tools
stringlengths 151
9.9k
| conversations
listlengths 1
11
| answer
stringlengths 33
4.78k
| raw_system
stringlengths 406
471
| data_source
stringclasses 1
value | prompt
listlengths 2
12
| ability
stringclasses 1
value | reward_model
dict | extra_info
dict |
|---|---|---|---|---|---|---|---|---|
[{"name": "getplatformlist", "description": "Fetches the list of platforms with their IDs and names from the API endpoint as JSON.", "parameters": {"page": {"description": "Page number to retrieve the platform list from the API.", "type": "int", "default": "0"}}}, {"name": "user_aliases", "description": "Gets a list of user aliases and the timestamps when they were changed.", "parameters": {"is_id": {"description": "The Steam ID of the user for whom the aliases are retrieved.", "type": "str", "default": "sindex"}}}, {"name": "gamesdata", "description": "Fetches data for a specific game using its APP_ID from the Steam Games Data API.", "parameters": {"app_id": {"description": "The unique identifier for the game whose data is to be fetched.", "type": "int", "default": "1581770"}}}, {"name": "get_card_by_slug", "description": "Fetches card information by slug from the Hearthstone API.", "parameters": {"slug": {"description": "The slug identifier for the card.", "type": "str", "default": "a-light-in-the-darkness"}}}]
|
[
{
"from": "human",
"value": "Retrieve platform list for page six. Also, find user aliases for the Steam ID '76561197998991414'. In addition, get game data for app with ID '730'. Finally, gather card information for the Hearthstone card with slug 'EX1_557'."
}
] |
[{"name": "getplatformlist", "arguments": {"page": 6}}, {"name": "user_aliases", "arguments": {"is_id": "76561197998991414"}}, {"name": "gamesdata", "arguments": {"app_id": 730}}, {"name": "get_card_by_slug", "arguments": {"slug": "EX1_557"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"getplatformlist\", \"description\": \"Fetches the list of platforms with their IDs and names from the API endpoint as JSON.\", \"parameters\": {\"page\": {\"description\": \"Page number to retrieve the platform list from the API.\", \"type\": \"int\", \"default\": \"0\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"user_aliases\", \"description\": \"Gets a list of user aliases and the timestamps when they were changed.\", \"parameters\": {\"is_id\": {\"description\": \"The Steam ID of the user for whom the aliases are retrieved.\", \"type\": \"str\", \"default\": \"sindex\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"gamesdata\", \"description\": \"Fetches data for a specific game using its APP_ID from the Steam Games Data API.\", \"parameters\": {\"app_id\": {\"description\": \"The unique identifier for the game whose data is to be fetched.\", \"type\": \"int\", \"default\": \"1581770\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"get_card_by_slug\", \"description\": \"Fetches card information by slug from the Hearthstone API.\", \"parameters\": {\"slug\": {\"description\": \"The slug identifier for the card.\", \"type\": \"str\", \"default\": \"a-light-in-the-darkness\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Retrieve platform list for page six. Also, find user aliases for the Steam ID '76561197998991414'. In addition, get game data for app with ID '730'. Finally, gather card information for the Hearthstone card with slug 'EX1_557'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"getplatformlist\", \"arguments\": {\"page\": 6}}, {\"name\": \"user_aliases\", \"arguments\": {\"is_id\": \"76561197998991414\"}}, {\"name\": \"gamesdata\", \"arguments\": {\"app_id\": 730}}, {\"name\": \"get_card_by_slug\", \"arguments\": {\"slug\": \"EX1_557\"}}]",
"style": "rule"
}
|
{
"index": 0,
"split": "train"
}
|
[{"name": "measure_distance", "description": "Calculates the distance between two geographic locations based on their latitude and longitude coordinates. The unit of measurement for the distance can be specified.", "parameters": {"lon2": {"description": "Longitude of the second location.", "type": "int", "default": "31.23788289124186"}, "lat1": {"description": "Latitude of the first location.", "type": "int", "default": "31.1991806"}, "lon1": {"description": "Longitude of the first location.", "type": "int", "default": "29.8951716"}, "lat2": {"description": "Latitude of the second location.", "type": "int", "default": "30.02313795"}, "unit": {"description": "Unit of distance measurement. Valid options are 'km' (default), 'mi', 'ft', and 'yd'.", "type": "str, optional", "default": "km"}}}, {"name": "timezones", "description": "Fetch a list of the world's time zones with various filtering options.", "parameters": {"offset": {"description": "Filter on non-daylight savings offset in hours.", "type": "str, optional", "default": ""}, "name": {"description": "Filter as \"contains\" on timezone name in the chosen language.", "type": "str, optional", "default": ""}, "code": {"description": "Filter on 3-letter non-daylight savings timezone code (not unique).", "type": "str, optional", "default": ""}, "timezone": {"description": "Filter on timezone name in English.", "type": "str, optional", "default": ""}, "daylights_offset": {"description": "Filter on daylight savings offset in hours.", "type": "str, optional", "default": ""}, "lang": {"description": "Language selection ISO 639-1 two-letter code. Overrides the Accept-Language header. Default is 'de'.", "type": "str, optional", "default": "de"}, "daylights_code": {"description": "Filter on daylight savings letter code (not unique).", "type": "str, optional", "default": ""}, "fields": {"description": "Select fields to return in the result. Default is 'offset,daylights_offset,daylights,daylights_code,timezone'.", "type": "str, optional", "default": "offset,daylights_offset,daylights,daylights_code,timezone"}, "daylights": {"description": "Filter on daylight savings/non-daylight savings time zones (`true` or `false`).", "type": "str, optional", "default": ""}}}, {"name": "get_countries", "description": "Fetches a list of countries based on various search criteria.", "parameters": {"keyword": {"description": "Search countries by keyword. Defaults to 'ind'.", "type": "str, optional", "default": "ind"}, "subregion_id": {"description": "Filter countries by subregion ID.", "type": "str, optional", "default": ""}, "state_id": {"description": "Filter countries by state ID.", "type": "str, optional", "default": ""}, "timezone_id": {"description": "Filter countries by timezone ID.", "type": "str, optional", "default": ""}, "region_id": {"description": "Filter countries by subregion ID.", "type": "str, optional", "default": ""}, "perpage": {"description": "Number of results per page. Defaults to 50.", "type": "int, optional", "default": "50"}, "is_id": {"description": "Filter countries by country ID.", "type": "str, optional", "default": ""}, "page": {"description": "Number of results per page. Defaults to 50.", "type": "int, optional", "default": "1"}}}]
|
[
{
"from": "human",
"value": "Retrieve time zones that are in daylight savings time and have a code containing 'EST'."
