Dataset Viewer
Auto-converted to Parquet Duplicate
id
large_stringlengths
23
28
split
large_stringclasses
1 value
modality
large_stringclasses
1 value
turn_type
large_stringclasses
1 value
dialogue
listlengths
1
8
input
dict
streaming
dict
target
dict
taxonomy
dict
quality
dict
source
dict
metadata
dict
audio_spokenwoz_v02_000001
train
audio
multi_turn
[ { "role": "user", "text": "i think i just want to you know make sure i could know about the entrance fee attraction type and phone number i don't really", "turn_id": 1 }, { "role": "assistant", "text": "care about which type", "turn_id": 2 }, { "role": "user", "text": "okay."...
{ "chunk_text": "Chunk 1: i think i just want to you know make sure i could know about the entrance fee attraction type and phone number i don't really\nChunk 2: care about which type\nChunk 3: okay.\nChunk 4: oh, great.\nChunk 5: um excellent.\nChunk 6: okay, got it. thank you.\nChunk 7: yes, i'm also looking for a ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "i think i just want to you know make sure i could know about the entrance fee attraction type and phone number i don't really", "role": "user", "streaming_reasoning": "T...
{ "answer": "Sure. I can help with the pricerange, food; first I need to identify the right restaurant option.", "reasoning": "Across the dialogue, the user is searching for restaurant. Key context: i think i just want to you know make sure i could know about the entrance fee attraction type and phone number i don'...
{ "category": "daily_life", "subcategory": "personal_preference", "difficulty": "medium", "intent_type": "opinion_sharing" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL0661", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000002
train
audio
multi_turn
[ { "role": "user", "text": "okay. um i also need some information on the place to build in town.", "turn_id": 1 }, { "role": "assistant", "text": "i'm looking for a trunk that serves a great curry located on the north end of town.", "turn_id": 2 }, { "role": "user", "text": "w...
{ "chunk_text": "Chunk 1: okay. um i also need some information on the place to build in town.\nChunk 2: i'm looking for a trunk that serves a great curry located on the north end of town.\nChunk 3: who. can i reserve a table for their place for four people on tuesday at 1215.\nChunk 4: ah sure. it's michael.\nChunk ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "okay. um i also need some information on the place to build in town.", "role": "user", "streaming_reasoning": "The user mentions talking about Okay. It is not yet clear ...
{ "answer": "I can help book that. Could you provide the departure, destination, time?", "reasoning": "Across the dialogue, the user is searching for trunk that serves a great curry located on the north end of town. Key context: um i also need some information on the place to build in town; i'm looking for a trunk ...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2363", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000001
train
audio
multi_turn
[ { "role": "user", "text": "Human: That's correct. The symbols and written script from the movie are based on an acutal script called nsibdi.", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: Amazing, I love when things are historical in a movie because it makes it more believable....
{ "chunk_text": "Chunk 1: Human: That's correct. The symbols and written script from the movie are based on an acutal script called nsibdi.\nChunk 2: Assistant: Amazing, I love when things are historical in a movie because it makes it more believable. Chadwick Boseman was accepted in to a prestigious summer theater p...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: That's correct. The symbols and written script from the movie are based on an acutal script called nsibdi.", "role": "user", "streaming_reasoning": "The user ment...
{ "answer": "Oh wow, noway! I didn't know that part of the story. How cool would that be!", "reasoning": "Across the dialogue, the user is asking for help with nsibdi. Key context: Human: That's correct; The symbols and written script from the movie are based on an acutal script called nsibdi; Assistant: Amazing, I...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_f106cb12-02dc-45b0-bde8-f06360172d9c", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000003
train
audio
multi_turn
[ { "role": "user", "text": "hello. hello.", "turn_id": 1 }, { "role": "assistant", "text": "hi, i'm traveling in cambridge and i'm looking for a particular attraction.", "turn_id": 2 }, { "role": "user", "text": "i'm looking for a traction called, um. single world is cinema.",...
{ "chunk_text": "Chunk 1: hello. hello.\nChunk 2: hi, i'm traveling in cambridge and i'm looking for a particular attraction.\nChunk 3: i'm looking for a traction called, um. single world is cinema.\nChunk 4: can i please get the entrance free and open hours.\nChunk 5: oh okay and i'm also looking for a place to dine...
{ "num_checkpoints": 7, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello. hello.", "role": "user", "streaming_reasoning": "The user mentions Hello. hello. It is not yet clear whether the user needs the full scope of the user's request. ...
{ "answer": "I can help with the restaurant. Could you share the area?", "reasoning": "Across the dialogue, the user is searching for particular attraction. Key context: hi, i'm traveling in cambridge and i'm looking for a particular attraction; i'm looking for a traction called, um; single world is cinema; can i p...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2246", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000004
train
audio
multi_turn
[ { "role": "user", "text": "hello.", "turn_id": 1 }, { "role": "assistant", "text": "hi there. i'm signing a trip to cambridge and i'm looking for a hotel, something, uh, something cheap, like some cheap guest house.", "turn_id": 2 } ]
{ "chunk_text": "Chunk 1: hello.\nChunk 2: hi there. i'm signing a trip to cambridge and i'm looking for a hotel, something, uh, something cheap, like some cheap guest house.", "dialogue_text": "Chunk 1: hello.\nChunk 2: hi there. i'm signing a trip to cambridge and i'm looking for a hotel, something, uh, something...
{ "num_checkpoints": 2, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello.", "role": "user", "streaming_reasoning": "The user mentions Hello. It is not yet clear whether the user needs the full scope of the user's request. The assistant ...
{ "answer": "I can help with the hotel. Could you share the area?", "reasoning": "Across the dialogue, the user is searching for hotel. Key context: i'm signing a trip to cambridge and i'm looking for a hotel, something, uh, something cheap, like some cheap guest house. The answer should focus on the latest user tu...
{ "category": "daily_life", "subcategory": "home_task", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2693", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000005
train
audio
multi_turn
[ { "role": "user", "text": "hi, how are you doing today.", "turn_id": 1 }, { "role": "assistant", "text": "so i'm not. came for a place to stay in cambridge. can you help me with that.", "turn_id": 2 }, { "role": "user", "text": "so i prefer it in the cheap price range and loc...
