\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.4, 1.73, "halt"]], "response_mode": "offline", "answer": "[{\"word\":\"halt\",\"start\":1.4,\"end\":1.73}]"} {"utterance_id": "6267-53049-0031", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6267/53049/6267-53049-0031.flac", "task_type": "lexical_detection", "query": "eye", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"eye\" with fields word, start, end (seconds). If none, reply [].", "labels": [[7.34, 7.66, "eye"]], "response_mode": "offline", "answer": "[{\"word\":\"eye\",\"start\":7.34,\"end\":7.66}]"} {"utterance_id": "1630-96099-0017", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/1630/96099/1630-96099-0017.flac", "task_type": "lexical_detection", "query": "chief", "prompt": "Process the full audio offline. Find every instance of the word \"chief\" and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.91, 2.38, "chief"]], "response_mode": "offline", "answer": "[{\"word\":\"chief\",\"start\":1.91,\"end\":2.38}]"} {"utterance_id": "4323-18416-0023", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/18416/4323-18416-0023.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "You will hear a complete utterance. When it ends, list every instance of a person's name you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.96, 1.42, "mister king"]], "response_mode": "offline", "answer": "[{\"word\":\"mister king\",\"start\":0.96,\"end\":1.42}]"} {"utterance_id": "3915-98647-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3915/98647/3915-98647-0008.flac", "task_type": "lexical_detection", "query": "finding", "prompt": "Listen to the audio in full, then emit a JSON array of every instance of the word \"finding\" with fields word, start, end (seconds). If none, reply [].", "labels": [[3.7, 4.17, "finding"]], "response_mode": "offline", "answer": "[{\"word\":\"finding\",\"start\":3.7,\"end\":4.17}]"} {"utterance_id": "5849-50873-0034", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0034.flac", "task_type": "lexical_detection", "query": "member", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"member\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[0.78, 1.1, "member"]], "response_mode": "offline", "answer": "[{\"word\":\"member\",\"start\":0.78,\"end\":1.1}]"} {"utterance_id": "5849-50963-0007", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50963/5849-50963-0007.flac", "task_type": "semantic_detection", "query": "person_name", "prompt": "Process the full audio offline. Find every instance of a person's name and give the surface word plus start/end timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.65, 2.49, "mister bobbsey"]], "response_mode": "offline", "answer": "[{\"word\":\"mister bobbsey\",\"start\":1.65,\"end\":2.49}]"} {"utterance_id": "8254-84205-0075", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/8254/84205/8254-84205-0075.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After the audio ends, extract all occurrences of a profession. Output each spoken word with its start and end time. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.37, 1.7, "lookout"]], "response_mode": "offline", "answer": "[{\"word\":\"lookout\",\"start\":1.37,\"end\":1.7}]"} {"utterance_id": "2506-11278-0019", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/2506/11278/2506-11278-0019.flac", "task_type": "lexical_detection", "query": "disposition", "prompt": "Listen carefully to the entire clip, then report every instance of the word \"disposition\" as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[3.38, 4.16, "disposition"]], "response_mode": "offline", "answer": "[{\"word\":\"disposition\",\"start\":3.38,\"end\":4.16}]"} {"utterance_id": "4323-13259-0004", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/4323/13259/4323-13259-0004.flac", "task_type": "lexical_detection", "query": "raymond", "prompt": "Do not answer until the audio ends. Then list each instance of the word \"raymond\" as {\"word\", \"start\", \"end\"}. If none, reply [].", "labels": [[1.37, 1.77, "raymond"]], "response_mode": "offline", "answer": "[{\"word\":\"raymond\",\"start\":1.37,\"end\":1.77}]"} {"utterance_id": "5849-50964-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50964/5849-50964-0008.flac", "task_type": "lexical_detection", "query": "bert", "prompt": "Once the audio has finished, identify every instance of the word \"bert\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.99, 2.34, "bert"]], "response_mode": "offline", "answer": "[{\"word\":\"bert\",\"start\":1.99,\"end\":2.34}]"} {"utterance_id": "6455-66379-0000", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6455/66379/6455-66379-0000.flac", "task_type": "semantic_detection", "query": "profession", "prompt": "After listening to the whole recording, output every instance of a profession with precise start and end times (seconds). Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.5, 0.9, "colonel"]], "response_mode": "offline", "answer": "[{\"word\":\"colonel\",\"start\":0.5,\"end\":0.9}]"} {"utterance_id": "7641-96252-0008", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7641/96252/7641-96252-0008.flac", "task_type": "lexical_detection", "query": "inevitable", "prompt": "Once the audio has finished, identify every instance of the word \"inevitable\". Return a JSON list of {\"word\", \"start\", \"end\"} in time order. If none, reply [].", "labels": [[1.68, 2.31, "inevitable"]], "response_mode": "offline", "answer": "[{\"word\":\"inevitable\",\"start\":1.68,\"end\":2.31}]"} {"utterance_id": "6123-59186-0027", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/6123/59186/6123-59186-0027.flac", "task_type": "lexical_detection", "query": "slain", "prompt": "Listen to the full audio. After it finishes, find every instance of the word \"slain\". For each match, report the spoken word and its start/end times in seconds. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[8.27, 8.84, "slain"]], "response_mode": "offline", "answer": "[{\"word\":\"slain\",\"start\":8.27,\"end\":8.84}]"} {"utterance_id": "3660-6517-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/3660/6517/3660-6517-0005.flac", "task_type": "semantic_counting", "query": "profession", "prompt": "Do not answer until the audio ends. Then return a JSON list of count events for a profession with word/start/end. If none, reply [].", "labels": [[3.02, 3.56, "one"], [4.94, 5.21, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":3.02,\"end\":3.56},{\"word\":\"two\",\"start\":4.94,\"end\":5.21}]"} {"utterance_id": "5849-50873-0016", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/5849/50873/5849-50873-0016.flac", "task_type": "semantic_detection", "query": "kinship", "prompt": "Listen carefully to the entire clip, then report every instance of a family relation as JSON objects with word, start, and end. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[1.61, 2.09, "uncle"]], "response_mode": "offline", "answer": "[{\"word\":\"uncle\",\"start\":1.61,\"end\":2.09}]"} {"utterance_id": "7601-101622-0005", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/7601/101622/7601-101622-0005.flac", "task_type": "lexical_detection", "query": "beasts", "prompt": "You will hear a complete utterance. When it ends, list every instance of the word \"beasts\" you heard with word-level timestamps. Reply with JSON only: a list of objects {\"word\": , \"start\": , \"end\": } in chronological order. If none, reply [].", "labels": [[0.66, 1.29, "beasts"]], "response_mode": "offline", "answer": "[{\"word\":\"beasts\",\"start\":0.66,\"end\":1.29}]"} {"utterance_id": "116-288045-0032", "split": "dev-other", "audio_path": "LibriSpeech/dev-other/116/288045/116-288045-0032.flac", "task_type": "semantic_counting", "query": "person_name", "prompt": "Listen to the audio in full, then emit a JSON array of count events for a person's name (word, start, end). If none, reply [].", "labels": [[1.83, 2.19, "one"], [5.0, 5.45, "two"]], "response_mode": "offline", "answer": "[{\"word\":\"one\",\"start\":1.83,\"end\":2.19},{\"word\":\"two\",\"start\":5.0,\"end\":5.45}]"}