}
] |
[{"name": "timezones", "arguments": {"daylights": "true", "code": "EST"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"measure_distance\", \"description\": \"Calculates the distance between two geographic locations based on their latitude and longitude coordinates. The unit of measurement for the distance can be specified.\", \"parameters\": {\"lon2\": {\"description\": \"Longitude of the second location.\", \"type\": \"int\", \"default\": \"31.23788289124186\"}, \"lat1\": {\"description\": \"Latitude of the first location.\", \"type\": \"int\", \"default\": \"31.1991806\"}, \"lon1\": {\"description\": \"Longitude of the first location.\", \"type\": \"int\", \"default\": \"29.8951716\"}, \"lat2\": {\"description\": \"Latitude of the second location.\", \"type\": \"int\", \"default\": \"30.02313795\"}, \"unit\": {\"description\": \"Unit of distance measurement. Valid options are 'km' (default), 'mi', 'ft', and 'yd'.\", \"type\": \"str, optional\", \"default\": \"km\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"timezones\", \"description\": \"Fetch a list of the world's time zones with various filtering options.\", \"parameters\": {\"offset\": {\"description\": \"Filter on non-daylight savings offset in hours.\", \"type\": \"str, optional\", \"default\": \"\"}, \"name\": {\"description\": \"Filter as \\\"contains\\\" on timezone name in the chosen language.\", \"type\": \"str, optional\", \"default\": \"\"}, \"code\": {\"description\": \"Filter on 3-letter non-daylight savings timezone code (not unique).\", \"type\": \"str, optional\", \"default\": \"\"}, \"timezone\": {\"description\": \"Filter on timezone name in English.\", \"type\": \"str, optional\", \"default\": \"\"}, \"daylights_offset\": {\"description\": \"Filter on daylight savings offset in hours.\", \"type\": \"str, optional\", \"default\": \"\"}, \"lang\": {\"description\": \"Language selection ISO 639-1 two-letter code. Overrides the Accept-Language header. Default is 'de'.\", \"type\": \"str, optional\", \"default\": \"de\"}, \"daylights_code\": {\"description\": \"Filter on daylight savings letter code (not unique).\", \"type\": \"str, optional\", \"default\": \"\"}, \"fields\": {\"description\": \"Select fields to return in the result. Default is 'offset,daylights_offset,daylights,daylights_code,timezone'.\", \"type\": \"str, optional\", \"default\": \"offset,daylights_offset,daylights,daylights_code,timezone\"}, \"daylights\": {\"description\": \"Filter on daylight savings/non-daylight savings time zones (`true` or `false`).\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"get_countries\", \"description\": \"Fetches a list of countries based on various search criteria.\", \"parameters\": {\"keyword\": {\"description\": \"Search countries by keyword. Defaults to 'ind'.\", \"type\": \"str, optional\", \"default\": \"ind\"}, \"subregion_id\": {\"description\": \"Filter countries by subregion ID.\", \"type\": \"str, optional\", \"default\": \"\"}, \"state_id\": {\"description\": \"Filter countries by state ID.\", \"type\": \"str, optional\", \"default\": \"\"}, \"timezone_id\": {\"description\": \"Filter countries by timezone ID.\", \"type\": \"str, optional\", \"default\": \"\"}, \"region_id\": {\"description\": \"Filter countries by subregion ID.\", \"type\": \"str, optional\", \"default\": \"\"}, \"perpage\": {\"description\": \"Number of results per page. Defaults to 50.\", \"type\": \"int, optional\", \"default\": \"50\"}, \"is_id\": {\"description\": \"Filter countries by country ID.\", \"type\": \"str, optional\", \"default\": \"\"}, \"page\": {\"description\": \"Number of results per page. Defaults to 50.\", \"type\": \"int, optional\", \"default\": \"1\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Retrieve time zones that are in daylight savings time and have a code containing 'EST'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"timezones\", \"arguments\": {\"daylights\": \"true\", \"code\": \"EST\"}}]",
"style": "rule"
}
|
{
"index": 1,
"split": "train"
}
|
[{"name": "get_train_data_mrt_id_id_get", "description": "Fetch train data using MRT ID from the RapidAPI service.", "parameters": {"is_id": {"description": "The MRT ID to fetch the train data for.", "type": "str", "default": ""}}}, {"name": "us_dc", "description": "Fetches current or historical gas price data for Washington D.C.", "parameters": {"date": {"description": "The date for which to fetch historical gas price data. Defaults to None.", "type": "str, optional", "default": ""}}}, {"name": "interior_colors", "description": "Fetches interior color details for a car based on various query parameters.", "parameters": {"model": {"description": "The model of the car.", "type": "str, optional", "default": ""}, "name": {"description": "The name of the color.", "type": "str, optional", "default": ""}, "trim": {"description": "The trim level of the car.", "type": "str, optional", "default": ""}, "page": {"description": "The page number for pagination.", "type": "int, optional", "default": ""}, "direction": {"description": "The sorting direction, either 'asc' or 'desc'. Default is 'asc'.", "type": "str, optional", "default": "asc"}, "limit": {"description": "The maximum number of results to return.", "type": "int, optional", "default": ""}, "make_model_trim_id": {"description": "The unique ID for make, model, and trim combination.", "type": "str, optional", "default": ""}, "year": {"description": "The year of the car.", "type": "str, optional", "default": ""}, "rgb": {"description": "The RGB value of the color.", "type": "str, optional", "default": ""}, "sort": {"description": "The field to sort by. Default is 'id'.", "type": "str, optional", "default": "id"}, "verbose": {"description": "If 'yes', includes detailed information about make, model, and trim. Default is 'yes'.", "type": "str, optional", "default": "yes"}, "json": {"description": "A URL encoded JSON array of conditions to filter results.", "type": "str, optional", "default": ""}, "make_id": {"description": "The ID of the car make.", "type": "str, optional", "default": ""}, "make": {"description": "The make of the car.", "type": "str, optional", "default": ""}, "make_model_id": {"description": "The ID for make and model combination.", "type": "str, optional", "default": ""}}}]
|
[
{
"from": "human",
"value": "Get current gas prices for Washington D.C."