{ "chunk_text": "Chunk 1: hi, how are you doing today.\nChunk 2: so i'm not. came for a place to stay in cambridge. can you help me with that.\nChunk 3: so i prefer it in the cheap price range and located in the west.", "dialogue_text": "Chunk 1: hi, how are you doing today.\nChunk 2: so i'm not. came for a place t...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hi, how are you doing today.", "role": "user", "streaming_reasoning": "The user mentions Hi, how are you doing today. It is not yet clear whether the user needs the full...
{ "answer": "I can help with the hotel. Could you share the hotel_type?", "reasoning": "Across the dialogue, the user is searching for hi, how are you doing today. Key context: hi, how are you doing today; came for a place to stay in cambridge; can you help me with that; so i prefer it in the cheap price range and ...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL1807", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000006
train
audio
multi_turn
[ { "role": "user", "text": "you are most welcome.", "turn_id": 1 }, { "role": "assistant", "text": "all right. and what is the entrance.", "turn_id": 2 }, { "role": "user", "text": "and may i get the phone number at.", "turn_id": 3 }, { "role": "assistant", "te...
{ "chunk_text": "Chunk 1: you are most welcome.\nChunk 2: all right. and what is the entrance.\nChunk 3: and may i get the phone number at.\nChunk 4: so what about their operating hours.\nChunk 5: oh, those who use really sex.\nChunk 6: ok, i will do that.\nChunk 7: no, i also need to find a restaurant.\nChunk 8: i'm...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "you are most welcome.", "role": "user", "streaming_reasoning": "The user mentions You are most welcome. It is not yet clear whether the user needs the full scope of the ...
{ "answer": "I can help with the restaurant. Could you share the area?", "reasoning": "Across the dialogue, the user is searching for to find a restaurant. Key context: you are most welcome; and what is the entrance; and may i get the phone number at; so what about their operating hours. The answer should focus on ...
{ "category": "information_extraction", "subcategory": "pros_cons_extraction", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3777", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000007
train
audio
multi_turn
[ { "role": "user", "text": "that's correct.", "turn_id": 1 }, { "role": "assistant", "text": "all right, could you provide me with the star rating of this hotel. is post code and the hotel type.", "turn_id": 2 }, { "role": "user", "text": "okay. is a three star rating hotel.",...
{ "chunk_text": "Chunk 1: that's correct.\nChunk 2: all right, could you provide me with the star rating of this hotel. is post code and the hotel type.\nChunk 3: okay. is a three star rating hotel.\nChunk 4: how about this post code.\nChunk 5: china, britain.\nChunk 6: to one.\nChunk 7: england, new zealand so c b2 ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "that's correct.", "role": "user", "streaming_reasoning": "The user mentions That's correct. It is not yet clear whether the user needs the full scope of the user's reque...
{ "answer": "Sure. I can provide the stars, postcode, type for the matching hotel option.", "reasoning": "Across the dialogue, the user is talking about that's correct. Key context: that's correct; all right, could you provide me with the star rating of this hotel; is post code and the hotel type; is a three star r...
{ "category": "daily_life", "subcategory": "home_task", "difficulty": "medium", "intent_type": "task_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3316", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000008
train
audio
multi_turn
[ { "role": "user", "text": "0, 3, 7. 9 8 8.", "turn_id": 1 }, { "role": "assistant", "text": "3 0, 2.", "turn_id": 2 }, { "role": "user", "text": "4 1 8.", "turn_id": 3 }, { "role": "assistant", "text": "3 0, 6, 5.", "turn_id": 4 }, { "role": "user"...
{ "chunk_text": "Chunk 1: 0, 3, 7. 9 8 8.\nChunk 2: 3 0, 2.\nChunk 3: 4 1 8.\nChunk 4: 3 0, 6, 5.\nChunk 5: i'm also looking for a place to go in town. it should not be in the east. and should be a cinema.\nChunk 6: how about a nightclub.\nChunk 7: and what's their entrance fee.\nChunk 8: oh, okay. thank you.", "di...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "0, 3, 7. 9 8 8.", "role": "user", "streaming_reasoning": "The user mentions 0, 3, 7. 9 8 8. It is not yet clear whether the user needs the full scope of the user's reque...
{ "answer": "Sure. I can provide the fee for the matching attraction option.", "reasoning": "Across the dialogue, the user is searching for place to go in town. Key context: i'm also looking for a place to go in town; it should not be in the east; and should be a cinema; how about a nightclub. The answer should foc...
{ "category": "daily_life", "subcategory": "home_task", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3008", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000009
train
audio
multi_turn
[ { "role": "user", "text": "hello.", "turn_id": 1 }, { "role": "assistant", "text": "oh, that would that would be great. um, i'd like to have email.", "turn_id": 2 }, { "role": "user", "text": "hi, i'm traveling in cambridge and so can you help me find a place to sleep.", ...
{ "chunk_text": "Chunk 1: hello.\nChunk 2: oh, that would that would be great. um, i'd like to have email.\nChunk 3: hi, i'm traveling in cambridge and so can you help me find a place to sleep.\nChunk 4: i need free parking lot and. um i think the hotel should have four star rank.\nChunk 5: also, i need in a rather. ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello.", "role": "user", "streaming_reasoning": "The user mentions Hello. It is not yet clear whether the user needs the full scope of the user's request. The assistant ...
{ "answer": "Sure. I can provide the type, phone for the matching hotel option.", "reasoning": "Across the dialogue, the user is searching for free parking lot and. Key context: oh, that would that would be great; um, i'd like to have email; hi, i'm traveling in cambridge and so can you help me find a place to slee...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "hard", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL0226", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000002
train
audio
multi_turn
[ { "role": "user", "text": "Human: That is pretty cool. I saw that Stan Lee ends every phone call he has with the word excelsior.", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: That's what I just said, lol. Have you seen the Garfield comic, Pet Force?", "turn_id": 2 }, {...
{ "chunk_text": "Chunk 1: Human: That is pretty cool. I saw that Stan Lee ends every phone call he has with the word excelsior.\nChunk 2: Assistant: That's what I just said, lol. Have you seen the Garfield comic, Pet Force?\nChunk 3: Human: I have not. What is it?", "dialogue_text": "Chunk 1: Human: That is pretty ...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: That is pretty cool. I saw that Stan Lee ends every phone call he has with the word excelsior.", "role": "user", "streaming_reasoning": "The user mentions talking...
