Datasets:
id string | audio audio | instruction string | tools list | choices list | choice_labels list | answer string | answer_index int32 | answer_label string | spoken_text string | spoken_language string | tts_text string | speech_category string | source_dataset string | source_id string | source_uuid string | source_subset string | target_tool string | source_split string | spoken_unit string | spoken_unit_rows int32 | tts_model string | tts_repo string | tts_revision string | tts_sdk_revision string | voice_id string | voice_kind string | voice_split string | tts_settings string | tts_seed int64 | duration_seconds float32 | n_segments int32 | audio_gain float32 | asr_text string | asr_wer float32 | asr_metric string | asr_model string | render_attempt int32 | passes_checks bool |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
when2call-train-03150 | Which response should the assistant give to the user's question, given the available tools? | [] | [
"I apologize, but I'm unable to perform tasks such as capturing screenshots. I'm here to provide information and answer questions to the best of my ability.",
"To proceed, I need to know if you prefer a specific region or section of the Twitter homepage for the screenshot."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Capture a non-fullscreen screenshot of the Twitter homepage with a height of 400 pixels and a width of 900 pixels. | en | Capture a non-fullscreen screenshot of the Twitter homepage with a height of 400 pixels and a width of 900 pixels. | long_number | When2Call | When2Call:train:3150 | null | null | null | train | When2Call:train:3150 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | f0e873bedaa3db5495e9c9fd2247fa2bb3721408b052083114d2e7ba58b54676 | clone | train | {} | 1,165,139,142 | 8.15 | 1 | 1 | Capture a non-full screen screenshot of the Twitter homepage with a height of 400 pixels and a width of 900 pixels. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07389 | Which response should the assistant give to the user's question, given the available tools? | [] | [
"I'm sorry, I can't provide the liquidity value of Uniswap in USD as I don't have real-time data or the ability to perform such calculations.",
"As of the latest data, Uniswap's liquidity is approximately $3 billion USD."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | What is the liquidity value of Uniswap in USD? | en | What is the liquidity value of Uniswap in USD? | plain | When2Call | When2Call:train:7389 | null | null | null | train | When2Call:train:7389 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | d73589f020154c6878176a5422f308b0b797c6a9497dba2d4891d586c0a23793 | clone | train | {} | 362,797,888 | 3.52 | 1 | 1 | What is the liquidity value of Uniswap and USD? | 0.1111 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00680 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"basketball_match_statistics\", \"description\": \"Fetches statistics for a given basketball match using the RapidAPI sports live scores service.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"match_id\": {\"description\": \"The identifier for the basketball match. Match IDs can be retrieve... | [
"<TOOLCALL>[{\"name\": \"basketball_match_statistics\", \"arguments\": {\"match_id\": \"euroleague_realmadrid_cska\"}}]</TOOLCALL>",
"To provide you with the statistics, I'll need the specific match ID for the Euroleague game between Real Madrid and CSKA Moscow. Could you please provide that?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Can you fetch the stats for the Euroleague match between Real Madrid and CSKA Moscow? | en | Can you fetch the stats for the Euroleague match between Real Madrid and CSKA Moscow? | plain | When2Call | When2Call:train:680 | null | null | null | train | When2Call:train:680 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 00a91ba0f768fca249c8a99ffa5765e4db9732d6b718d02ab3b753699749aeb3 | clone | train | {} | 3,673,884,015 | 4.32 | 1 | 1 | Can you fetch the stats for the EuroLeague match between Real Madrid and CSK and Moscow? | 0.1333 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05864 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"products_search\", \"description\": \"Search for products by keywords with various filtering and sorting options.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"keywords\": {\"description\": \"The search terms or phrases to find relevant products.\", \"type\": \"str\", \"default\": \"Compu... | [
"To help you better, could you please specify the keywords for the type of dresses you're interested in?",
"I found several summer dresses within your budget. The most affordable option is a Casual Sleeveless Dress for $25.99. Would you like more details?"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Search for dresses with a maximum price of $50 and sort by price from low to high in the US store. | en | Search for dresses with a maximum price of $50 and sort by price from low to high in the US store. | plain | When2Call | When2Call:train:5864 | null | null | null | train | When2Call:train:5864 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | F2 | preset | train | {"speed": 1.05, "total_step": 5} | 3,802,711,589 | 6.888 | 1 | 1 | Search for dresses with a maximum price of $50 and sort by price from low to high in the U.S. store. | 0.1364 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00238 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"quotes_get_mashup\", \"description\": \"Fetch additional information for a specific market quote.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"symbol\": {\"description\": \"The market quote symbol for which the information is to be fetched. Only one symbol is allowed.\", \"type\": \"str\... | [
"<TOOLCALL>[{\"name\": \"symbols_press_releases\", \"arguments\": {\"ticker_slug\": \"AMZN\", \"page_number\": 2}}]</TOOLCALL>",
"Sure, I can help with that. However, could you please specify if you're interested in a particular type of press release?"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Fetch the second page of press releases for Amazon (AMZN) from Seeking Alpha. | en | Fetch the second page of press releases for Amazon (AMZN) from Seeking Alpha. | plain | When2Call | When2Call:train:238 | null | null | null | train | When2Call:train:238 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | 7fe2e4a18cf5c9c47d0be6c7254fa7f0fc7f05b44161889fec8f74eafb01e1fb | clone | train | {} | 1,734,306,104 | 4.59 | 1 | 1 | Betch the second page of press releases for Amazon MZN from Seeking Alpha. | 0.1538 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00887 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"v1_mortgagecalculator\", \"description\": \"Calculates mortgage-related information using the API Ninjas Mortgage Calculator API endpoint.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"interest_rate\": {\"description\": \"Annual interest rate (in %). For example, a 3.5% interest rate woul... | [
"To provide you with the most relevant results, could you please specify the base asset you're interested in?",
"<TOOLCALL>[{\"name\": \"market\", \"arguments\": {}}]</TOOLCALL>"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Find all available markets on Brave New Coin. | en | Find all available markets on Brave New Coin. | plain | When2Call | When2Call:train:887 | null | null | null | train | When2Call:train:887 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | 796c6d03d4d23b79f63fe25fec88093b812c3b9f8a4cf916d5c2f6b6dcdb3d21 | clone | train | {} | 2,718,128,558 | 4.213 | 1 | 1 | Find all available markets on Brave New Coin. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00500 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"stock_get_stock_profile_data\", \"description\": \"Fetches the stock profile data for a specified ticker symbol and region.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"symbol\": {\"description\": \"The ticker symbol of the stock (e.g., \\\"AAPL\\\" for Apple Inc.).\", \"type\": \"str\",... | [
"Apologies, I'm unable to provide direct access to the Walnut Trading Dashboard. Please contact your system administrator for assistance.",
"<TOOLCALL>[{\"name\": \"free_access\", \"arguments\": {\"get_dash_url\": \"https://example.com/dashboard\"}}]</TOOLCALL>"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | I need access to the Walnut Trading Dashboard via a unique URL. | en | I need access to the Walnut Trading Dashboard via a unique URL. | plain | When2Call | When2Call:train:500 | null | null | null | train | When2Call:train:500 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | c2f1136c6be619e50b107aed8c98348538d157fa38a10d1c31e06c5543fcc793 | clone | train | {} | 347,532,086 | 4 | 1 | 1 | I need access to the walnut trading dashboard via a unique URL. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07610 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_user_dialog_string\", \"description\": \"Fetches the dialog string for a given user from the Waifu API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"user_id\": {\"description\": \"A unique identifier for the user.\", \"type\": \"str\", \"default\": \"sample_user_id\"}}}, \"required\"... | [
"To assist you better, could you please specify the exact line or phrase you're interested in? In this case, is it 'To be or not to be'?",
"The poem containing the line 'To be or not to be' is William Shakespeare's Hamlet, Act 3, Scene 1. It's a soliloquy, not a standalone poem."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Retrieve all poems containing the phrase 'To be or not to be'. | en | Retrieve all poems containing the phrase 'To be or not to be'. | plain | When2Call | When2Call:train:7610 | null | null | null | train | When2Call:train:7610 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 7c81dc22e359f5ac78beeabb0a1da6ddc36fae0c32ff4f99c8e01e600360bc9d | clone | train | {} | 4,080,268,792 | 4.08 | 1 | 1 | retrieve all poems containing the phrase to be or not to be. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03598 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"topic_headlines\", \"description\": \"Fetches the latest news headlines for a specified topic.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"topic\": {\"description\": \"The topic for which to get news headlines. Available topics include:\", \"type\": \"str\", \"default\": \"WORLD\"}, \"l... | [
"<TOOLCALL>[{\"name\": \"topic_headlines\", \"arguments\": {\"topic\": \"BUSINESS\", \"lang\": \"de\", \"country\": \"DE\"}}]</TOOLCALL>",