}
] |
[{"name": "us_dc", "arguments": {}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_train_data_mrt_id_id_get\", \"description\": \"Fetch train data using MRT ID from the RapidAPI service.\", \"parameters\": {\"is_id\": {\"description\": \"The MRT ID to fetch the train data for.\", \"type\": \"str\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"us_dc\", \"description\": \"Fetches current or historical gas price data for Washington D.C.\", \"parameters\": {\"date\": {\"description\": \"The date for which to fetch historical gas price data. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"interior_colors\", \"description\": \"Fetches interior color details for a car based on various query parameters.\", \"parameters\": {\"model\": {\"description\": \"The model of the car.\", \"type\": \"str, optional\", \"default\": \"\"}, \"name\": {\"description\": \"The name of the color.\", \"type\": \"str, optional\", \"default\": \"\"}, \"trim\": {\"description\": \"The trim level of the car.\", \"type\": \"str, optional\", \"default\": \"\"}, \"page\": {\"description\": \"The page number for pagination.\", \"type\": \"int, optional\", \"default\": \"\"}, \"direction\": {\"description\": \"The sorting direction, either 'asc' or 'desc'. Default is 'asc'.\", \"type\": \"str, optional\", \"default\": \"asc\"}, \"limit\": {\"description\": \"The maximum number of results to return.\", \"type\": \"int, optional\", \"default\": \"\"}, \"make_model_trim_id\": {\"description\": \"The unique ID for make, model, and trim combination.\", \"type\": \"str, optional\", \"default\": \"\"}, \"year\": {\"description\": \"The year of the car.\", \"type\": \"str, optional\", \"default\": \"\"}, \"rgb\": {\"description\": \"The RGB value of the color.\", \"type\": \"str, optional\", \"default\": \"\"}, \"sort\": {\"description\": \"The field to sort by. Default is 'id'.\", \"type\": \"str, optional\", \"default\": \"id\"}, \"verbose\": {\"description\": \"If 'yes', includes detailed information about make, model, and trim. Default is 'yes'.\", \"type\": \"str, optional\", \"default\": \"yes\"}, \"json\": {\"description\": \"A URL encoded JSON array of conditions to filter results.\", \"type\": \"str, optional\", \"default\": \"\"}, \"make_id\": {\"description\": \"The ID of the car make.\", \"type\": \"str, optional\", \"default\": \"\"}, \"make\": {\"description\": \"The make of the car.\", \"type\": \"str, optional\", \"default\": \"\"}, \"make_model_id\": {\"description\": \"The ID for make and model combination.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Get current gas prices for Washington D.C.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"us_dc\", \"arguments\": {}}]",
"style": "rule"
}
|
{
"index": 2,
"split": "train"
}
|
[{"name": "monthly_mortgage_payment", "description": "Calculates the monthly mortgage payment based on the loan amount, annual interest rate, and loan term.", "parameters": {"loan_amount": {"description": "The total amount of the loan.", "type": "float"}, "annual_interest_rate": {"description": "The annual interest rate as a decimal (e.g., 0.05 for 5%).", "type": "float"}, "loan_term_years": {"description": "The loan term in years.", "type": "int"}}}]
|
[
{
"from": "human",
"value": "Compute the monthly mortgage payment for a $350,000 loan with an annual interest rate of 2.9% for 15 years."
}
] |
[{"name": "monthly_mortgage_payment", "arguments": {"loan_amount": 350000, "annual_interest_rate": 0.029, "loan_term_years": 15}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"monthly_mortgage_payment\", \"description\": \"Calculates the monthly mortgage payment based on the loan amount, annual interest rate, and loan term.\", \"parameters\": {\"loan_amount\": {\"description\": \"The total amount of the loan.\", \"type\": \"float\"}, \"annual_interest_rate\": {\"description\": \"The annual interest rate as a decimal (e.g., 0.05 for 5%).\", \"type\": \"float\"}, \"loan_term_years\": {\"description\": \"The loan term in years.\", \"type\": \"int\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Compute the monthly mortgage payment for a $350,000 loan with an annual interest rate of 2.9% for 15 years.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"monthly_mortgage_payment\", \"arguments\": {\"loan_amount\": 350000, \"annual_interest_rate\": 0.029, \"loan_term_years\": 15}}]",
"style": "rule"
}
|
{
"index": 3,
"split": "train"
}
|
[{"name": "can_attend_all_meetings", "description": "Determines if a person can attend all meetings given a list of meeting time intervals.", "parameters": {"intervals": {"description": "A list of meeting time intervals, where each interval is represented as [start_time, end_time].", "type": "List[List[int]]"}}}, {"name": "cell_density", "description": "Calculates the cell density based on the optical density (OD) and dilution factor.", "parameters": {"od": {"description": "The optical density of the sample.", "type": "float", "default": 1000000000.0}, "dilution": {"description": "The dilution factor applied to the sample.", "type": "int", "default": 1000000000.0}, "factor": {"description": "The calibration factor for converting OD to cell density. Defaults to 1e9.", "type": "float, optional"}}}]
|
[
{
"from": "human",
"value": "Determine if a person can attend all of their meetings scheduled from 9:00 AM to 10:00 AM, 10:30 AM to 11:30 AM, and 12:00 PM to 1:00 PM."