{ "answer": "Garfield and his friends are super heroes. It's cute. There is a superhero called \"normalman\", he lives on a planet full of superheroes, and yet he has none.", "reasoning": "Across the dialogue, the user discusses \"That is pretty cool. I saw that Stan Lee ends every phone.\" over 3 turns. The answer...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "fact_lookup" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_31ae7fc7-a365-4adc-ac24-5465cc6c7414", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000010
train
audio
multi_turn
[ { "role": "user", "text": "hello.", "turn_id": 1 }, { "role": "assistant", "text": "i'm good. how are you.", "turn_id": 2 }, { "role": "user", "text": "i'm looking for a train.", "turn_id": 3 }, { "role": "assistant", "text": "the departing from cambridge.", ...
{ "chunk_text": "Chunk 1: hello.\nChunk 2: i'm good. how are you.\nChunk 3: i'm looking for a train.\nChunk 4: the departing from cambridge.\nChunk 5: um we should go to london team's car.\nChunk 6: on tuesday.\nChunk 7: um if it can leave off this quarter to 12.", "dialogue_text": "Chunk 1: hello.\nChunk 2: i'm go...
{ "num_checkpoints": 7, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello.", "role": "user", "streaming_reasoning": "The user mentions Hello. It is not yet clear whether the user needs the full scope of the user's request. The assistant ...
{ "answer": "I can help find a matching train option with those details.", "reasoning": "Across the dialogue, the user is searching for train. Key context: how are you; i'm looking for a train; the departing from cambridge; um we should go to london team's car. The answer should focus on the latest user turn while ...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2854", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000011
train
audio
multi_turn
[ { "role": "user", "text": "good day. how are you.", "turn_id": 1 }, { "role": "assistant", "text": "okay, thank you. i'd like to book a hotel. a three star hotel in the opposite side of cambridge south in cheap price range.", "turn_id": 2 }, { "role": "user", "text": "ok, can...
{ "chunk_text": "Chunk 1: good day. how are you.\nChunk 2: okay, thank you. i'd like to book a hotel. a three star hotel in the opposite side of cambridge south in cheap price range.\nChunk 3: ok, can you please submit one inexpensive. price range.\nChunk 4: yes. uh, my name is william.\nChunk 5: um can i give you my...
{ "num_checkpoints": 5, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "good day. how are you.", "role": "user", "streaming_reasoning": "The user mentions Good day. how are you. It is not yet clear whether the user needs the full scope of th...
{ "answer": "I can help book that. Could you provide the hotel_type, day, stay_length?", "reasoning": "Across the dialogue, the user is wanting to book hotel. Key context: how are you; okay, thank you; i'd like to book a hotel; a three star hotel in the opposite side of cambridge south in cheap price range. The ans...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL0854", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000003
train
audio
multi_turn
[ { "role": "user", "text": "Human: That is interesting. I imagine the sequel coming out will be a very popular movie. Do the same actors play in each movie?", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: No. sometimes ant man will make an appearance in captain america movie, but...
{ "chunk_text": "Chunk 1: Human: That is interesting. I imagine the sequel coming out will be a very popular movie. Do the same actors play in each movie?\nChunk 2: Assistant: No. sometimes ant man will make an appearance in captain america movie, but not all of them are in every movie. Infinity war will be super pop...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: That is interesting. I imagine the sequel coming out will be a very popular movie. Do the same actors play in each movie?", "role": "user", "streaming_reasoning":...
{ "answer": "Yes, a huge falling out, the group was divided. This is an interesting fact: When robert downey jr. was on the set of the avengers he would hide food on the sets.", "reasoning": "Across the dialogue, the user is asking for help with for help is the. Key context: Human: That is interesting; I imagine th...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "hard", "intent_type": "information_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_529a20c2-af07-415d-8b27-4d161f5ce475", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000012
train
audio
multi_turn
[ { "role": "user", "text": "hi, i'm looking for some information.", "turn_id": 1 }, { "role": "assistant", "text": "i'm looking to attend some sort of college attraction while i'm in town in cambridge.", "turn_id": 2 }, { "role": "user", "text": "oh, okay. by chance. do you ha...
{ "chunk_text": "Chunk 1: hi, i'm looking for some information.\nChunk 2: i'm looking to attend some sort of college attraction while i'm in town in cambridge.\nChunk 3: oh, okay. by chance. do you happen to know the operating hours of the college.\nChunk 4: okay. and do you know the address of the college.", "dial...
{ "num_checkpoints": 4, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hi, i'm looking for some information.", "role": "user", "streaming_reasoning": "The user mentions Hi, i'm looking for some information. It is not yet clear whether the u...
{ "answer": "Sure. I can provide the address for the matching attraction option.", "reasoning": "Across the dialogue, the user is searching for information. Key context: hi, i'm looking for some information; i'm looking to attend some sort of college attraction while i'm in town in cambridge; do you happen to know ...
{ "category": "tech_support", "subcategory": "device_troubleshooting", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "SNG1017", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000013
train
audio
multi_turn
[ { "role": "user", "text": "hi, my name is maggie. i'm planning a trip to camera.", "turn_id": 1 }, { "role": "assistant", "text": "i am looking for a train from cambridge to standard airport.", "turn_id": 2 }, { "role": "user", "text": "sure thing. thank you.", "turn_id":...
{ "chunk_text": "Chunk 1: hi, my name is maggie. i'm planning a trip to camera.\nChunk 2: i am looking for a train from cambridge to standard airport.\nChunk 3: sure thing. thank you.\nChunk 4: is there a train available on thursday that will arrive by quarter past 5 p m.\nChunk 5: oh, excellent. can i please have th...
{ "num_checkpoints": 5, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hi, my name is maggie. i'm planning a trip to camera.", "role": "user", "streaming_reasoning": "The user mentions Hi, my name is maggie. i'm planning a trip to camera. I...
{ "answer": "Sure. I can provide the travel_time for the matching train option.", "reasoning": "Across the dialogue, the user is searching for train from cambridge to standard airport. Key context: hi, my name is maggie; i'm planning a trip to camera; i am looking for a train from cambridge to standard airport; is ...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2087", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000004
train
audio
multi_turn