"Hier sind die neuesten Wirtschaftsnachrichten-Schlagzeilen aus Deutschland:\n\n1. \"Deutsche Bank meldet Gewinnanstieg im zweiten Quartal\"\n2. \"BMW steigert Absatz in China... | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Fetch the latest business news headlines in Germany in German. | en | Fetch the latest business news headlines in Germany in German. | plain | When2Call | When2Call:train:3598 | null | null | null | train | When2Call:train:3598 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | 796c6d03d4d23b79f63fe25fec88093b812c3b9f8a4cf916d5c2f6b6dcdb3d21 | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 354,166,936 | 4.16 | 1 | 1 | Fetch the latest business news headlines in Germany in German. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00661 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"teamlastmatches\", \"description\": \"Fetch the last matches for a specified cricket team using the RapidAPI cricket API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"page\": {\"description\": \"The page number (starting from 0) of the last matches you want to retrieve.\", \"type\": \"in... | [
"Apologies, but I'm unable to provide football match predictions. My capabilities are limited to providing information and assistance based on available data, not predicting future events or performing tasks beyond my programming.",
"To provide the most accurate predictions, could you please specify the league id... | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Get the football match predictions for the market type 'bts', on page 4, without filtering by league or date. | en | Get the football match predictions for the market type 'bts', on page 4, without filtering by league or date. | plain | When2Call | When2Call:train:661 | null | null | null | train | When2Call:train:661 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | 22494c57853223dc516642b45e14330c4edb08ac344005c06d7685f07d44f75b | clone | train | {} | 793,059,934 | 5.68 | 1 | 1 | Get the football match predictions for the market type BTS on page 4 without filtering by league or date. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-01877 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"fight\", \"description\": \"Fetches details of a single fight by its ID using the Spectation Sports Events API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"is_id\": {\"description\": \"The ID of the fight to fetch details for.\", \"type\": \"int\", \"default\": \"728\"}}}, \"required\":... | [
"<TOOLCALL>[{\"name\": \"getmatchesbyteam\", \"arguments\": {\"team\": \"Royal Challengers Bangalore\"}}]</TOOLCALL>",
"To provide you with the correct information, could you please specify the name of the team you're interested in?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Could you list all matches played by the team in the IPL? | en | Could you list all matches played by the team in the IPL? | plain | When2Call | When2Call:train:1877 | null | null | null | train | When2Call:train:1877 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | F5 | preset | train | {"speed": 1.05, "total_step": 5} | 2,362,525,863 | 3.584 | 1 | 1 | Could you list all matches played by the team in the IPL? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07982 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"kompas_search\", \"description\": \"Performs a search on the Kompas news API using the provided parameters.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"limit\": {\"description\": \"The number of search results to retrieve.\", \"type\": \"int\", \"default\": \"10\"}, \"page\": {\"descrip... | [
"To provide you with the most accurate results, could you please specify the page number for the Kompas search?",
"<TOOLCALL>[{\"name\": \"kompas_search\", \"arguments\": {\"limit\": 5, \"page\": 1, \"command\": \"Indonesian government\"}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Display the top 5 articles about the Indonesian government from the Kompas search results. | en | Display the top 5 articles about the Indonesian government from the Kompas search results. | plain | When2Call | When2Call:train:7982 | null | null | null | train | When2Call:train:7982 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | 419a9ce7406075fbf0f74c41cdaf7369cc4f34da8dde4f4813d4c213003b5ff2 | clone | train | {} | 1,605,186,803 | 5.6 | 1 | 1 | Display the top five articles about the Indonesian government from the Compost search results. | 0.0714 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07671 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"products_get_reviews\", \"description\": \"Fetches brief reviews of a product from the Shein API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"goods_spu\": {\"description\": \"The value of 'productRelationID' returned in the /products/list or /products/search endpoints. Defaults to 'm220... | [
"<TOOLCALL>[{\"name\": \"auto_complete\", \"arguments\": {\"query\": \"unique coffee table\"}}]</TOOLCALL>",
"I'm sorry for the inconvenience, but I'm unable to provide auto-complete suggestions from Wayfair. I recommend visiting their website for personalized results."
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | I'm looking for a new coffee table, but I want something unique. Can you give me some auto-complete suggestions for 'unique coffee table' from Wayfair? | en | I'm looking for a new coffee table, but I want something unique. Can you give me some auto-complete suggestions for 'unique coffee table' from Wayfair? | plain | When2Call | When2Call:train:7671 | null | null | null | train | When2Call:train:7671 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | ce00270c445f1a0789e9542db88ee0fd81ba98273436a956cdcc331d5a1a198c | clone | train | {} | 2,962,775,961 | 9.04 | 1 | 1 | I'm looking for a new coffee table, but I want something unique. Can you give me some auto-complete suggestions for a unique coffee table from Wayfair? | 0.0385 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-04468 | Which response should the assistant give to the user's question, given the available tools? | [] | [
"I'm sorry, I'm unable to provide that information as I don't have the capability to access or interpret data from a travel hacking tool.",
"Japan (JP) is a popular travel destination with Tokyo, Osaka, and Kyoto as major cities. The local currency is Japanese Yen (JPY). The international dialing code is +81. Tim... | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | What are the details of Japan (JP) according to the travel hacking tool? | en | What are the details of Japan (JP) according to the travel hacking tool? | plain | When2Call | When2Call:train:4468 | null | null | null | train | When2Call:train:4468 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | 3f2c311dd7d24b6f0ee4a0fd5fa9a629dca514e1e3fd998d0be8c67b5f72ff3e | clone | train | {} | 2,130,042,530 | 4.981 | 1 | 1 | What are the details of Japan, JP, according to the travel hacking tool? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-04201 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"uniquestageimage\", \"description\": \"Retrieves the image for a specific Motorsport unique stage using the provided stage ID.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"is_id\": {\"description\": \"The unique stage ID for which to retrieve the image.\", \"type\": \"int\", \"default\":... | [
"Sure, I can help with that. To provide the most accurate information, could you please specify the date of the game?",
"<TOOLCALL>[{\"name\": \"nhl_summary\", \"arguments\": {\"is_id\": \"20230101\"}}]</TOOLCALL>"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | I need the game summary for the NHL game with ID '20230101'. | en | I need the game summary for the NHL game with ID '20230101'. | long_number | When2Call | When2Call:train:4201 | null | null | null | train | When2Call:train:4201 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | 0d370e11aea4d9fe335096f4119750fc390d75c36c4cb3777360b147c5ba8c69 | clone | train | {} | 3,581,032,335 | 5.36 | 1 | 1 | I need the game summary for the NHL game with ID 230101. | 0.1053 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-04226 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"simulate_query_database\", \"description\": \"Simulates querying a database based on certain conditions.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"table\": {\"description\": \"Name of the table to query.\", \"type\": \"str\"}, \"conditions\": {\"description\": \"Conditions for the que... | [
"<TOOLCALL>[{\"name\": \"structural_analysis\", \"arguments\": {\"building_id\": \"B123\", \"floor_numbers\": [1, 3, 5]}}]</TOOLCALL>",
"Which specific aspect of the structural analysis would you like me to focus on? For example, stress, strain, or deflection."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Perform a static structural analysis on the building with ID 'B123' for floors 1, 3, and 5. | en | Perform a static structural analysis on the building with ID 'B123' for floors 1, 3, and 5. | long_number | When2Call | When2Call:train:4226 | null | null | null | train | When2Call:train:4226 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | 9478b606993d9d537eb6a099494d093be5db4b48f9b603e6485198f516639f81 | clone | train | {} | 1,464,107,371 | 6.6 | 1 | 1 | Perform a static structural analysis on the building with IDB123 for floors 1, 3 and 5. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03326 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_4_date_endpoint\", \"description\": \"Fetches the exchange rates for specified currencies on a given date.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"base\": {\"description\": \"The base currency for which to get the exchange rates.\", \"type\": \"str\", \"default\": \"eur\"}, \"sy... | [
"Amazon's ESG score as of now is 22 out of 100, according to Refinitiv. This score reflects its environmental, social, and governance performance.",
"I'm sorry, I'm unable to provide real-time data or perform live calculations. Please check a reliable financial source for the most current ESG score of Amazon (AMZ... | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | What is the ESG score for Amazon (AMZN) on its current date? | en | What is the ESG score for Amazon (AMZN) on its current date? | plain | When2Call | When2Call:train:3326 | null | null | null | train | When2Call:train:3326 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | aa81eb18be6e8e9482a87789245790debfdcae87345a59f4ffb140a8cb2c36ba | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 1,795,725,424 | 4.66 | 1 | 1 | What is the ESG score for Amazon AMZEN on its current date? | 0.0833 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-06394 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"teamlastmatches\", \"description\": \"Fetch the last matches for a specified cricket team using the RapidAPI cricket API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"page\": {\"description\": \"The page number (starting from 0) of the last matches you want to retrieve.\", \"type\": \"in... | [
"To provide the most accurate predictions, could you please specify the league identifier?",
"Apologies, but I'm unable to provide football match predictions. I can't perform tasks that involve speculation or predicting outcomes."