}
] |
[{"name": "can_attend_all_meetings", "arguments": {"intervals": [[900, 1000], [1030, 1130], [1200, 1300]]}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"can_attend_all_meetings\", \"description\": \"Determines if a person can attend all meetings given a list of meeting time intervals.\", \"parameters\": {\"intervals\": {\"description\": \"A list of meeting time intervals, where each interval is represented as [start_time, end_time].\", \"type\": \"List[List[int]]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"cell_density\", \"description\": \"Calculates the cell density based on the optical density (OD) and dilution factor.\", \"parameters\": {\"od\": {\"description\": \"The optical density of the sample.\", \"type\": \"float\", \"default\": 1000000000.0}, \"dilution\": {\"description\": \"The dilution factor applied to the sample.\", \"type\": \"int\", \"default\": 1000000000.0}, \"factor\": {\"description\": \"The calibration factor for converting OD to cell density. Defaults to 1e9.\", \"type\": \"float, optional\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Determine if a person can attend all of their meetings scheduled from 9:00 AM to 10:00 AM, 10:30 AM to 11:30 AM, and 12:00 PM to 1:00 PM.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"can_attend_all_meetings\", \"arguments\": {\"intervals\": [[900, 1000], [1030, 1130], [1200, 1300]]}}]",
"style": "rule"
}
|
{
"index": 4,
"split": "train"
}
|
[{"name": "latest", "description": "Fetches the latest news from Google News for a specified language region.", "parameters": {"lr": {"description": "Language region to get news for, e.g., 'en-US'.", "type": "str", "default": "en-US"}}}, {"name": "top_headlines", "description": "Fetches top headline news articles based on specified parameters.", "parameters": {"language": {"description": "Language code to filter news articles. Default is 'en'. Options include:", "type": "str", "default": "en"}, "category": {"description": "Category of news articles to filter by. Default is 'sports'. Options include:", "type": "str", "default": "sports"}, "country": {"description": "Country code to filter news articles. Default is 'us'. Options include:", "type": "str", "default": "us"}, "sortby": {"description": "Criterion to sort news articles by. Default is 'published_date'.", "type": "str", "default": ""}, "pagesize": {"description": "Number of news articles to return per page. Default is 10.", "type": "int", "default": "10"}}}, {"name": "news_source", "description": "Fetch the latest news stories for a given news source in chronological order.", "parameters": {"source_id": {"description": "The identifier of the news source for which to retrieve the latest news stories.", "type": "str", "default": "bbc"}}}]
|
[
{
"from": "human",
"value": "Please get me the top health news in German from Germany, and also the latest news from the BBC source."
}
] |
[{"name": "top_headlines", "arguments": {"language": "de", "country": "de", "category": "health"}}, {"name": "news_source", "arguments": {"source_id": "bbc"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"latest\", \"description\": \"Fetches the latest news from Google News for a specified language region.\", \"parameters\": {\"lr\": {\"description\": \"Language region to get news for, e.g., 'en-US'.\", \"type\": \"str\", \"default\": \"en-US\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"top_headlines\", \"description\": \"Fetches top headline news articles based on specified parameters.\", \"parameters\": {\"language\": {\"description\": \"Language code to filter news articles. Default is 'en'. Options include:\", \"type\": \"str\", \"default\": \"en\"}, \"category\": {\"description\": \"Category of news articles to filter by. Default is 'sports'. Options include:\", \"type\": \"str\", \"default\": \"sports\"}, \"country\": {\"description\": \"Country code to filter news articles. Default is 'us'. Options include:\", \"type\": \"str\", \"default\": \"us\"}, \"sortby\": {\"description\": \"Criterion to sort news articles by. Default is 'published_date'.\", \"type\": \"str\", \"default\": \"\"}, \"pagesize\": {\"description\": \"Number of news articles to return per page. Default is 10.\", \"type\": \"int\", \"default\": \"10\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"news_source\", \"description\": \"Fetch the latest news stories for a given news source in chronological order.\", \"parameters\": {\"source_id\": {\"description\": \"The identifier of the news source for which to retrieve the latest news stories.\", \"type\": \"str\", \"default\": \"bbc\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Please get me the top health news in German from Germany, and also the latest news from the BBC source.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"top_headlines\", \"arguments\": {\"language\": \"de\", \"country\": \"de\", \"category\": \"health\"}}, {\"name\": \"news_source\", \"arguments\": {\"source_id\": \"bbc\"}}]",
"style": "rule"
}
|
{
"index": 5,
"split": "train"
}
|
[{"name": "get_movies_by_cast_name", "description": "Fetches movies associated with a given cast member's name from the IMDB API.", "parameters": {"cast_name": {"description": "Name of the cast member to search for. Defaults to None.", "type": "str, optional", "default": ""}}}, {"name": "get_movies_by_director", "description": "Gets a list of movies by a specified director using the IMDb API.", "parameters": {"movie_director": {"description": "The name of the director. If not provided, an empty query will be sent.", "type": "str, optional", "default": ""}}}, {"name": "season_episodes", "description": "Fetch episodes for a given season from the Netflix RapidAPI.", "parameters": {"ids": {"description": "Season IDs (separated by commas) for which episodes are to be fetched.", "type": "str", "default": "80077209,80117715"}, "lang": {"description": "Language code for the episodes. Defaults to 'en'.", "type": "str, optional", "default": "en"}, "limit": {"description": "The maximum number of episodes to return. Defaults to 25.", "type": "int, optional", "default": "25"}, "offset": {"description": "The offset for pagination. Defaults to 0.", "type": "int, optional", "default": "0"}}}, {"name": "actors_random", "description": "Fetches a random list of actors from the Movies Database RapidAPI.", "parameters": {"limit": {"description": "The number of random actors to retrieve. Defaults to None, and the API will use its default value if not specified.", "type": "int, optional", "default": ""}}}]
|
[
{
"from": "human",
"value": "I'm trying to find all the movies directed by Steven Spielberg and the ones starring Tom Hanks. Can you also get me the episodes of the first season of Breaking Bad? And while you're at it, fetch me a list of 5 random actors."