[ { "role": "user", "text": "Human: Hello, have you heard of the possibility of stimulating the brain with zaps to decrease aggressive instincts?", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: That is quite amazing, like something from science fiction. I just wonder if it will re...
{ "chunk_text": "Chunk 1: Human: Hello, have you heard of the possibility of stimulating the brain with zaps to decrease aggressive instincts?\nChunk 2: Assistant: That is quite amazing, like something from science fiction. I just wonder if it will really translate to real life. It seems the test was to zap the prefr...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: Hello, have you heard of the possibility of stimulating the brain with zaps to decrease aggressive instincts?", "role": "user", "streaming_reasoning": "The user m...
{ "answer": "In some ways I agree, like creating this computerized device to electrically stimulate the brain. Its a bit scary that external devices can manipulate the brain's wiring.", "reasoning": "Across the dialogue, the user is asking for help with human: hello, have you heard of the possibility of stimulating...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "hard", "intent_type": "information_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_2feb802c-49da-49c7-bf30-fe333035c4f2", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000014
train
audio
multi_turn
[ { "role": "user", "text": "sure. it's 0 4, 4.", "turn_id": 1 }, { "role": "assistant", "text": "9, 7, 1.", "turn_id": 2 }, { "role": "user", "text": "3, 6, 3, 6.", "turn_id": 3 }, { "role": "assistant", "text": "and i would like a taxi to commute between those...
{ "chunk_text": "Chunk 1: sure. it's 0 4, 4.\nChunk 2: 9, 7, 1.\nChunk 3: 3, 6, 3, 6.\nChunk 4: and i would like a taxi to commute between those two locations and i would like the taxi to arrive at a restaurant reservation time\nChunk 5: of 4 pm\nChunk 6: and do you know what type of car the taxi is.\nChunk 7: okay, ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "sure. it's 0 4, 4.", "role": "user", "streaming_reasoning": "The user mentions Sure. it's 0 4, 4. It is not yet clear whether the user needs the full scope of the user's...
{ "answer": "Sure. I can provide the car for the matching taxi option.", "reasoning": "Across the dialogue, the user is talking about taxi to commute between those two locations. Key context: it's 0 4, 4; and i would like a taxi to commute between those two locations and i would like the taxi to arrive at a restaur...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3049", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000015
train
audio
multi_turn
[ { "role": "user", "text": "hello.", "turn_id": 1 }, { "role": "assistant", "text": "can you give a particular hotel.", "turn_id": 2 }, { "role": "user", "text": "it's called the cambridge, um. both um frank.", "turn_id": 3 }, { "role": "assistant", "text": "ca...
{ "chunk_text": "Chunk 1: hello.\nChunk 2: can you give a particular hotel.\nChunk 3: it's called the cambridge, um. both um frank.\nChunk 4: can i make a booking.\nChunk 5: i like the book for seven people.", "dialogue_text": "Chunk 1: hello.\nChunk 2: can you give a particular hotel.\nChunk 3: it's called the cam...
{ "num_checkpoints": 5, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello.", "role": "user", "streaming_reasoning": "The user mentions Hello. It is not yet clear whether the user needs the full scope of the user's request. The assistant ...
{ "answer": "I can help book that. Could you provide the area, price, hotel_type?", "reasoning": "Across the dialogue, the user is wanting to book for seven people. Key context: can you give a particular hotel; it's called the cambridge, um; both um frank; can i make a booking. The answer should focus on the latest...
{ "category": "daily_life", "subcategory": "home_task", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "SNG0963", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000016
train
audio
multi_turn
[ { "role": "user", "text": "good day. good day.", "turn_id": 1 }, { "role": "assistant", "text": "could you please assist me in finding some places to go to in town.", "turn_id": 2 }, { "role": "user", "text": "yes, it should be any type of a cinema and also be. um located in ...
{ "chunk_text": "Chunk 1: good day. good day.\nChunk 2: could you please assist me in finding some places to go to in town.\nChunk 3: yes, it should be any type of a cinema and also be. um located in the south.\nChunk 4: okay, not a problem.\nChunk 5: okay. um. that sounds fine. fine. may i please get the entrance fe...
{ "num_checkpoints": 6, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "good day. good day.", "role": "user", "streaming_reasoning": "The user mentions Good day. good day. It is not yet clear whether the user needs the full scope of the user...
{ "answer": "I can help with the hotel. Could you share the area, price, hotel_type?", "reasoning": "Across the dialogue, the user is searching for place to stay too. Key context: could you please assist me in finding some places to go to in town; yes, it should be any type of a cinema and also be; um located in th...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "hard", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3801", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000005
train
audio
multi_turn
[ { "role": "user", "text": "Human: I think so. Darnold is not the answer in NY in my opinion.", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: Some seem curious as to why the Jets did not want to fork over the money for a sure thing since they had the funds.", "turn_id": 2 }...
{ "chunk_text": "Chunk 1: Human: I think so. Darnold is not the answer in NY in my opinion.\nChunk 2: Assistant: Some seem curious as to why the Jets did not want to fork over the money for a sure thing since they had the funds.\nChunk 3: Human: agree. especially seeing how much Garoppolo got for only 7 NFL starts. s...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: I think so. Darnold is not the answer in NY in my opinion.", "role": "user", "streaming_reasoning": "The user mentions talking about Human, Darnold. It is not yet...
{ "answer": "Minnesota was predicted to be themost appealing destination for Cousins.", "reasoning": "Across the dialogue, the user discusses \"I think so. Darnold is not the answer in NY in my.\" over 3 turns. The answer should respond to the latest user turn naturally, maintaining the conversational tone.", "re...
{ "category": "information_extraction", "subcategory": "long_note_to_summary", "difficulty": "medium", "intent_type": "opinion_sharing" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_aa8b3134-2b25-49f9-a0e6-dabff137a5e4", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000017
train
audio
multi_turn
[ { "role": "user", "text": "hello, how are you.", "turn_id": 1 }, { "role": "assistant", "text": "i'm good. thank you. um i'd like to book a first a hotel with an occupied packing spot.", "turn_id": 2 }, { "role": "user", "text": "i i would like to book for four people.", ...
{ "chunk_text": "Chunk 1: hello, how are you.\nChunk 2: i'm good. thank you. um i'd like to book a first a hotel with an occupied packing spot.\nChunk 3: i i would like to book for four people.\nChunk 4: uh, starting from sunday.", "dialogue_text": "Chunk 1: hello, how are you.\nChunk 2: i'm good. thank you. um i'd...