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Get the football match predictions for the market type 'OU25', on page 3, for the date '2022-08-14'. | en | Get the football match predictions for the market type 'OU25', on page 3, for the date '2022-08-14'. | long_number | When2Call | When2Call:train:6394 | null | null | null | train | When2Call:train:6394 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | 90b9b2bfc3bafda7e46715ea35c8b692feff2ff8e672d87e8460eda330b22cb3 | clone | train | {} | 399,213,414 | 8.72 | 1 | 1 | Get the football match predictions for the market type 025 on page 3 for the date 2022 0814. | 0.0385 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05901 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"count\", \"description\": \"Fetches the total number of expired followers from an OnlyFans API endpoint.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"signstart\": {\"description\": \"The start of the signature used for authorization.\", \"type\": \"str\", \"default\": \"\"}, \"apptoken\"... | [
"I'm sorry for the inconvenience, but I'm unable to fetch that information. I can't perform tasks that require specific account access or data retrieval.",
"To proceed, I'll need the application token (apptoken) which is required to fetch the total number of expired followers. Could you please provide it?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Fetch the total number of expired followers for an OnlyFans account with the following parameters: signstart='abc123', signend='def456', xbc='ghi456', timezone='UTC', useragent='Mozilla/5.0', auth_id='123456', sess='789456'. | en | Fetch the total number of expired followers for an OnlyFans account with the following parameters: signstart=abc123, signend=def456, xbc=ghi456, timezone=UTC, useragent=Mozilla/5.0, auth_id=123456, sess=789456. | id_token | When2Call | When2Call:train:5901 | null | null | null | train | When2Call:train:5901 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | 2ab011ef4e209ac751790af7e9d78ebfa66824596e183b3175c3da5e61070f6a | clone | train | {} | 822,542,918 | 24.16 | 1 | 1 | Fetch the total number of expired followers for an only fans account with the following parameters. Sign Start equals ABC 123, signant D456, XBC equals GI456, time zone equals UTC user agent, Mozilla 5.0, auth ID equals 123456, CESS equals 789456. | 0.3333 | wer | Systran/faster-whisper-base.en | 8 | true | |
when2call-train-04894 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"getmoviestatus\", \"description\": \"Fetches the status of a movie using the HelioViewer API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"is_id\": {\"description\": \"Unique movie identifier (provided by the response to a `queueMovie` request).\", \"type\": \"str\", \"default\": \"VXvX5... | [
"To provide you with the fact, could you please specify the month for the 1st day?",
"{\"fact\": \"On June 1st, 1980, CNN, the first 24-hour cable news channel, was launched.\"}"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Get a JSON-formatted fact about the 1st day as a complete sentence. | en | Get a JSON-formatted fact about the 1st day as a complete sentence. | plain | When2Call | When2Call:train:4894 | null | null | null | train | When2Call:train:4894 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | F5 | preset | train | {"speed": 1.05, "total_step": 5} | 3,139,341,919 | 5.037 | 1 | 1 | Get a JSON formatted fact about the first day as a complete sentence. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05834 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"autocomplete_zipcodes_zip4\", \"description\": \"Fetches autocomplete suggestions for USA Zip+4 zip codes using the provided zipcode and API key.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"zipcode\": {\"description\": \"The base ZIP code for which to fetch autocomplete suggestions.\", ... | [
"<TOOLCALL>[{\"name\": \"get_mempool_v2\", \"arguments\": {\"blockchain\": \"bitcoin\"}}]</TOOLCALL>",
"To provide the most accurate information, could you please specify if you'd like the results limited to a certain number of transactions per call?"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | List all unconfirmed transactions in the Bitcoin blockchain. | en | List all unconfirmed transactions in the Bitcoin blockchain. | plain | When2Call | When2Call:train:5834 | null | null | null | train | When2Call:train:5834 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 13653ba8d92cd9f77f185aa94d5d0fd1707873cfa771a0c3f5b9866b607cf14b | clone | train | {} | 3,929,258,536 | 3.28 | 1 | 1 | List all unconfirmed transactions in the Bitcoin blockchain. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00994 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_company_data_by_linkedin_url\", \"description\": \"Fetch company data using a given LinkedIn URL via the ToolBench RapidAPI.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"linkedin_url\": {\"description\": \"The LinkedIn URL of the company to retrieve data for.\", \"type\": \"str\", \"... | [
"<TOOLCALL>[{\"name\": \"get_company_data_by_linkedin_url\", \"arguments\": {\"linkedin_url\": \"https://www.linkedin.com/company/tesla\"}}]</TOOLCALL>",
"Apologies, I'm unable to perform that task as I can't access or analyze external websites."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Can you get the company details from Tesla's LinkedIn URL? | en | Can you get the company details from Tesla's LinkedIn URL? | plain | When2Call | When2Call:train:994 | null | null | null | train | When2Call:train:994 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | 07b95955eec0a4d8b1567845bae9e9694456bcdc9cde4518de67015678aa25ae | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 3,979,341,770 | 2.96 | 1 | 1 | Can you get the company duties from Tesla's LinkedIn URL? | 0.1 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03507 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"shorts_sound_attribution\", \"description\": \"Fetches the Shorts original sound attribution listing from the YouTube API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"x_cachebypass\": {\"description\": \"Cache bypass value, must be set to '1' if used.\", \"type\": \"str, optional\", \"d... | [
"To provide you with the detailed consumer information, I'll need the unique ID for 'user456'. Could you please provide that?",
"To access detailed consumer information for 'user456' on TrustPilot, you'll need to use their API. Here's a simplified step-by-step guide:\n\n1. Register for a TrustPilot Business accou... | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | I need the detailed consumer information for 'user456' from TrustPilot. | en | I need the detailed consumer information for user456 from TrustPilot. | id_token | When2Call | When2Call:train:3507 | null | null | null | train | When2Call:train:3507 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | 564a6dc60d17106f8ce2c6299e8d8dabf0807e6bc16e4beaccabd7f0181273b5 | clone | train | {} | 617,594,246 | 4.88 | 1 | 1 | I need the detailed consumer information for User 456 from Trust Pilot. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05265 | Which response should the assistant give to the user's question, given the available tools? | [] | [
"Apologies, but I'm unable to perform statistical calculations. You may use a chi-square test to determine the association between the variables.",
"<TOOLCALL>[{\"name\": \"chi_square_independence_test\", \"arguments\": {\"contingency_table\": [[75, 50], [50, 75]], \"significance_level\": 0.05}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Is there a significant association between two variables, given the contingency table [[75, 50], [50, 75]] and a significance level of 0.05? | en | Is there a significant association between two variables, given the contingency table 75, 50; 50, 75 and a significance level of 0.05? | list | When2Call | When2Call:train:5265 | null | null | null | train | When2Call:train:5265 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | f819cd5af79550405ed05a94532f4104b500c6d75cb9b8e05bcffaa01c2b82f0 | clone | train | {} | 3,183,933,696 | 11.68 | 1 | 0.9991 | Is there a significant association between two variables given the contingency table 75, 50, 50, 75 and the significance level of 0.05? | 0.0357 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05429 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"v1_hobbies\", \"description\": \"Fetches a random hobby and a Wikipedia link detailing the hobby from the Hobbies by API Ninjas API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"category\": {\"description\": \"The category of hobbies to fetch. Possible values are 'general', 'sports_and_o... | [
"<TOOLCALL>[{\"name\": \"v1_hobbies\", \"arguments\": {\"category\": \"education\"}}]</TOOLCALL>",
"Hobby: Origami, the traditional Japanese art of paper folding. It's educational as it involves geometry and thinking in three dimensions. Wikipedia link: https://en.wikipedia.org/wiki/Origami"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Get details about a random hobby from the 'education' category and its Wikipedia link. | en | Get details about a random hobby from the 'education' category and its Wikipedia link. | plain | When2Call | When2Call:train:5429 | null | null | null | train | When2Call:train:5429 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | e58580c5cd910c6600e252d3e156b905b339d51722c747ec677445e9e39484e9 | clone | train | {} | 1,319,694,942 | 5.195 | 1 | 1 | Get details about a random hobby from the education category and its Wikipedia link. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-01424 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"v1_dogs\", \"description\": \"Fetch a list of dog breeds matching specified parameters from the API. Returns at most 20 results by default.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"offset\": {\"description\": \"Number of results to offset for pagination. Defaults to None.\", \"type\"... | [
"To proceed, could you please specify the keyword for the channel search?",
"Here are some German technology YouTube channels you might find interesting: Technikfaultier, Inside Digital, and Technik News."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Search for YouTube channels focused on topics, in German, without specifying a sorting method. | en | Search for YouTube channels focused on topics, in German, without specifying a sorting method. | plain | When2Call | When2Call:train:1424 | null | null | null | train | When2Call:train:1424 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | 0d370e11aea4d9fe335096f4119750fc390d75c36c4cb3777360b147c5ba8c69 | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 3,043,872,456 | 6.1 | 1 | 1 | Search for YouTube channels focused on topics in German without specifying a sorting method. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-01796 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"provider2_live_match_id\", \"description\": \"Fetches live match details from the provider2 API for a given match ID.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"is_id\": {\"description\": \"The match ID for which to fetch the live details.\", \"type\": \"str\", \"default\": \"\"}}}, \"... | [
"I'm sorry, I can't provide live betting odds. I'm designed to assist with information and tasks, but I don't have real-time access to betting platforms.",
"To provide you with the live betting odds, I'll need the match ID. Could you please provide me with that?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | What are the live betting odds for the match? | en | What are the live betting odds for the match? | plain | When2Call | When2Call:train:1796 | null | null | null | train | When2Call:train:1796 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | 6c08f4a61aa9cc86d2fb7871fa7f7c42753fd30de609f0bf95f0a97a2c51a179 | clone | train | {} | 2,722,473,292 | 2.64 | 1 | 1 | What are the live betting odds for the match? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03846 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_random_question\", \"description\": \"Fetches a random question from the API, with an optional parameter to specify the type of the challenge.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"type\": {\"description\": \"The type of question to fetch. Options include 'funny' and 'dirty'. ... | [