}
] |
[{"name": "get_movies_by_director", "arguments": {"movie_director": "Steven Spielberg"}}, {"name": "get_movies_by_cast_name", "arguments": {"cast_name": "Tom Hanks"}}, {"name": "season_episodes", "arguments": {"ids": "1"}}, {"name": "actors_random", "arguments": {"limit": 5}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_movies_by_cast_name\", \"description\": \"Fetches movies associated with a given cast member's name from the IMDB API.\", \"parameters\": {\"cast_name\": {\"description\": \"Name of the cast member to search for. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"get_movies_by_director\", \"description\": \"Gets a list of movies by a specified director using the IMDb API.\", \"parameters\": {\"movie_director\": {\"description\": \"The name of the director. If not provided, an empty query will be sent.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"season_episodes\", \"description\": \"Fetch episodes for a given season from the Netflix RapidAPI.\", \"parameters\": {\"ids\": {\"description\": \"Season IDs (separated by commas) for which episodes are to be fetched.\", \"type\": \"str\", \"default\": \"80077209,80117715\"}, \"lang\": {\"description\": \"Language code for the episodes. Defaults to 'en'.\", \"type\": \"str, optional\", \"default\": \"en\"}, \"limit\": {\"description\": \"The maximum number of episodes to return. Defaults to 25.\", \"type\": \"int, optional\", \"default\": \"25\"}, \"offset\": {\"description\": \"The offset for pagination. Defaults to 0.\", \"type\": \"int, optional\", \"default\": \"0\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"actors_random\", \"description\": \"Fetches a random list of actors from the Movies Database RapidAPI.\", \"parameters\": {\"limit\": {\"description\": \"The number of random actors to retrieve. Defaults to None, and the API will use its default value if not specified.\", \"type\": \"int, optional\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "I'm trying to find all the movies directed by Steven Spielberg and the ones starring Tom Hanks. Can you also get me the episodes of the first season of Breaking Bad? And while you're at it, fetch me a list of 5 random actors.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"get_movies_by_director\", \"arguments\": {\"movie_director\": \"Steven Spielberg\"}}, {\"name\": \"get_movies_by_cast_name\", \"arguments\": {\"cast_name\": \"Tom Hanks\"}}, {\"name\": \"season_episodes\", \"arguments\": {\"ids\": \"1\"}}, {\"name\": \"actors_random\", \"arguments\": {\"limit\": 5}}]",
"style": "rule"
}
|
{
"index": 6,
"split": "train"
}
|
[{"name": "get_3_summary", "description": "Fetches a summary of stock performance and related metrics from the Stocky API.", "parameters": {"keyword": {"description": "The stock keyword to search for. Default is 'GOOG'.", "type": "str", "default": "GOOG"}}}, {"name": "get_liquidity_value", "description": "Retrieves the liquidity value of a specified cryptocurrency in US Dollars.", "parameters": {"symbol": {"description": "The ticker symbol of the cryptocurrency.", "type": "str", "default": "BTC"}}}, {"name": "market_trends", "description": "Fetches market trends data from the Joj Finance API.", "parameters": {"t": {"description": "Tab name, specifying the category of market trends. One of the following:", "type": "str", "default": "indexes"}, "gl": {"description": "Geolocation for the data, default is 'US'.", "type": "str, optional", "default": "US"}, "s": {"description": "Subtab name for 'indexes' tab to specify the region. One of the following:", "type": "str, optional", "default": ""}, "hl": {"description": "Language code, default is 'en'.", "type": "str, optional", "default": "en"}}}]
|
[
{
"from": "human",
"value": "I need the liquidity value of Stellar (XLM) and the market trends for indexes in the Americas. Also, could you give me the summary of the stock performance for Facebook (FB)?"
}
] |
[{"name": "get_liquidity_value", "arguments": {"symbol": "XLM"}}, {"name": "market_trends", "arguments": {"t": "indexes", "s": "americas"}}, {"name": "get_3_summary", "arguments": {"keyword": "FB"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_3_summary\", \"description\": \"Fetches a summary of stock performance and related metrics from the Stocky API.\", \"parameters\": {\"keyword\": {\"description\": \"The stock keyword to search for. Default is 'GOOG'.\", \"type\": \"str\", \"default\": \"GOOG\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"get_liquidity_value\", \"description\": \"Retrieves the liquidity value of a specified cryptocurrency in US Dollars.\", \"parameters\": {\"symbol\": {\"description\": \"The ticker symbol of the cryptocurrency.\", \"type\": \"str\", \"default\": \"BTC\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"market_trends\", \"description\": \"Fetches market trends data from the Joj Finance API.\", \"parameters\": {\"t\": {\"description\": \"Tab name, specifying the category of market trends. One of the following:\", \"type\": \"str\", \"default\": \"indexes\"}, \"gl\": {\"description\": \"Geolocation for the data, default is 'US'.\", \"type\": \"str, optional\", \"default\": \"US\"}, \"s\": {\"description\": \"Subtab name for 'indexes' tab to specify the region. One of the following:\", \"type\": \"str, optional\", \"default\": \"\"}, \"hl\": {\"description\": \"Language code, default is 'en'.\", \"type\": \"str, optional\", \"default\": \"en\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "I need the liquidity value of Stellar (XLM) and the market trends for indexes in the Americas. Also, could you give me the summary of the stock performance for Facebook (FB)?",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"get_liquidity_value\", \"arguments\": {\"symbol\": \"XLM\"}}, {\"name\": \"market_trends\", \"arguments\": {\"t\": \"indexes\", \"s\": \"americas\"}}, {\"name\": \"get_3_summary\", \"arguments\": {\"keyword\": \"FB\"}}]",
"style": "rule"
}
|
{
"index": 7,
"split": "train"
}
|
[{"name": "aerospace_material_optimization", "description": "Optimize material usage for aerospace 3D printing projects based on specific requirements.", "parameters": {"type": "dict", "properties": {"project_id": {"type": "string", "description": "Unique identifier for the 3D printing project."}, "material_requirements": {"type": "array", "description": "List of material specifications needed for the project.", "items": {"type": "dict", "properties": {"material_type": {"type": "string", "description": "Type of material required (e.g., titanium, aluminum)."}, "quantity": {"type": "float", "description": "Required quantity in kilograms."}, "properties": {"type": "dict", "properties": {"strength": {"type": "int", "description": "Required strength measured in MPa."}, "temperature_resistance": {"type": "int", "description": "Maximum temperature resistance required (in degrees Celsius)."}}, "required": ["strength", "temperature_resistance"]}}, "required": ["material_type", "quantity"]}}, "optimization_time": {"type": "string", "enum": ["Immediate", "End of Day", "Within 24 Hours"], "description": "Preferred time for optimization results.", "default": "Immediate"}}, "required": ["project_id", "material_requirements"]}, "required": null}]
|
[
{
"from": "human",
"value": "I need an optimization for three aerospace 3D printing projects. The details are as follows:\n1. Project ID: A123, Materials: Titanium (50 kg, strength: 900 MPa, temperature resistance: 500°C), Aluminum (30 kg, strength: 400 MPa, temperature resistance: 300°C)\n2. Project ID: B456, Materials: Carbon Fiber (20 kg, strength: 1500 MPa, temperature resistance: 200°C)\n3. Project ID: C789, Materials: Titanium (40 kg, strength: 850 MPa, temperature resistance: 450°C), Steel (25 kg, strength: 1200 MPa, temperature resistance: 700°C)\nAll optimizations should be immediate."