{ "num_checkpoints": 4, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello, how are you.", "role": "user", "streaming_reasoning": "The user mentions Hello, how are you. It is not yet clear whether the user needs the full scope of the user...
{ "answer": "I can help book that. Could you provide the area, price, stay_length?", "reasoning": "Across the dialogue, the user is wanting to book first a hotel with an occupied packing spot. Key context: hello, how are you; um i'd like to book a first a hotel with an occupied packing spot; i i would like to book ...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL1961", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000018
train
audio
multi_turn
[ { "role": "user", "text": "um but you give food.", "turn_id": 1 }, { "role": "assistant", "text": "okay. how about one that serves seafood.", "turn_id": 2 }, { "role": "user", "text": "okay. um, it should be in the expensive price range and in the center.", "turn_id": 3 ...
{ "chunk_text": "Chunk 1: um but you give food.\nChunk 2: okay. how about one that serves seafood.\nChunk 3: okay. um, it should be in the expensive price range and in the center.\nChunk 4: okay. can i please make a a booking that, uh, for the same number of people.\nChunk 5: emma liyi. double m.\nChunk 6: a.\nChunk ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "um but you give food.", "role": "user", "streaming_reasoning": "The user mentions Um but you give food. It is not yet clear whether the user needs the full scope of the ...
{ "answer": "I can help book that. Could you provide the day, time?", "reasoning": "Across the dialogue, the user is wanting to book um but you give food. Key context: um but you give food; how about one that serves seafood; um, it should be in the expensive price range and in the center; can i please make a a book...
{ "category": "daily_life", "subcategory": "shopping_decision", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2845", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000019
train
audio
multi_turn
[ { "role": "user", "text": "the train should leave on tuesday.", "turn_id": 1 }, { "role": "assistant", "text": "and to depart. from the legendary university.", "turn_id": 2 }, { "role": "user", "text": "anything arriving by 8 45.", "turn_id": 3 }, { "role": "assis...
{ "chunk_text": "Chunk 1: the train should leave on tuesday.\nChunk 2: and to depart. from the legendary university.\nChunk 3: anything arriving by 8 45.\nChunk 4: um can i book for 7 people, please.\nChunk 5: 8 0 5 oh sorry, 5 0 2.\nChunk 6: 2, 8, 7.\nChunk 7: 1 7 2.\nChunk 8: 1, 2, 0.", "dialogue_text": "Chunk 1:...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "the train should leave on tuesday.", "role": "user", "streaming_reasoning": "The user mentions The train should leave on tuesday. It is not yet clear whether the user ne...
{ "answer": "I can help book that. Could you provide the destination?", "reasoning": "Across the dialogue, the user is wanting to book for 7 people. Key context: the train should leave on tuesday; and to depart; from the legendary university; anything arriving by 8 45. The answer should focus on the latest user tur...
{ "category": "education", "subcategory": "course_planning", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3011", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000006
train
audio
multi_turn
[ { "role": "user", "text": "Human: That is very true. I wish that our government would implement some type of gun control or background checks.", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: I believe firearms are big business in America the governments do not want to stop the m...
{ "chunk_text": "Chunk 1: Human: That is very true. I wish that our government would implement some type of gun control or background checks.\nChunk 2: Assistant: I believe firearms are big business in America the governments do not want to stop the money train.\nChunk 3: Human: I feel the exact same way. It is terri...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: That is very true. I wish that our government would implement some type of gun control or background checks.", "role": "user", "streaming_reasoning": "The user me...
{ "answer": "Yes mental health needs to be addessed also", "reasoning": "The dialogue is about \"That is very true. I wish that our government would.\" The user later explains \"I feel the exact same way. It is terrible that money is.\" The answer should treat this as a discussion of the topic, not as a task-planni...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "easy", "intent_type": "explain_cause" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_21e15a0a-21e6-4174-8143-8c8eb4335e5a", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000020
train
audio
multi_turn
[ { "role": "user", "text": "hello, i'm looking for some information.", "turn_id": 1 }, { "role": "assistant", "text": "i'm looking for place to dine.", "turn_id": 2 }, { "role": "user", "text": "it should serve with grant.", "turn_id": 3 }, { "role": "assistant", ...
{ "chunk_text": "Chunk 1: hello, i'm looking for some information.\nChunk 2: i'm looking for place to dine.\nChunk 3: it should serve with grant.\nChunk 4: and not necessarily only the moderate price range.\nChunk 5: what about modern european food.\nChunk 6: i would like to book it for two people.", "dialogue_text...
{ "num_checkpoints": 6, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello, i'm looking for some information.", "role": "user", "streaming_reasoning": "The user mentions Hello, i'm looking for some information. It is not yet clear whether...
{ "answer": "I can help book that. Could you provide the area, day, time?", "reasoning": "Across the dialogue, the user is searching for information. Key context: hello, i'm looking for some information; i'm looking for place to dine; it should serve with grant; and not necessarily only the moderate price range. Th...
{ "category": "daily_life", "subcategory": "shopping_decision", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3125", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_sdeval_v02_000001
train
audio
multi_turn
[ { "role": "user", "text": "How can I maintain my fitness routine when it's raining non-stop?", "turn_id": 1 } ]
{ "chunk_text": "Chunk 1: How can I maintain my fitness routine when it's raining non-stop?", "dialogue_text": "Chunk 1: How can I maintain my fitness routine when it's raining non-stop?", "instruction": "Help the user complete a spoken task-oriented dialogue based on gradually revealed audio/transcript chunks.",...
{ "num_checkpoints": 1, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "How can I maintain my fitness routine when it's raining non-stop?", "role": "user", "streaming_reasoning": "The user mentions How can I maintain my fitness routine when ...
{ "answer": "Consider indoor workouts such as yoga, pilates, or using an exercise app.", "reasoning": "Across the dialogue, the user discusses \"How can I maintain my fitness routine when it's raining.\" over 1 turns. The answer should respond to the latest user turn naturally, maintaining the conversational tone."...