"I'm afraid I can't provide the exact compatibility score without the API key. However, based on general compatibility factors, Lily and Jack have a high potential for a harmonious relationship due to their shared interests and complementary personality traits.",
"I'm sorry for the inconvenience, but I'm unable t... | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | I want to know the relationship compatibility score for Lily and Jack, but I don't have the API key, can you still help? | en | I want to know the relationship compatibility score for Lily and Jack, but I don't have the API key, can you still help? | plain | When2Call | When2Call:train:3846 | null | null | null | train | When2Call:train:3846 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | M4 | preset | train | {"speed": 1.05, "total_step": 5} | 3,710,927,642 | 7.015 | 1 | 1 | I want to know the relationship compatibility score for Lillian Jack, but I don't have the API key, can you still help? | 0.087 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07252 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"extract_entities\", \"description\": \"Extracts named entities from the provided text using the specified model and RapidAPI key.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"text\": {\"description\": \"The text from which to extract named entities.\", \"type\": \"str\", \"default\": \"U... | [
"Apologies, I'm unable to provide the observation as I can't perform real-time analysis or access external APIs.",
"<TOOLCALL>[{\"name\": \"main\", \"arguments\": {\"text\": \"Microsoft announced its new product launch\"}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | What is the observation from the NeuroNet NLP API for the text 'Microsoft announced its new product launch'? | en | What is the observation from the NeuroNet NLP API for the text 'Microsoft announced its new product launch'? | plain | When2Call | When2Call:train:7252 | null | null | null | train | When2Call:train:7252 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 76fec9cb426cb09752beca2b172d20c5e7276c9315e8b7d236d6c5a816fe2875 | clone | train | {} | 940,168,127 | 6.56 | 1 | 1 | What is the observation from the NeuroNet NLP API for the text Microsoft announced its new product launch? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03723 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"independent_samples_t_test\", \"description\": \"Conducts a two-sample independent t-test and returns the t-statistic, p-value, and conclusion.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"sample1\": {\"description\": \"The first sample of observations.\", \"type\": \"List[float]\", \"de... | [
"<TOOLCALL>[{\"name\": \"independent_samples_t_test\", \"arguments\": {\"sample1\": [50000, 55000, 60000, 52000, 58000], \"sample2\": [45000, 48000, 50000, 47000, 52000], \"alpha\": 0.05}}]</TOOLCALL>",
"Apologies, but I'm unable to perform calculations or statistical analyses. I recommend using a statistical sof... | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Perform a t-test to compare the average salaries of two departments: Department A [50000, 55000, 60000, 52000, 58000] dollars and Department B [45000, 48000, 50000, 47000, 52000] dollars with a significance level of 0.05. | en | Perform a t-test to compare the average salaries of two departments: Department A 50000, 55000, 60000, 52000, 58000 dollars and Department B 45000, 48000, 50000, 47000, 52000 dollars with a significance level of 0.05. | list | When2Call | When2Call:train:3723 | null | null | null | train | When2Call:train:3723 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | f4b8216176b0ece39ffd4d64cf5078cf7a3e7c07808a8b2ec4eacb9b745ea1c8 | clone | train | {} | 3,892,722,393 | 23.115 | 1 | 1 | Perform a T-test to compare the average salaries of two departments. Department A 50,000, 55,000, 60,000, 52,000, 58,000 dollars and Department B 45,000, 48,000, 50,000, 47,000, 52,000 dollars with a significance level of 0.05. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-02273 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"bus_detours\", \"description\": \"Fetches the bus detours from the SEPTA API based on the specified route number.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"req1\": {\"description\": \"The route number for which to retrieve detour information. Defaults to 2.\", \"type\": \"int, optiona... | [
"<TOOLCALL>[{\"name\": \"get_stations\", \"arguments\": {\"distance\": \"200\", \"longitude\": \"-118.2436849\", \"latitude\": \"34.0522342\"}}]</TOOLCALL>",
"I found 250 charging stations within 200 miles of Los Angeles. 100 are Tesla Superchargers, 50 are ChargePoint stations, and the rest are a mix of other pr... | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | List all vehicle charging stations within 200 miles of Los Angeles. | en | List all vehicle charging stations within 200 miles of Los Angeles. | long_number | When2Call | When2Call:train:2273 | null | null | null | train | When2Call:train:2273 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | aa145973ea0371e9f5e54ab1086610c84e617fed354b112dc5d6b1e9242d488f | clone | train | {} | 3,344,713,619 | 5.75 | 1 | 1 | list all vehicle charging stations within 200 miles of Los Angeles. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-02992 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"discover\", \"description\": \"Discover files using the FilePursuit API with optional filters for pagination, result type, and specific links.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"start\": {\"description\": \"The starting point for result pagination.\", \"type\": \"int, optional\... | [
"I apologize, but I'm unable to perform tasks such as taking screenshots. I'm here to provide information and answer questions to the best of my ability.",
"<TOOLCALL>[{\"name\": \"screenshot_endpoint\", \"arguments\": {\"height\": \"768\", \"url\": \"https://example.com\", \"width\": \"1024\"}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Hey, can you grab a screenshot of this page: https://example.com, with a width of 1024px and height of 768px? | en | Hey, can you grab a screenshot of this page: https://example.com, with a width of 1024px and height of 768px? | url | When2Call | When2Call:train:2992 | null | null | null | train | When2Call:train:2992 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | c3c0ba5a31d1e6a1386951f2a6db0bb39c5e956a75698f33a063df56b6c0133f | clone | train | {} | 2,765,815,899 | 14.48 | 1 | 1 | Hey, can you grab a screenshot of this page, httpps.com with a width of 124px and height of 76.8px. | 0.1034 | wer | Systran/faster-whisper-base.en | 11 | true | |
when2call-train-06172 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_cigar_average_length\", \"description\": \"Fetches the average length of cigars based on specified filters.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"wrapper\": {\"description\": \"The wrapper type of the cigar. Default is 'Connecticut Shade, Ecuador'.\", \"type\": \"str\", \"defa... | [
"<TOOLCALL>[{\"name\": \"get_cigar_average_length\", \"arguments\": {\"strength\": \"Medium\", \"wrapper\": \"Ecuador\"}}]</TOOLCALL>",
"To provide you with the most accurate information, could you please specify the color of the cigar?"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Can you determine the average length of cigars with a Medium strength and an Ecuador wrapper? | en | Can you determine the average length of cigars with a Medium strength and an Ecuador wrapper? | plain | When2Call | When2Call:train:6172 | null | null | null | train | When2Call:train:6172 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | F2 | preset | train | {"speed": 1.05, "total_step": 5} | 583,048,108 | 5.671 | 1 | 1 | Can you determine the average length of cigars with a medium strength and an Ecuador wrapper? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-01492 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"air_quality_history\", \"description\": \"Fetches the past 24 hours of air quality observations for a specified geographic location.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"lat\": {\"description\": \"Latitude of the location for which to retrieve air quality data.\", \"type\": \"int... | [
"I'm sorry, I'm unable to provide the ski weather forecast. I can only assist with general information and tasks. Please check a reliable weather service for the most accurate forecast.",
"To provide the forecast, could you please specify the location or the coordinates for the ski area?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Fetch the ski weather forecast for the next 4 days in English. | en | Fetch the ski weather forecast for the next 4 days in English. | plain | When2Call | When2Call:train:1492 | null | null | null | train | When2Call:train:1492 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | f72bc96782350def2ebba1a905c6b69e7aae00e7c0d5e45b83b3ddc013d0547c | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 1,299,885,925 | 4.12 | 1 | 1 | Fetch the ski weather forecast for the next four days in English. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-08174 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"ultimateoscillator\", \"description\": \"Fetches the Ultimate Oscillator (UO) indicator data for a given cryptocurrency symbol from the Qvantana API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"exchange\": {\"description\": \"The exchange to retrieve data from.\", \"type\": \"str\", \"d... | [
"<TOOLCALL>[{\"name\": \"supply_chain\", \"arguments\": {\"ticker\": \"MSFT\"}}]</TOOLCALL>",
"To provide the most accurate information, could you please confirm the stock ticker for Microsoft Corporation?"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Fetch supply chain details for Microsoft Corporation using its stock ticker. | en | Fetch supply chain details for Microsoft Corporation using its stock ticker. | plain | When2Call | When2Call:train:8174 | null | null | null | train | When2Call:train:8174 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 650bcc4d5689f084f62e93b394a67a221a4446f0a2fca14848a713bc35143d78 | clone | train | {} | 1,196,383,111 | 4.64 | 1 | 1 | fetch supply chain details for Microsoft Corporation using its stock ticker. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-06390 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"qr_code\", \"description\": \"Creates a QR code by making a GET request to the QR code API with the given URL.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"url\": {\"description\": \"The URL to be included in the QR code. Defaults to 'https://rapidapi.com/'.\", \"type\": \"str\", \"defau... | [
"I have generated the QR code for the URL 'https://www.example.com/product123'. You can scan it to access the webpage directly.",
"To generate the QR code, I need the specific URL for the product. Could you please provide it?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Generate a QR code for the product. | en | Generate a QR code for the product. | plain | When2Call | When2Call:train:6390 | null | null | null | train | When2Call:train:6390 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | c41a6582267dd93df8cac594ce933e878fdd4a56336c7e9c3c603956ea5101a3 | clone | train | {} | 1,451,355,056 | 2.477 | 1 | 1 | Generate a QR code for the product. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-04996 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"reserve_hotel_room\", \"description\": \"Reserves a hotel room with the specified details.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"room_type\": {\"description\": \"The type of room to reserve.\", \"type\": \"str\"}, \"nightly_rate\": {\"description\": \"The nightly rate for the room... | [
"<TOOLCALL>[{\"name\": \"place_safeway_order\", \"arguments\": {\"location\": \"Palo Alto, CA\", \"items\": [\"milk\", \"bread\", \"eggs\"], \"quantity\": [2, 1, 3]}}]</TOOLCALL>",
"To complete your order, I'll need to know the location of the Safeway store. Could you please provide that?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Order 'milk', 'bread', and 'eggs' from Safeway with quantities '2', '1', and '3' respectively. | en | Order 'milk', 'bread', and 'eggs' from Safeway with quantities '2', '1', and '3' respectively. | plain | When2Call | When2Call:train:4996 | null | null | null | train | When2Call:train:4996 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | F5 | preset | train | {"speed": 1.05, "total_step": 5} | 3,992,111,845 | 4.716 | 1 | 1 | order milk, bread and eggs from Safeway with quantities 2, 1 and 3 respectively. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-02008 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"getplatformlist\", \"description\": \"Fetches the list of platforms with their IDs and names from the API endpoint as JSON.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"page\": {\"description\": \"Page number to retrieve the platform list from the API.\", \"type\": \"int\", \"default\": ... | [