}
] |
[{"name": "aerospace_material_optimization", "arguments": {"project_id": "A123", "material_requirements": [{"material_type": "Titanium", "quantity": 50.0, "properties": {"strength": 900, "temperature_resistance": 500}}, {"material_type": "Aluminum", "quantity": 30.0, "properties": {"strength": 400, "temperature_resistance": 300}}], "optimization_time": "Immediate"}}, {"name": "aerospace_material_optimization", "arguments": {"project_id": "B456", "material_requirements": [{"material_type": "Carbon Fiber", "quantity": 20.0, "properties": {"strength": 1500, "temperature_resistance": 200}}], "optimization_time": "Immediate"}}, {"name": "aerospace_material_optimization", "arguments": {"project_id": "C789", "material_requirements": [{"material_type": "Titanium", "quantity": 40.0, "properties": {"strength": 850, "temperature_resistance": 450}}, {"material_type": "Steel", "quantity": 25.0, "properties": {"strength": 1200, "temperature_resistance": 700}}], "optimization_time": "Immediate"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out.
The current time is 2026-12-22 09:05:35.Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out.\nThe current time is 2026-12-22 09:05:35.Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"aerospace_material_optimization\", \"description\": \"Optimize material usage for aerospace 3D printing projects based on specific requirements.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"project_id\": {\"type\": \"string\", \"description\": \"Unique identifier for the 3D printing project.\"}, \"material_requirements\": {\"type\": \"array\", \"description\": \"List of material specifications needed for the project.\", \"items\": {\"type\": \"dict\", \"properties\": {\"material_type\": {\"type\": \"string\", \"description\": \"Type of material required (e.g., titanium, aluminum).\"}, \"quantity\": {\"type\": \"float\", \"description\": \"Required quantity in kilograms.\"}, \"properties\": {\"type\": \"dict\", \"properties\": {\"strength\": {\"type\": \"int\", \"description\": \"Required strength measured in MPa.\"}, \"temperature_resistance\": {\"type\": \"int\", \"description\": \"Maximum temperature resistance required (in degrees Celsius).\"}}, \"required\": [\"strength\", \"temperature_resistance\"]}}, \"required\": [\"material_type\", \"quantity\"]}}, \"optimization_time\": {\"type\": \"string\", \"enum\": [\"Immediate\", \"End of Day\", \"Within 24 Hours\"], \"description\": \"Preferred time for optimization results.\", \"default\": \"Immediate\"}}, \"required\": [\"project_id\", \"material_requirements\"]}, \"required\": null}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "I need an optimization for three aerospace 3D printing projects. The details are as follows:\n1. Project ID: A123, Materials: Titanium (50 kg, strength: 900 MPa, temperature resistance: 500°C), Aluminum (30 kg, strength: 400 MPa, temperature resistance: 300°C)\n2. Project ID: B456, Materials: Carbon Fiber (20 kg, strength: 1500 MPa, temperature resistance: 200°C)\n3. Project ID: C789, Materials: Titanium (40 kg, strength: 850 MPa, temperature resistance: 450°C), Steel (25 kg, strength: 1200 MPa, temperature resistance: 700°C)\nAll optimizations should be immediate.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"aerospace_material_optimization\", \"arguments\": {\"project_id\": \"A123\", \"material_requirements\": [{\"material_type\": \"Titanium\", \"quantity\": 50.0, \"properties\": {\"strength\": 900, \"temperature_resistance\": 500}}, {\"material_type\": \"Aluminum\", \"quantity\": 30.0, \"properties\": {\"strength\": 400, \"temperature_resistance\": 300}}], \"optimization_time\": \"Immediate\"}}, {\"name\": \"aerospace_material_optimization\", \"arguments\": {\"project_id\": \"B456\", \"material_requirements\": [{\"material_type\": \"Carbon Fiber\", \"quantity\": 20.0, \"properties\": {\"strength\": 1500, \"temperature_resistance\": 200}}], \"optimization_time\": \"Immediate\"}}, {\"name\": \"aerospace_material_optimization\", \"arguments\": {\"project_id\": \"C789\", \"material_requirements\": [{\"material_type\": \"Titanium\", \"quantity\": 40.0, \"properties\": {\"strength\": 850, \"temperature_resistance\": 450}}, {\"material_type\": \"Steel\", \"quantity\": 25.0, \"properties\": {\"strength\": 1200, \"temperature_resistance\": 700}}], \"optimization_time\": \"Immediate\"}}]",
"style": "rule"
}
|
{
"index": 8,
"split": "train"
}
|
[{"name": "find_minimum_rotated_sorted_array", "description": "Finds the minimum element in a rotated sorted array.", "parameters": {"nums": {"description": "The rotated sorted array.", "type": "List[int]"}}}, {"name": "average", "description": "Calculates the arithmetic mean of a list of numbers.", "parameters": {"numbers": {"description": "The list of numbers.", "type": "List[float]"}}}, {"name": "greatest_common_divisor", "description": "Computes the greatest common divisor (GCD) of two non-negative integers.", "parameters": {"a": {"description": "The first non-negative integer.", "type": "int"}, "b": {"description": "The second non-negative integer.", "type": "int"}}}]
|
[
{
"from": "human",
"value": "Determine the greatest common divisor of 105 and 24."