{ "category": "information_extraction", "subcategory": "constraints_extraction", "difficulty": "easy", "intent_type": "information_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SD-Eval", "dialogue_id": "89_4", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000021
train
audio
multi_turn
[ { "role": "user", "text": "hi, hello. is that customer service center.", "turn_id": 1 }, { "role": "assistant", "text": "yes, i'm looking for a taxi. could you help me to book 1.", "turn_id": 2 } ]
{ "chunk_text": "Chunk 1: hi, hello. is that customer service center.\nChunk 2: yes, i'm looking for a taxi. could you help me to book 1.", "dialogue_text": "Chunk 1: hi, hello. is that customer service center.\nChunk 2: yes, i'm looking for a taxi. could you help me to book 1.", "instruction": "Help the user com...
{ "num_checkpoints": 2, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hi, hello. is that customer service center.", "role": "user", "streaming_reasoning": "The user mentions Hi, hello. is that customer service center. It is not yet clear w...
{ "answer": "I can help with the taxi. Could you share the departure, destination, time?", "reasoning": "Across the dialogue, the user is searching for taxi. Key context: is that customer service center; yes, i'm looking for a taxi; could you help me to book 1. The answer should focus on the latest user turn while ...
{ "category": "finance_consumer", "subcategory": "budget_planning", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "SNG0400", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000007
train
audio
multi_turn
[ { "role": "user", "text": "Human: Crazy stuff! I recently learned there was a fantasy league for congress where players draft congressmen and get points for legislative successes. I cannot even imagine!!!", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: That would actually be pre...
{ "chunk_text": "Chunk 1: Human: Crazy stuff! I recently learned there was a fantasy league for congress where players draft congressmen and get points for legislative successes. I cannot even imagine!!!\nChunk 2: Assistant: That would actually be pretty fun. There are even three subway lines exclusively for Congress...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: Crazy stuff! I recently learned there was a fantasy league for congress where players draft congressmen and get points for legislative successes. I cannot even imagine!!!", ...
{ "answer": "I agree. Until 1998, British politicians had to wear a top hat if making a point of order in the House of Commons. I wonder if a hat maker made that rule?", "reasoning": "Across the dialogue, the user discusses \"Crazy stuff! I recently learned there was a fantasy league.\" over 3 turns. The answer sho...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_dd15c85d-001b-4fa8-8a7f-47b5cc2261d6", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_multidialog_v02_000008
train
audio
multi_turn
[ { "role": "user", "text": "Human: Yeah i don't get to watch it much anymore though. I hear bowlers used to make more than top nfl players in the 60's", "turn_id": 1 }, { "role": "assistant", "text": "Assistant: Thats amazing to learn. I think that has changed just a little.", "turn_id": ...
{ "chunk_text": "Chunk 1: Human: Yeah i don't get to watch it much anymore though. I hear bowlers used to make more than top nfl players in the 60's\nChunk 2: Assistant: Thats amazing to learn. I think that has changed just a little.\nChunk 3: Human: Yeah as for Tom Brady he swears by Avocado ice cream.", "dialogue...
{ "num_checkpoints": 3, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "Human: Yeah i don't get to watch it much anymore though. I hear bowlers used to make more than top nfl players in the 60's", "role": "user", "streaming_reasoning": "The ...
{ "answer": "I think I read that somewhere. i never actually tried it though. Is it good.", "reasoning": "Across the dialogue, the user discusses \"Yeah i don't get to watch it much anymore though. I hear.\" over 3 turns. The answer should respond to the latest user turn naturally, maintaining the conversational to...
{ "category": "daily_life", "subcategory": "home_task", "difficulty": "medium", "intent_type": "opinion_sharing" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "MultiDialog", "dialogue_id": "t_8a13af8f-8f53-47d2-8ab2-76e6beed9fa5", "language": "en", "license": "cc", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000022
train
audio
multi_turn
[ { "role": "user", "text": "hello, you speaking with marianne and i'm looking for some information in cambridge. can you help me.", "turn_id": 1 }, { "role": "assistant", "text": "and i'm looking to try out a few local israel.", "turn_id": 2 }, { "role": "user", "text": "i.", ...
{ "chunk_text": "Chunk 1: hello, you speaking with marianne and i'm looking for some information in cambridge. can you help me.\nChunk 2: and i'm looking to try out a few local israel.\nChunk 3: i.\nChunk 4: thank you. yes. um. yes, i'm looking for. um one in the expensive price range.\nChunk 5: okay. lover to be in ...
{ "num_checkpoints": 6, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello, you speaking with marianne and i'm looking for some information in cambridge. can you help me.", "role": "user", "streaming_reasoning": "The user mentions searchi...
{ "answer": "I can help find a matching restaurant option with those details.", "reasoning": "Across the dialogue, the user is searching for information in cambridge. Key context: hello, you speaking with marianne and i'm looking for some information in cambridge; can you help me; and i'm looking to try out a few l...
{ "category": "daily_life", "subcategory": "personal_preference", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL0609", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000023
train
audio
multi_turn
[ { "role": "user", "text": "hello, i'm looking for a trend.", "turn_id": 1 }, { "role": "assistant", "text": "the train should leave after 9 45 and should leave on wednesday. the train should depart from cambridge and should go to london kings cross.", "turn_id": 2 }, { "role": "u...
{ "chunk_text": "Chunk 1: hello, i'm looking for a trend.\nChunk 2: the train should leave after 9 45 and should leave on wednesday. the train should depart from cambridge and should go to london kings cross.\nChunk 3: may you please provide me with the travel time.\nChunk 4: can you please provide me with a travel i...
{ "num_checkpoints": 5, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello, i'm looking for a trend.", "role": "user", "streaming_reasoning": "The user mentions Hello, i'm looking for a trend. It is not yet clear whether the user needs th...
{ "answer": "Sure. I can provide the travel_time, id, arriveby for the matching train option.", "reasoning": "Across the dialogue, the user is searching for trend. Key context: hello, i'm looking for a trend; the train should leave after 9 45 and should leave on wednesday; the train should depart from cambridge and...
{ "category": "daily_life", "subcategory": "home_task", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL2673", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000024
train
audio
multi_turn