"Apologies, but I'm unable to fetch trivia questions as I don't have the capability to perform such tasks.",
"<TOOLCALL>[{\"name\": \"v1_trivia\", \"arguments\": {\"limit\": 10, \"category\": \"general\"}}]</TOOLCALL>"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Fetch 10 trivia questions from the 'general' category. | en | Fetch 10 trivia questions from the 'general' category. | plain | When2Call | When2Call:train:2008 | null | null | null | train | When2Call:train:2008 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | aa81eb18be6e8e9482a87789245790debfdcae87345a59f4ffb140a8cb2c36ba | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 610,118,385 | 3.14 | 1 | 1 | Fetch 10 trivia questions from the general category. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-02418 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"getteamroster\", \"description\": \"Retrieves the roster of a hockey team for a given season using the RapidAPI NHL Stats and Live Data API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"is_id\": {\"description\": \"The ID of the team for which to retrieve the roster.\", \"type\": \"int\"... | [
"<TOOLCALL>[{\"name\": \"basketball_match_statistics\", \"arguments\": {\"match_id\": \"BASKET_MATCH_ID_123\"}}]</TOOLCALL>",
"To provide you with the statistics, I need to confirm the match ID. Is 'BASKET_MATCH_ID_123' the correct identifier for the basketball match?"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | What are the statistics for the basketball match with ID 'BASKET_MATCH_ID_123'? | en | What are the statistics for the basketball match with ID BASKET_MATCH_ID_123? | id_token | When2Call | When2Call:train:2418 | null | null | null | train | When2Call:train:2418 | 1 | qwen | Qwen/Qwen3-TTS-12Hz-0.6B-Base | 5d83992436eae1d760afd27aff78a71d676296fc | 022e286b98fbec7e1e916cb940cdf532cd9f488e | 63916561e73648a3806eecf49ecbf1889fbb34773ee8dbf99de19f096c19c782 | clone | train | {} | 2,391,769,851 | 4.4 | 1 | 1 | What are the statistics for the basketball match with ID basket match ID 12-3? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-06960 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"matchschedules\", \"description\": \"Fetches the schedules of ice hockey matches for a specified date, including match timings, teams, and other relevant information.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"day\": {\"description\": \"The day of the month for which you want to retrie... | [
"<TOOLCALL>[{\"name\": \"matchschedules\", \"arguments\": {\"day\": 25, \"month\": 12, \"year\": 2022}}]</TOOLCALL>",
"I'm sorry, I can't provide that information as I don't have real-time data or the ability to perform web searches."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Can you provide the match schedules for December 25, 2022? | en | Can you provide the match schedules for December 25, 2022? | long_number | When2Call | When2Call:train:6960 | null | null | null | train | When2Call:train:6960 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | aa81eb18be6e8e9482a87789245790debfdcae87345a59f4ffb140a8cb2c36ba | clone | train | {} | 2,270,919,506 | 4.39 | 1 | 1 | Can you provide the match schedules for December 25, 2022? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07250 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"investment_profit\", \"description\": \"Calculates the profit from an investment based on the initial amount, annual return rate, and time.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"amount\": {\"description\": \"The initial investment amount.\", \"type\": \"float\"}, \"rate\": {\"desc... | [
"To calculate the area, I need the ending x-value for the integration. Could you please provide the end_x value?",
"Apologies, but I'm unable to calculate that for you. I can't perform mathematical computations, including applying the Simpson's rule to find the area under a curve."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | What is the area under the curve of the function 'x^2' starting from x=1 using the Simpson's rule? | en | What is the area under the curve of the function 'x^2' starting from x=1 using the Simpson's rule? | plain | When2Call | When2Call:train:7250 | null | null | null | train | When2Call:train:7250 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | 90b9b2bfc3bafda7e46715ea35c8b692feff2ff8e672d87e8460eda330b22cb3 | clone | train | {} | 400,316,000 | 7.456 | 1 | 1 | What is the area under the curve of the function XCA2 starting from X1 using the Simpsons rule? | 0.1 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03443 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"points\", \"description\": \"Fetches official FedExCup points earned per player for a given tournament ID and year.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"tournid\": {\"description\": \"The tournament ID for which to fetch the points.\", \"type\": \"str\", \"default\": \"475\"}, \"... | [
"Here are the scheduled daily matches for December 12, 2022:\n\n1. Team A vs Team B at 10:00 AM\n2. Team C vs Team D at 1:00 PM\n3. Team E vs Team F at 4:00 PM\n4. Team G vs Team H at 7:00 PM",
"To provide you with the list of scheduled daily matches, I'll need the specific date. Could you please provide the date... | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Could you fetch the list of scheduled daily matches? | en | Could you fetch the list of scheduled daily matches? | plain | When2Call | When2Call:train:3443 | null | null | null | train | When2Call:train:3443 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | ea5942f5833c18b6de5a5c94e4127bf7e145d2c101cb001565937efa41fe6fa3 | clone | train | {} | 199,669,895 | 4.043 | 1 | 1 | Could you fetch the list of scheduled daily matches? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05637 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"symbols_get_profile\", \"description\": \"Fetch summary information for a stock quote, index, exchange, etc., given an issue ID.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"issueid\": {\"description\": \"The value of the issueId field obtained from previous endpoints such as .../auto-co... | [
"<TOOLCALL>[{\"name\": \"screeners_details\", \"arguments\": {\"screener_id\": \"High Yield ETFs\"}}]</TOOLCALL>",
"I'm sorry, I'm unable to provide detailed information about the 'High Yield ETFs' screener as I don't have the capability to perform such tasks."
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Retrieve detailed information about the 'High Yield ETFs' screener. | en | Retrieve detailed information about the 'High Yield ETFs' screener. | plain | When2Call | When2Call:train:5637 | null | null | null | train | When2Call:train:5637 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | cd7eff8d04a9f1dad071c265715a45169e0553b15b2afd92d21f670b9ec07c8c | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 3,710,388,250 | 4.48 | 1 | 1 | Retrieve detailed information about the high-yield ETF's screener. | 0.1111 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07965 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_articles\", \"description\": \"Fetches articles from the climate news feed with given parameters for limit, excluded source, and specific source.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"limit\": {\"description\": \"Number of articles to fetch. Default is 50.\", \"type\": \"int, ... | [
"To provide you with the most relevant science news in English, I need to know the specific language region. Could you please specify the language region code, such as 'en-US'?",
"<TOOLCALL>[{\"name\": \"science\", \"arguments\": {\"lr\": \"en-US\"}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Fetch science news in English. | en | Fetch science news in English. | plain | When2Call | When2Call:train:7965 | null | null | null | train | When2Call:train:7965 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 54d1aa4453f20afe0d3761c0700ed6f94df86932d633348f83fe09cb1f11673b | clone | train | {} | 3,561,061,430 | 2.08 | 1 | 1 | fetch science news in English. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-08491 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"is_valid_sudoku\", \"description\": \"Checks if a 9x9 Sudoku board is valid.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"board\": {\"description\": \"The Sudoku board represented as a 2D list of strings.\", \"type\": \"List[List[str]]\"}}}, \"required\": [\"board\"]}",
"{\"name\": \"e... | [
"To assist you better, could you please provide me with the list of numbers?",
"<TOOLCALL>[{\"name\": \"find_next_greater_element\", \"arguments\": {\"nums\": [1, 3, 2, 4]}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Find the next greater element for each number in the given list. | en | Find the next greater element for each number in the given list. | plain | When2Call | When2Call:train:8491 | null | null | null | train | When2Call:train:8491 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | fd65afa816d1dbe8d6fb934b24e0b6c31aecd859a62be0b665f229fc1b2d269c | clone | train | {} | 109,345,632 | 3.44 | 1 | 1 | Find the next greater element for each number in the given list. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-02510 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_coin_price_difference\", \"description\": \"Fetches the price difference of a specified cryptocurrency from the Crypto Arbitrage Trading API using the provided RapidAPI key.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"name\": {\"description\": \"The name of the cryptocurrency for wh... | [
"I'm sorry for the inconvenience, but I'm unable to generate a Litecoin wallet. It's important to use trusted wallet software to securely generate and manage your cryptocurrency wallets.",
"To proceed, I'll need your mnemonic phrase. This is a 24-word phrase used to generate the wallet. Please provide it."
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Generate a Litecoin wallet. | en | Generate a Litecoin wallet. | plain | When2Call | When2Call:train:2510 | null | null | null | train | When2Call:train:2510 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 8cd79b938c5910ba4f1afa307ce18d92aa37b2e537437a8e86ddf785b1b0bc15 | clone | train | {} | 2,528,847,199 | 1.92 | 1 | 1 | Generate a Litecoin wallet. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-02849 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"is_subset\", \"description\": \"Checks if one set is a subset of another set.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"set1\": {\"description\": \"The first set.\", \"type\": \"set\"}, \"set2\": {\"description\": \"The second set.\", \"type\": \"set\"}}}, \"required\": [\"set1\", \"s... | [
"<TOOLCALL>[{\"name\": \"find_max_subarray_sum\", \"arguments\": {\"nums\": [1, -3, 2, 1, -1]}}]</TOOLCALL>",
"To proceed, I need to know if the list you provided is the complete input or if there are more integers to consider."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Find the maximum sum of a contiguous subarray within the list [1, -3, 2, 1, -1]. | en | Find the maximum sum of a contiguous subarray within the list 1, -3, 2, 1, -1. | list | When2Call | When2Call:train:2849 | null | null | null | train | When2Call:train:2849 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | f59cd6603efeec1b0239a90a2794d52761e0b3e1ab975cdb4bd1806dfda3731d | clone | train | {} | 925,884,026 | 5.707 | 1 | 1 | Find the maximum sum of a contiguous subarray within the list 13-215-1. | 0.0625 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-07005 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"search_php\", \"description\": \"Search for geocoding information using the specified Geokeo Forward Geocoding API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"api\": {\"description\": \"The API key for accessing the Geokeo Forward Geocoding service.\", \"type\": \"str\", \"default\": \... | [
"To provide you with the address, could you please specify the format in which you'd like the address to be presented?",
"Apologies, I'm unable to provide the address for those coordinates as I don't have real-time geolocation capabilities."