}
] |
[{"name": "greatest_common_divisor", "arguments": {"a": 105, "b": 24}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"find_minimum_rotated_sorted_array\", \"description\": \"Finds the minimum element in a rotated sorted array.\", \"parameters\": {\"nums\": {\"description\": \"The rotated sorted array.\", \"type\": \"List[int]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"average\", \"description\": \"Calculates the arithmetic mean of a list of numbers.\", \"parameters\": {\"numbers\": {\"description\": \"The list of numbers.\", \"type\": \"List[float]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"greatest_common_divisor\", \"description\": \"Computes the greatest common divisor (GCD) of two non-negative integers.\", \"parameters\": {\"a\": {\"description\": \"The first non-negative integer.\", \"type\": \"int\"}, \"b\": {\"description\": \"The second non-negative integer.\", \"type\": \"int\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Determine the greatest common divisor of 105 and 24.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"greatest_common_divisor\", \"arguments\": {\"a\": 105, \"b\": 24}}]",
"style": "rule"
}
|
{
"index": 9,
"split": "train"
}
|
[{"name": "List User Followings", "description": "Retrieve a list of users that a given user is following on TikTok", "parameters": {"type": "dict", "properties": {"user_id": {"description": "The ID of the user to retrieve followings for", "type": "string"}, "sec_uid": {"description": "The secure user ID of the user to retrieve followings for", "type": "string"}, "proxy": {"description": "Optional proxy URL for the request", "type": "string", "default": ""}, "offset": {"description": "Optional offset for pagination", "type": "float", "default": 0}, "count": {"description": "Optional count for pagination", "type": "float", "default": 10}}, "required": ["user_id", "sec_uid"]}, "required": null}, {"name": "Twitter Auto Complete API", "description": "Returns autocomplete results for Twitter search queries.", "parameters": {"type": "dict", "properties": {"query": {"description": "The search query to autocomplete", "type": "string"}}, "required": ["query"]}, "required": null}, {"name": "Get User Tweets and Replies by REST ID", "description": "Retrieves a user's tweets and replies by providing a REST ID.", "parameters": {"type": "dict", "properties": {"user_id": {"description": "The ID of the user whose tweets and replies are to be retrieved.", "type": "string"}, "limit": {"description": "The maximum number of tweets and replies to return.", "type": "string", "default": "40"}, "cursor": {"description": "The cursor value to use for pagination.", "type": "string", "default": ""}}, "required": ["user_id"]}, "required": null}, {"name": "Search Followers by PK and Query", "description": "Search for users who are followers of a specific Instagram account, given the account's public key (pk) and a search query.", "parameters": {"type": "dict", "properties": {"pk": {"description": "Public key of the Instagram account", "type": "float"}, "query": {"description": "Search query to filter followers", "type": "string"}}, "required": ["pk", "query"]}, "required": null}, {"name": "Performer Model Information", "description": "Retrieves detailed information about a performer or model on the onlyfans platform.", "parameters": {"type": "dict", "properties": {"username": {"description": "The username or ID of the performer or model.", "type": "string"}}, "required": ["username"]}, "required": null}, {"name": "Twitter Space Stream URL API", "description": "This API receives the m3u8 URL for a Twitter Space stream by providing the mediaId.", "parameters": {"type": "dict", "properties": {"mediaId": {"description": "The mediaId obtained from the Twitter Space details and media_key value.", "type": "string"}}, "required": ["mediaId"]}, "required": null}]
|
[
{
"from": "human",
"value": "Role definition:\n Inquirer: A user who raises an inquiry.\n Response assistant: Communicates with the inquirer and provides answers and solutions.\n\n Historical dialog data is as follows:\nInquirer: I want to listen to the upcoming Twitter Space on AI advancements, but I'm struggling with finding the stream URL. Can you help me with that?\nResponse assistant: Can you provide the mediaId for the Twitter Space you mentioned?\nInquirer: Yes, the mediaId is \"12345\"\n\n Please continue your answer given the historical dialog. Your role is the Response assistant."
}
] |
[{"name": "Twitter Space Stream URL API", "arguments": {"mediaId": "12345"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out.