[ { "role": "user", "text": "hello.", "turn_id": 1 }, { "role": "assistant", "text": "i'm looking for a train going to cambridge.", "turn_id": 2 }, { "role": "user", "text": "please and leaving from stevenage and i'd like something leaving on wednesday october 12.", "turn_i...
{ "chunk_text": "Chunk 1: hello.\nChunk 2: i'm looking for a train going to cambridge.\nChunk 3: please and leaving from stevenage and i'd like something leaving on wednesday october 12.\nChunk 4: and i'd like it leaving after 1015.\nChunk 5: sounds great. i'd like to book it please for five people.\nChunk 6: yes, ca...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello.", "role": "user", "streaming_reasoning": "The user mentions Hello. It is not yet clear whether the user needs the full scope of the user's request. The assistant ...
{ "answer": "I can help book that with the details you gave. Would you like me to confirm the reservation?", "reasoning": "Across the dialogue, the user is searching for train going to cambridge. Key context: i'm looking for a train going to cambridge; please and leaving from stevenage and i'd like something leavin...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3385", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000025
train
audio
multi_turn
[ { "role": "user", "text": "yes, and this train should go to norwich.", "turn_id": 1 }, { "role": "assistant", "text": "and should depart from cambridge, please.", "turn_id": 2 }, { "role": "user", "text": "yes. have you talked about that.", "turn_id": 3 }, { "role...
{ "chunk_text": "Chunk 1: yes, and this train should go to norwich.\nChunk 2: and should depart from cambridge, please.\nChunk 3: yes. have you talked about that.\nChunk 4: yes, i want to get the price travel time and the train id. could you. uh, could you recommend they recommend for me.\nChunk 5: yes, of course.\nC...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "yes, and this train should go to norwich.", "role": "user", "streaming_reasoning": "The user mentions Yes, and this train should go to norwich. It is not yet clear wheth...
{ "answer": "I can help with the attraction. Could you share the area, type?", "reasoning": "Across the dialogue, the user is asking for recommendations about to get the price travel time. Key context: yes, and this train should go to norwich; and should depart from cambridge, please; have you talked about that; ye...
{ "category": "daily_life", "subcategory": "shopping_decision", "difficulty": "medium", "intent_type": "task_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3905", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000026
train
audio
multi_turn
[ { "role": "user", "text": "hello. have i reached customer service.", "turn_id": 1 }, { "role": "assistant", "text": "hi, i'm looking for some information.", "turn_id": 2 }, { "role": "user", "text": "um i'm looking for a phone number of a particular hospital department.", ...
{ "chunk_text": "Chunk 1: hello. have i reached customer service.\nChunk 2: hi, i'm looking for some information.\nChunk 3: um i'm looking for a phone number of a particular hospital department.\nChunk 4: and it's ah the hospital culinary care unit is the department that i would like to reach.\nChunk 5: oh, that is e...
{ "num_checkpoints": 6, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hello. have i reached customer service.", "role": "user", "streaming_reasoning": "The user mentions Hello. have i reached customer service. It is not yet clear whether t...
{ "answer": "I can help find a matching hospital option with those details.", "reasoning": "Across the dialogue, the user is searching for information. Key context: have i reached customer service; hi, i'm looking for some information; um i'm looking for a phone number of a particular hospital department; and it's ...
{ "category": "social_communication", "subcategory": "message_reply", "difficulty": "medium", "intent_type": "search_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "SNG1334", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000027
train
audio
multi_turn
[ { "role": "user", "text": "my name is brittany vincent. i was buying for you bri.", "turn_id": 1 }, { "role": "assistant", "text": "ta ny.", "turn_id": 2 }, { "role": "user", "text": "vi n.", "turn_id": 3 }, { "role": "assistant", "text": "s o n.", "turn_i...
{ "chunk_text": "Chunk 1: my name is brittany vincent. i was buying for you bri.\nChunk 2: ta ny.\nChunk 3: vi n.\nChunk 4: s o n.\nChunk 5: i don't need. so i'm also looking for a taxi to go with them these two places to make the attraction by.\nChunk 6: and then 15 pm.\nChunk 7: ya, it's 4 0, 7.\nChunk 8: 9 3 1.", ...
{ "num_checkpoints": 8, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "my name is brittany vincent. i was buying for you bri.", "role": "user", "streaming_reasoning": "The user mentions My name is brittany vincent. i was buying for you bri....
{ "answer": "I can help with the taxi. Could you share the destination?", "reasoning": "Across the dialogue, the user is searching for taxi to go with them these two places to make the attraction by. Key context: my name is brittany vincent; i was buying for you bri; i don't need; so i'm also looking for a taxi to ...
{ "category": "finance_consumer", "subcategory": "budget_planning", "difficulty": "medium", "intent_type": "conversational_understanding" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "MUL3213", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
audio_spokenwoz_v02_000028
train
audio
multi_turn
[ { "role": "user", "text": "hi, my name is elma garro.", "turn_id": 1 }, { "role": "assistant", "text": "hi manoeuvre. i'm looking forward to booking attacks in.", "turn_id": 2 } ]
{ "chunk_text": "Chunk 1: hi, my name is elma garro.\nChunk 2: hi manoeuvre. i'm looking forward to booking attacks in.", "dialogue_text": "Chunk 1: hi, my name is elma garro.\nChunk 2: hi manoeuvre. i'm looking forward to booking attacks in.", "instruction": "Help the user complete a spoken task-oriented dialogu...
{ "num_checkpoints": 2, "turn_checkpoints": [ { "checkpoint_id": "C1", "confidence": "medium", "observed_text": "hi, my name is elma garro.", "role": "user", "streaming_reasoning": "The user mentions Hi, my name is elma garro. It is not yet clear whether the user needs the full sco...
{ "answer": "I can help book that. Could you provide the departure, destination, time?", "reasoning": "Across the dialogue, the user is searching for hi, my name is elma garro. Key context: hi, my name is elma garro; hi manoeuvre; i'm looking forward to booking attacks in. The answer should focus on the latest user...
{ "category": "social_communication", "subcategory": "conflict_deescalation", "difficulty": "easy", "intent_type": "booking_request" }
{ "audio_quality": "standard", "is_high_quality": true, "natural_reasoning": true, "reasoning_quality": "high", "sft_ready": true, "taxonomy_quality": "high", "timestamp_quality": "standard", "target_grounded": true }
{ "audio_id": null, "dataset": "SpokenWOZ", "dialogue_id": "SNG1523", "language": "en", "license": "cc-by-nc-4.0", "turn_ids": [] }
{ "created_by": "rule_based_pipeline_with_content_grounding", "notes": null, "release_version": "vFinal" }
End of preview. Expand in Data Studio