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | What is the address corresponding to the coordinates 34.0522° N, 118.2437° W? | en | What is the address corresponding to the coordinates 34.0522° N, 118.2437° W? | long_number | When2Call | When2Call:train:7005 | null | null | null | train | When2Call:train:7005 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | aa81eb18be6e8e9482a87789245790debfdcae87345a59f4ffb140a8cb2c36ba | clone | train | {} | 3,973,207,019 | 7.15 | 1 | 1 | What is the address? Corresponding to the coordinates 34 poo. Zoo on 18 2437 degrees woo. | 0.3478 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00146 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"latest_rates\", \"description\": \"Fetches the latest real-time exchange rates for given symbols based on a specified base currency.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"symbols\": {\"description\": \"Comma-separated list of currency or commodity codes to retrieve rates for.\", \... | [
"Apologies, I'm unable to provide real-time currency exchange rates as I don't have the capability to perform live data retrieval.",
"To provide you with the latest rates, could you please specify the base currency you'd like to use?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Can you tell me the latest rates for USD, EUR, and JPY against GBP? | en | Can you tell me the latest rates for USD, EUR, and JPY against GBP? | plain | When2Call | When2Call:train:146 | null | null | null | train | When2Call:train:146 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | M4 | preset | train | {"speed": 1.05, "total_step": 5} | 2,326,065,121 | 4.712 | 1 | 1 | Can you tell me the latest rates for USD, EUR, and JPY against GBP? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00041 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"by_zip_code\", \"description\": \"Fetches weather data based on a given zip code using the Toolbench RapidAPI service.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"zip\": {\"description\": \"The zip code for which to fetch the weather data. If the country is not specified, the search def... | [
"To assist you better, could you please provide me with the zip code for Paris, France?",
"<TOOLCALL>[{\"name\": \"by_zip_code\", \"arguments\": {\"zip\": \"75000,FR\", \"lang\": \"fr\", \"units\": \"metric\"}}]</TOOLCALL>"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Fetch the current weather in Paris, France in French and using metric units. | en | Fetch the current weather in Paris, France in French and using metric units. | plain | When2Call | When2Call:train:41 | null | null | null | train | When2Call:train:41 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | f5e00e51a31bdcc7491c52c87127d5e53a990cbe232d7e2ed133d4ffb4a36726 | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 31,052,897 | 4.78 | 1 | 1 | fetch the current weather in Paris, France in French and using metric units. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05982 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"get_ip_zipcode\", \"description\": \"Retrieves the ZIP code of a given IP address using the ip-api.com API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"ip\": {\"description\": \"The IP address to locate.\", \"type\": \"str\"}}}, \"required\": [\"ip\"]}",
"{\"name\": \"get_ip_location\... | [
"<TOOLCALL>[{\"name\": \"get_ip_location\", \"arguments\": {\"ip\": \"55.66.77.88\"}}]</TOOLCALL>",
"The IP address 55.66.77.88 corresponds to a location in Moscow, Russia, with approximate coordinates 55.7558° N, 37.6176° E."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | I need the latitude and longitude coordinates of the IP address 55.66.77.88. Can you help me with that? | en | I need the latitude and longitude coordinates of the IP address 55.66.77.88. Can you help me with that? | long_number | When2Call | When2Call:train:5982 | null | null | null | train | When2Call:train:5982 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | 54e81b1c5452501ae8c335897b6695f8e4ec7e0bbc99427ef66cbdfd35a24573 | clone | train | {} | 973,934,225 | 6.52 | 1 | 1 | I need the latitude and longitude coordinates of the IP address 55667788. Can you help me with that? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-01252 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"prematch_event\", \"description\": \"Fetches pre-match event data from a sports betting API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"lang\": {\"description\": \"The language code for the API response.\", \"type\": \"str\", \"default\": \"en\"}, \"evt\": {\"description\": \"The event... | [
"<TOOLCALL>[{\"name\": \"prematch_event\", \"arguments\": {\"lang\": \"pt\", \"evt\": 24681}}]</TOOLCALL>",
"To provide you with the pre-match information, could you please specify the event ID for the volleyball game?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | What is the pre-match information for a volleyball game in Portuguese? | en | What is the pre-match information for a volleyball game in Portuguese? | plain | When2Call | When2Call:train:1252 | null | null | null | train | When2Call:train:1252 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | F2 | preset | train | {"speed": 1.05, "total_step": 5} | 2,202,615,316 | 4.336 | 1 | 1 | What is the pre-match information for a volleyball game in Portuguese? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03732 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"gethardquestion\", \"description\": \"Fetches the number of hard questions solved by a specified user on LeetCode using the RapidAPI service.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"username\": {\"description\": \"The LeetCode username of the user whose hard question count is to be ... | [
"I'm sorry, I can't provide that information as I don't have the ability to access or analyze data from LeetCode.",
"<TOOLCALL>[{\"name\": \"gethardquestion\", \"arguments\": {\"username\": \"algorithmMaster\"}}]</TOOLCALL>"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | How many hard questions has user 'algorithmMaster' solved on LeetCode? | en | How many hard questions has user 'algorithmMaster' solved on LeetCode? | plain | When2Call | When2Call:train:3732 | null | null | null | train | When2Call:train:3732 | 1 | chatterbox | ResembleAI/chatterbox | 5bb1f6ee58e50c3b8d408bc82a6d3740c2db6e18 | 5de7a54aa4e5e2baadb0182dde554908b48b85c2 | ce00270c445f1a0789e9542db88ee0fd81ba98273436a956cdcc331d5a1a198c | clone | train | {"cfg_weight": 0.5, "exaggeration": 0.5, "temperature": 0.8} | 2,491,358,863 | 4.08 | 1 | 0.9678 | How many hard questions has user algorithm master solved on leak code? | 0.2 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-04490 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"nal_query\", \"description\": \"Converts a given number to its corresponding word representation in Spanish using the Numeros-a-Letras API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"num\": {\"description\": \"The number to be converted to words.\", \"type\": \"str\", \"default\": \"0\... | [
"On '2022-12-31', the exchange rate of IDR to USD was approximately 14,050 IDR for 1 USD.",
"Apologies, I'm unable to provide the exact exchange rate for that specific date as I don't have real-time data access."
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | What was the exchange rate of IDR to USD on '2022-12-31'? | en | What was the exchange rate of IDR to USD on '2022-12-31'? | long_number | When2Call | When2Call:train:4490 | null | null | null | train | When2Call:train:4490 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | a6706458854ee66a8c4dd339e5b4f5f09db3d517b81ef4d46ceb9bb75bbcadec | clone | train | {} | 1,592,269,065 | 5.09 | 1 | 1 | What was the exchange rate of IDR to USD on 2022 1231? | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03933 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"software_assets\", \"description\": \"Fetch details of SaaS/Software tools based on specified feature tags or software name.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"q\": {\"description\": \"Query string representing the list of feature tags or the name of the software for which deta... | [
"Apologies, but I'm unable to provide a solution as I don't have the capability to execute or modify functions like 'convert_url_to_image'.",
"<TOOLCALL>[{\"name\": \"convert_url_to_image\", \"arguments\": {\"url\": \"https://www.example.com\", \"timeout\": 10000}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | You are a web developer who wants to convert a URL to an image but the page takes a long time to load. You want to set a timeout of 10000 milliseconds. How can you use the 'convert_url_to_image' function to achieve this? | en | You are a web developer who wants to convert a URL to an image but the page takes a long time to load. You want to set a timeout of 10000 milliseconds. How can you use the 'convert_url_to_image' function to achieve this? | handle | When2Call | When2Call:train:3933 | null | null | null | train | When2Call:train:3933 | 1 | supertonic | Supertone/supertonic-3 | 3cadd1ee6394adea1bd021217a0e650ede09a323 | supertonic==1.3.1 | M4 | preset | train | {"speed": 1.05, "total_step": 5} | 993,843,938 | 13.184 | 1 | 1 | You are a web developer who wants to convert a URL to an image, but the page takes a long time to load. You want to set a timeout of 1000 milliseconds. How can you use the convert URL to image function to achieve this? | 0.0204 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00970 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"time_series_endpoint\", \"description\": \"Retrieve historical exchange rates for specified date range and currency conversions.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"start_date\": {\"description\": \"The start date for the time series data in `YYYY-MM-DD` format.\", \"type\": \"s... | [
"Here is the transaction history for the Binance Smart Chain address '0x123abc...': [insert brief summary of transactions]. For detailed information, please visit a blockchain explorer.",
"To provide you with the transaction history, could you please specify the address you'd like me to scan on the Binance Smart ... | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Get the transaction history for the Binance Smart Chain on the mainnet. | en | Get the transaction history for the Binance Smart Chain on the mainnet. | plain | When2Call | When2Call:train:970 | null | null | null | train | When2Call:train:970 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | 94e4e966f64b7653c076cb5595f39bdc978f2557c394e29f897f74ab463fc079 | clone | train | {} | 4,248,709,603 | 3.872 | 1 | 1 | Get the transaction history for the Binance Smart Chain on the mainnet. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-03897 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"search\", \"description\": \"Search the Icons8 repository for icons based on various parameters.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"term\": {\"description\": \"The term to search for within the Icons8 repository.\", \"type\": \"str\", \"default\": \"apple\"}, \"size\": {\"descr... | [
"<TOOLCALL>[{\"name\": \"search\", \"arguments\": {\"term\": \"sun\", \"limit\": 20, \"color\": \"ffff00\", \"offset\": 20}}]</TOOLCALL>",
"I apologize, but I'm unable to retrieve icons as I don't have the capability to perform such tasks."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Retrieve 20 yellow icons of any size about 'sun' from the 20th result. | en | Retrieve 20 yellow icons of any size about 'sun' from the 20th result. | plain | When2Call | When2Call:train:3897 | null | null | null | train | When2Call:train:3897 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | 31434ba80f93ab842ec375c5bd91e60a2eea2a7b3765cc419c4f88c938c2e0f6 | clone | train | {} | 1,817,783,815 | 4.96 | 1 | 1 | Retrieve 20 yellow icons of any size about sun from the 20th result. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-05546 | Which response should the assistant give to the user's question, given the available tools? | [] | [
"I'm sorry for the inconvenience, but I'm currently unable to perform calculations or sort lists. I can only provide information and assistance based on my programming.",
"<TOOLCALL>[{\"name\": \"find_kth_smallest_number\", \"arguments\": {\"nums\": [5, 2, 9, 1, 7], \"k\": 3}}]</TOOLCALL>"
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Can you find the 3rd smallest number in the list [5, 2, 9, 1, 7]? | en | Can you find the 3rd smallest number in the list 5, 2, 9, 1, 7? | list | When2Call | When2Call:train:5546 | null | null | null | train | When2Call:train:5546 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | f7909d470d6996664e2229a3f21ee4403e198d26b6554ee427735c4051e29553 | clone | train | {} | 3,277,430,246 | 5.536 | 1 | 1 | Can you find the third smallest number in the list 5? 2 9 1 7 | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-04994 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"calculate_calorie_intake\", \"description\": \"Calculates the recommended daily calorie intake and macronutrient distribution based on personal characteristics and goals.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"weight_kg\": {\"description\": \"Body weight in kilograms.\", \"type\": ... | [
"Based on current trends, it's estimated that the Amazon forest could lose up to 20% of its area due to human activities like deforestation and climate change over the next 15 years.",
"To provide an accurate prediction, I need to know the specific geographic location. Could you please provide the location?"
] | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | Predict the change in forest area for the next 15 years, considering human impact. | en | Predict the change in forest area for the next 15 years, considering human impact. | plain | When2Call | When2Call:train:4994 | null | null | null | train | When2Call:train:4994 | 1 | pocket | kyutai/pocket-tts | 19f95fe2df36e79fbd9f10008595cc4c977a0fcc | pocket-tts==3.3.0 | b1e2e3f2622e4566610c2fb3a61bbd2d3cef063bcde4fc4a99e9a3730cd216e0 | clone | train | {} | 4,191,426,006 | 4.16 | 1 | 1 | Predict the change in forest area for the next 15 years, considering human impact. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00364 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"stock_get_key_statistics_data\", \"description\": \"Gets key statistics data for a given stock symbol and region using the RapidAPI service.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"symbol\": {\"description\": \"The ticker symbol for the stock (e.g., 'AAPL' for Apple Inc.).\", \"type... | [
"To fetch data for aggressive small cap stocks starting from index 100, you can use a stock screener tool. Set the market cap filter to small, the investment style to aggressive, and start your search from index 100.",
"I'm sorry for the inconvenience, but I'm unable to perform that task as I don't have the capab... | [
"rejected",
"chosen"
] | option_1 | 1 | chosen | I want to fetch data for aggressive small cap stocks starting from index 100. | en | I want to fetch data for aggressive small cap stocks starting from index 100. | long_number | When2Call | When2Call:train:364 | null | null | null | train | When2Call:train:364 | 1 | omni | k2-fsa/OmniVoice | c5fdb5ccb189668d56333f77ba2629f4cd7535f4 | 08be0b4ccbac3e13e374e86fbfead4b4cac343e2 | f7909d470d6996664e2229a3f21ee4403e198d26b6554ee427735c4051e29553 | clone | train | {} | 2,535,205,824 | 4.32 | 1 | 0.99 | I want to fetch data for aggressive small cap stocks starting from index 100. | 0 | wer | Systran/faster-whisper-base.en | 0 | true | |
when2call-train-00677 | Which response should the assistant give to the user's question, given the available tools? | [
"{\"name\": \"auto_complete_deprecated\", \"description\": \"Fetches suggestions for a given term or phrase from the Trading View API.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"text\": {\"description\": \"Any term or phrase for which to get suggestions.\", \"type\": \"str\", \"default\": \"tesla\"}... | [
"To provide the most relevant suggestions, could you please specify the topic or term related to cryptocurrencies?",
"Aquí tienes algunas sugerencias: Bitcoin Cash, Bitcoin Gold y Bitcoin SV. Estas criptomonedas están relacionadas con Bitcoin y comparten su nombre."
] | [
"chosen",
"rejected"
] | option_0 | 0 | chosen | Obtain suggestions for cryptocurrencies related to the given topic in Spanish. | en | Obtain suggestions for cryptocurrencies related to the given topic in Spanish. | plain | When2Call | When2Call:train:677 | null | null | null | train | When2Call:train:677 | 1 | sopro | samuel-vitorino/sopro-v2-turbo | 6c4aa6bba196ea3c8ee8e3e8b5dabe7a715fac92 | 7bfcf9a0539d274f6593959a21da948f506ceee1 | 26edc049ea290460fe30f7a8355d915fd77baa12ad7d50ba0623464029d732e8 | clone | train | {} | 3,935,197,794 | 5.579 | 1 | 1 | Obtain suggestions for cryptocurrencies related to the given topic in Spanish. | 0 | wer | Systran/faster-whisper-base.en | 0 | true |
Decider Audio – When2Call (spoken)
Part of the Decider Audio collection.
The model hears a user question. It reads the assistant's tool definitions and the candidate replies as text. Then it must decide which reply the assistant should give. The reply can call a tool, ask a follow-up question, or say that the assistant cannot help. The words alone can solve the task, so the challenge is to understand them from speech and to match them with the tools.
12,622 rows from NVIDIA's When2Call, about 21.67 hours of audio over all rows. 11,269 distinct spoken questions were rendered by 6 text-to-speech models.
| Split | Source | Rows | Spoken clips | Distinct voices | Hours | Re-rendered | Failing checks |
|---|---|---|---|---|---|---|---|
train |
When2Call train_pref (2 options) |
8,970 | 8,970 | 204 | 14.81 | 199 | 11 |
test |
When2Call test MCQ (4 options, official evaluation set) |
3,652 | 2,299 | 202 | 6.87 | 56 | 1 |
No speaker appears in more than one split. The test rows are spoken by test-split voices only.
Task format
| Input | audio (the spoken question) + instruction + tools + choices as text |
| Target text | answer, an option key such as option_3 |
| Resolved label | answer_label, which is tool_call, request_for_info or cannot_answer in test and chosen in train |
Here is a suggested prompt. The model sees the audio at the marked position.
Available tools:
[{"name": "...", "description": "...", "parameters": {...}}, ...]
User question:
<audio>
Which response should the assistant give to the user's question, given the available tools?
option_0: ...
option_1: ...
Answer with the option key.
spoken_text is the original question. It is included for quality control and ASR baselines. Do not give it to a model as input when training an audio model.
How it is made
- Text.
test= When2Calltest/when2call_test_mcq.jsonl(3,652 rows, the official multiple-choice evaluation, must not be used for training). It has four candidate replies in the fixed orderdirect,tool_call,request_for_info,cannot_answer, anddirectis never correct.train=train/when2call_train_pref.jsonl(9,000 rows). It has the chosen and the rejected reply in a seeded shuffle per row. Questions, options and answer keys are identical to the Decision Index projection of When2Call. The SFT training file (single replies, no options) is not used. - Content-only speech. Only the user question is spoken. Tools, instruction, options and answer never enter the audio.
- One clip per test question. When2Call asks many test questions twice or more. Once with the tool it needs (
tool_call) and once with that tool removed (cannot_answer), sometimes also asrequest_for_info. These rows share one clip (spoken_unit,spoken_unit_rows> 1), so the audio is identical and only the tools decide. 2,640testrows share their clip with another row. 2,299 distinct test questions were spoken.trainrows are spoken one by one. No row was merged. - Model and voice assignment by question type. The six models are not equally good at everything. A pilot test rendered 43 hard questions with every model and the same voice. On plain questions all models are equally good, but lists, ID tokens, URLs, handles and long numbers separate them. Each question is classified (
speech_category, the first matching type in the order list, URL, ID token, handle, long number) and gets the best-ranked model of its type that still has room. Types are assigned from the smallest to the largest, and plain questions fill the remaining slots least-used-first. Every model still renders the same number of clips (±1) per split. The assignment does not look at the answer. Lists and ID tokens are spoken without list brackets and without the quotes around values (tts_text, e.g. "Split the list ['a', 'b']" becomes "Split the list a, b"). This made lists clearly easier for every model except OmniVoice. The voice is picked by a seeded hash from the pool of the matching split.
| Question type | Spoken as | Ranking from the pilot | Clips | chatterbox |
omni |
pocket |
qwen |
sopro |
supertonic |
|---|---|---|---|---|---|---|---|---|---|
| handle | original | supertonic > omni > qwen > sopro > chatterbox > pocket | 218 | 1 | 0 | 0 | 1 | 0 | 216 |
| url | original | supertonic > omni > chatterbox > qwen > sopro > pocket | 372 | 1 | 3 | 1 | 3 | 1 | 363 |
| list | cleaned | sopro > pocket > chatterbox > qwen > supertonic > omni | 518 | 0 | 0 | 0 | 0 | 518 | 0 |
| id token | cleaned | qwen > omni > pocket > chatterbox > sopro > supertonic | 625 | 0 | 0 | 0 | 625 | 0 | 0 |
| long number | original | omni > pocket > qwen > supertonic > chatterbox > sopro | 2,499 | 0 | 1,830 | 669 | 0 | 0 | 0 |
| plain | original | balanced | 6,988 | 1,878 | 0 | 1,209 | 1,253 | 1,360 | 1,288 |
| non-English | own language | OmniVoice only | 49 | 0 | 49 | 0 | 0 | 0 | 0 |
- Non-English questions. 49 spoken questions (85 rows) are not English. They are 11
mixed, 11zh, 6ko, 6id, 3vi, 3es, 3th, 2hi, 2pt, 1hi-latnand 1de. They were found by script (Han, kana, Hangul, Thai, Devanagari), a language detector and a keyword search, and reviewed by hand. They are always spoken by OmniVoice, the only multilingual model, in their language (language-agnostic mode formixed). It uses the same English reference voices (cross-lingual cloning). They were moved to OmniVoice after the balanced assignment, so the other questions keep their models. They are checked withSystran/faster-whisper-large-v3and the character error rate. - Voices. Voice-cloning models use English reference clips from a private voice collection (not redistributed). Speakers are assigned to
train,validationandtestbefore any voice is chosen, and every row uses only voices of its own split, so test speakers never occur in training. Preset-voice models use the presets of their split's partition. 406 distinct voices appear in total. - Audio. 24 kHz mono, 16-bit FLAC. Clips whose peak exceeded full scale were scaled to 0.99 (
audio_gain). Flat-topped (clipped) renders were rejected instead. - Re-rendering, never removal. Every render was checked. A render fails if the ASR word error rate is above 0.5, if there are more than 30 characters per second, or if 3 or more extra words appear that the question does not contain (10 % of its words for questions longer than 30 words). For non-English questions the error rate is the character error rate and the extra-word check does not apply. A failed question was rendered again with a different voice and seed, using the same model (
render_attempt> 0, up to 5 attempts). 255 rows carry such a re-render. 12 rows still failed after the last attempt and keep their best render withpasses_checks = false. A second repair pass rendered 137 clips again. 42 of them still failed a check and 95 passed the word check but leave out part of their text. A clip got up to 10 more attempts on its own model. Clips on Supertonic and clips that a model could not render at all then moved to another model for up to 3 more. 12 clips now come from another model, whichtts_modelshows. A clip now also passes when its only problem is the word error rate or extra words and its transcript sounds like the text. This check compares phonemes after numbers and units are spelled out, so "topup" and "top up" or "1960" and "nineteen sixty" match. A clip that leaves out part of its text does not pass this way. A truncated clip is replaced only when the new render leaves out less.