LifeAudioMultiTurnStreamingCoT

Version: vFinal

A multi-turn audio dataset with streaming chain-of-thought reasoning for SFT. 13349 active training rows with audio references.

Schema

Top-level fields: id, split, modality, turn_type, dialogue, audio, input, streaming, target, taxonomy, quality, source, metadata

  • dialogue (list of turns with role/text)
  • audio (path, reference, duration_sec, loadability)
  • input (instruction, length_bucket)
  • target (reasoning, answer, response)
  • streaming (checkpoints with streaming_reasoning)
  • taxonomy (category, subcategory, difficulty, intent_type)

Target Format

{
  "reasoning": "Natural language reasoning about the task/input...",
  "answer": "The actual task output...",
  "response": "Reasoning: ...\n\nAnswer: ..."
}

Use target.response for SFT training. It includes both reasoning and final answer.

Quality

Metric Value
Active rows 13,349
Train 10,679
Eval 2,670
High quality 13,349
SFT-ready 100.0%
Target grounded 100.0%

High-Quality Configuration

The high_quality config contains a filtered subset of default rows where quality.sft_ready = true and quality.is_high_quality = true. It is not additional unique data.

Limitations

  • Audio is external reference only (HF URLs). No bundled audio bytes.
  • Natural-language reasoning is template-generated, not LLM-written.
  • Row counts reflect quality-filtered active splits suitable for direct SFT usage.

Usage

from datasets import load_dataset

# Load default config
ds = load_dataset("skyzhou06/LifeAudioMultiTurnStreamingCoT")

# Load high-quality subset
ds_hq = load_dataset("skyzhou06/LifeAudioMultiTurnStreamingCoT", "high_quality")
Downloads last month
69