TTS models
| TTS model | Checkpoint | Voice mode | Rows | Distinct voices | Hours | Mean WER | Exact | WER ≤ 0.2 | Re-rendered |
|---|---|---|---|---|---|---|---|---|---|
chatterbox |
ResembleAI/chatterbox | clone | 2,103 | 376 | 2.85 | 0.0407 | 70.0% | 95.0% | 14 |
omni |
k2-fsa/OmniVoice | clone | 2,125 | 377 | 4.48 | 0.0394 | 59.2% | 97.0% | 29 |
pocket |
kyutai/pocket-tts | clone | 2,109 | 370 | 3.50 | 0.0390 | 65.7% | 96.5% | 49 |
qwen |
Qwen/Qwen3-TTS-12Hz-0.6B-Base | clone | 2,107 | 371 | 3.63 | 0.0558 | 59.3% | 91.6% | 87 |
sopro |
samuel-vitorino/sopro-v2-turbo | clone | 2,110 | 371 | 3.70 | 0.0334 | 72.7% | 96.1% | 32 |
supertonic |
Supertone/supertonic-3 | preset | 2,068 | 7 | 3.52 | 0.0556 | 60.3% | 92.9% | 44 |
Counts are rows. A shared test clip counts once per row. OmniVoice, Chatterbox and Qwen are the original PyTorch checkpoints.
Checks
Every clip was transcribed with Systran/faster-whisper-base.en and compared with spoken_text after lower-casing, removing punctuation and spelling out digits.
- Mean error rate over rows 0.0439 (word error rate, and character error rate for the 85 non-English rows). 8,144 rows (64.5%) transcribe exactly. The 90th percentile is 0.143 and the 99th is 0.385.
- 94.9% of the rows have WER ≤ 0.2. Questions contain IDs, handles, URLs and file names ("@jane_doe", "project ID badae") that TTS reads in its own way and ASR spells differently. Part of the remaining error is in such tokens. Filter on
asr_werorpasses_checksfor a cleaner subset.
ASR agreement shows that the words were spoken. It does not show that the voice sounds natural or that it resembles the reference voice. Only a pilot sample was reviewed by listening. Voice similarity differs between models.
Rows left out
The dataset follows When2Call row by row with one exception. 30 train questions about Sudoku spell out the whole board ([[5,3,'.','.',7,...],...], '53..7....'). Read aloud, that is a minute of digits and dots that no model rendered intelligibly in five attempts, so these rows are not included. Sudoku questions without a board are kept. The source lines of the left-out train rows are 48, 708, 1179, 1225, 1237, 1459, 1781, 2241, 2341, 2768, 2885, 3097, 3351, 3444, 3459, 3531, 3840, 3850, 4514, 4528, 4816, 5074, 5628, 6494, 6762, 6818, 8352, 8436, 8519 and 8832.
Peculiarities of the source
- Prompt-like questions. 3
testrows have a long system prompt ("You are MemGPT, …") as their question and 15 rows contain an artifact "Revised question: …" that repeats the question. They are spoken in full, which gives the longest clips (41 rows have more than 400 characters). - Repeated questions. Besides the shared test clips, 657 questions occur more than once inside
train(with different tools or replies), and 2trainrows have a question that also occurs intest. To drop those, use this code.
import re
from datasets import load_dataset
def key(text):
return " ".join(re.findall(r"[a-z0-9']+", text.lower().replace("-", " ")))
train = load_dataset("SebastianBodza/decider-audio-when2call", split="train")
test = load_dataset("SebastianBodza/decider-audio-when2call", split="test")
test_questions = {key(t) for t in test["spoken_text"]}
train_clean = train.filter(lambda row: key(row["spoken_text"]) not in test_questions)
# optionally also keep only renders that passed every check
train_clean = train_clean.filter(lambda row: row["passes_checks"])
Fields
| Field | Type | Description |
|---|---|---|
id |
string | when2call-<split>-<index>, index = line of the source file |
audio |
Audio, 24 kHz mono FLAC | Spoken question |
instruction |
string | Task text |
tools |
list of strings | Tool definitions as in the source, one JSON object per entry (may be empty) |
choices |
list of strings | Candidate replies. choices[i] belongs to option key option_i. Tool calls are JSON (test) or <TOOLCALL>[...]</TOOLCALL> (train), as in the source |
choice_labels |
list of strings | What each candidate is. Values are direct, tool_call, request_for_info, cannot_answer (test) and chosen, rejected (train) |
answer |
string | Option key of the correct reply |
answer_index |
int | Index of the correct reply in choices |
answer_label |
string | choice_labels[answer_index] |
spoken_text |
string | The user question as in the source |
tts_text |
string | Text the TTS model got. This is spoken_text, or for lists and ID tokens spoken_text without list brackets and delimiter quotes |
speech_category |
string | Question type that chose the model (list, url, id_token, handle, long_number or plain) |
spoken_language |
string | en, a language code (zh, ko, th, hi, id, vi, es, pt, de), hi-latn (Hindi written in Latin script) or mixed (English with words or a passage in another script) |
source_dataset, source_id, source_split |
string | Provenance (When2Call:<split>:<line>) |
source_uuid, source_subset, target_tool |
string | Test only. When2Call uuid, the BFCL subset it was built from, and the target tool (JSON) |
spoken_unit, spoken_unit_rows |
string, int | Id of the spoken clip and the number of rows that share it |
tts_model |
string | Project key of the TTS model |
tts_repo, tts_revision, tts_sdk_revision |
string | Checkpoint, pinned revision and runtime revision |
voice_id |
string | Reference-voice id (clone models) or preset name |
voice_kind |
clone or preset |
How the voice was chosen |
voice_split |
string | Split the voice belongs to. Always equal to the row's split |
tts_settings |
JSON string | Generation settings |
tts_seed |
int | Seed of the generation call |
duration_seconds, n_segments |
float, int | Length of the clip. Questions above 400 characters are split into ordered segments joined with 180 ms of silence |
audio_gain |
float | 1.0 unless the clip overshot full scale and was scaled down (823 rows) |
asr_text, asr_wer, asr_metric, asr_model |
string, float, string, string | Round-trip check, see above. asr_metric says whether asr_wer is a word (wer) or character error rate (cer, non-English rows) |
render_attempt |
int | 0 for the first render, higher if the question was rendered again after a failed check |
passes_checks |
bool | False if even the last attempt failed a check |
Clips are 0.52–303.71 s long (mean 6.18 s per row).
Limitations
- Synthetic speech only. All audio comes from TTS models, there is no human speech. Expect a domain gap to real users, accents, noise and telephone channels.
- Train and test differ in format.
trainhas two options (chosen vs. rejected reply, no category labels),testfour (one per category). A model trained only ontrainnever sees four options. This is the source's design. - Answer distribution follows the source.
testhas 1,295tool_call, 1,062request_for_infoand 1,295cannot_answer. - Questions were written as text. Many look like typed requests (IDs, quotes, URLs, JSON-like values) rather than natural speech.
- Mostly English. 85 rows are in other languages or mixed. All are spoken by OmniVoice with English reference voices, so they carry an English accent of varying strength. Speaker attributes (gender, age, accent) are not annotated.
License and attribution
The dataset mixes several upstream terms. Check them before any use beyond private research.
- Text. When2Call, NVIDIA, CC BY 4.0 (paper, data). Its test questions and tools derive from BFCL v2 Live.
- Speech from OmniVoice. The upstream weights are reported as CC BY-NC 4.0 (non-commercial). Verify on
k2-fsa/OmniVoice. Thetts_repocolumn shows which checkpoint rendered a clip. - Other generators. Qwen3-TTS Apache-2.0, Sopro Apache-2.0, Supertonic OpenRAIL (use restrictions apply), Pocket TTS CC BY 4.0 (attribution required, and Kyutai's PyTorch weights are gated behind their terms), Chatterbox MIT.
- Reference voices. Taken from a private collection and not redistributed.
Because of the non-commercial and use-restricted components the dataset is marked license: other. Do not use it commercially before the terms above are resolved.
@inproceedings{ross-etal-2025-when2call,
title = {When2Call: When (not) to Call Tools},
author = {Ross, Hayley and Mahabaleshwarkar, Ameya Sunil and Suhara, Yoshi},
booktitle = {Proceedings of NAACL 2025},
year = {2025},
url = {https://aclanthology.org/2025.naacl-long.174/}
}
- Downloads last month
- 20