The current time is 2023-10-19 17:34:44.Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out.\nThe current time is 2023-10-19 17:34:44.Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"List User Followings\", \"description\": \"Retrieve a list of users that a given user is following on TikTok\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"user_id\": {\"description\": \"The ID of the user to retrieve followings for\", \"type\": \"string\"}, \"sec_uid\": {\"description\": \"The secure user ID of the user to retrieve followings for\", \"type\": \"string\"}, \"proxy\": {\"description\": \"Optional proxy URL for the request\", \"type\": \"string\", \"default\": \"\"}, \"offset\": {\"description\": \"Optional offset for pagination\", \"type\": \"float\", \"default\": 0}, \"count\": {\"description\": \"Optional count for pagination\", \"type\": \"float\", \"default\": 10}}, \"required\": [\"user_id\", \"sec_uid\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Twitter Auto Complete API\", \"description\": \"Returns autocomplete results for Twitter search queries.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"query\": {\"description\": \"The search query to autocomplete\", \"type\": \"string\"}}, \"required\": [\"query\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Get User Tweets and Replies by REST ID\", \"description\": \"Retrieves a user's tweets and replies by providing a REST ID.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"user_id\": {\"description\": \"The ID of the user whose tweets and replies are to be retrieved.\", \"type\": \"string\"}, \"limit\": {\"description\": \"The maximum number of tweets and replies to return.\", \"type\": \"string\", \"default\": \"40\"}, \"cursor\": {\"description\": \"The cursor value to use for pagination.\", \"type\": \"string\", \"default\": \"\"}}, \"required\": [\"user_id\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Search Followers by PK and Query\", \"description\": \"Search for users who are followers of a specific Instagram account, given the account's public key (pk) and a search query.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"pk\": {\"description\": \"Public key of the Instagram account\", \"type\": \"float\"}, \"query\": {\"description\": \"Search query to filter followers\", \"type\": \"string\"}}, \"required\": [\"pk\", \"query\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Performer Model Information\", \"description\": \"Retrieves detailed information about a performer or model on the onlyfans platform.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"username\": {\"description\": \"The username or ID of the performer or model.\", \"type\": \"string\"}}, \"required\": [\"username\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Twitter Space Stream URL API\", \"description\": \"This API receives the m3u8 URL for a Twitter Space stream by providing the mediaId.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"mediaId\": {\"description\": \"The mediaId obtained from the Twitter Space details and media_key value.\", \"type\": \"string\"}}, \"required\": [\"mediaId\"]}, \"required\": null}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Role definition:\n Inquirer: A user who raises an inquiry.\n Response assistant: Communicates with the inquirer and provides answers and solutions.\n\n Historical dialog data is as follows:\nInquirer: I want to listen to the upcoming Twitter Space on AI advancements, but I'm struggling with finding the stream URL. Can you help me with that?\nResponse assistant: Can you provide the mediaId for the Twitter Space you mentioned?\nInquirer: Yes, the mediaId is \"12345\"\n\n Please continue your answer given the historical dialog. Your role is the Response assistant.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"Twitter Space Stream URL API\", \"arguments\": {\"mediaId\": \"12345\"}}]",
"style": "rule"
}
|
{
"index": 10,
"split": "train"
}
|
[{"name": "greatest_common_divisor", "description": "Computes the greatest common divisor (GCD) of two non-negative integers.", "parameters": {"a": {"description": "The first non-negative integer.", "type": "int"}, "b": {"description": "The second non-negative integer.", "type": "int"}}}, {"name": "linear_regression_prediction", "description": "Predicts the y-value for a given x-value using linear regression.", "parameters": {"x_values": {"description": "The x-coordinates of the data points.", "type": "List[float]"}, "y_values": {"description": "The y-coordinates of the data points.", "type": "List[float]"}, "target_x": {"description": "The x-value for which to predict the y-value.", "type": "float"}}}, {"name": "count_bits", "description": "Counts the number of set bits (1's) in the binary representation of a number.", "parameters": {"num": {"description": "The input number.", "type": "int"}}}, {"name": "merge_sorted_lists", "description": "Merges two sorted lists into a single sorted list.", "parameters": {"list1": {"description": "The first sorted list.", "type": "List[int]"}, "list2": {"description": "The second sorted list.", "type": "List[int]"}}}, {"name": "investment_profit", "description": "Calculates the profit from an investment based on the initial amount, annual return rate, and time.", "parameters": {"amount": {"description": "The initial investment amount.", "type": "float"}, "rate": {"description": "The annual return rate expressed as a decimal.", "type": "float"}, "years": {"description": "The number of years the investment is held.", "type": "int"}}}]
|
[
{
"from": "human",
"value": "Merge two sorted lists: [1, 3, 5] and [2, 4, 6] into a single sorted list."
}
] |
[{"name": "merge_sorted_lists", "arguments": {"list1": [1, 3, 5], "list2": [2, 4, 6]}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"greatest_common_divisor\", \"description\": \"Computes the greatest common divisor (GCD) of two non-negative integers.\", \"parameters\": {\"a\": {\"description\": \"The first non-negative integer.\", \"type\": \"int\"}, \"b\": {\"description\": \"The second non-negative integer.\", \"type\": \"int\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"linear_regression_prediction\", \"description\": \"Predicts the y-value for a given x-value using linear regression.\", \"parameters\": {\"x_values\": {\"description\": \"The x-coordinates of the data points.\", \"type\": \"List[float]\"}, \"y_values\": {\"description\": \"The y-coordinates of the data points.\", \"type\": \"List[float]\"}, \"target_x\": {\"description\": \"The x-value for which to predict the y-value.\", \"type\": \"float\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"count_bits\", \"description\": \"Counts the number of set bits (1's) in the binary representation of a number.\", \"parameters\": {\"num\": {\"description\": \"The input number.\", \"type\": \"int\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"merge_sorted_lists\", \"description\": \"Merges two sorted lists into a single sorted list.\", \"parameters\": {\"list1\": {\"description\": \"The first sorted list.\", \"type\": \"List[int]\"}, \"list2\": {\"description\": \"The second sorted list.\", \"type\": \"List[int]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"investment_profit\", \"description\": \"Calculates the profit from an investment based on the initial amount, annual return rate, and time.\", \"parameters\": {\"amount\": {\"description\": \"The initial investment amount.\", \"type\": \"float\"}, \"rate\": {\"description\": \"The annual return rate expressed as a decimal.\", \"type\": \"float\"}, \"years\": {\"description\": \"The number of years the investment is held.\", \"type\": \"int\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Merge two sorted lists: [1, 3, 5] and [2, 4, 6] into a single sorted list.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"merge_sorted_lists\", \"arguments\": {\"list1\": [1, 3, 5], \"list2\": [2, 4, 6]}}]",
"style": "rule"
}
|
{
"index": 11,
"split": "train"
}
|
[{"name": "generate_meme", "description": "Generate a meme with a specified image and text", "parameters": {"type": "dict", "properties": {"image": {"type": "string", "description": "The URL or path of the image to be used"}, "text": {"type": "string", "description": "The text to be added to the meme"}}, "required": ["image", "text"]}, "required": null}]
|
[
{
"from": "human",
"value": "Can you create two memes for me? Use this image (https://example.com/image1.jpg) for the first meme with the text \"When you realize it's Friday!\", and this image (https://example.com/image2.jpg) for the second meme with the text \"Monday mornings be like...\"."
}
] |
[{"name": "generate_meme", "arguments": {"image": "https://example.com/image1.jpg", "text": "When you realize it's Friday!"}}, {"name": "generate_meme", "arguments": {"image": "https://example.com/image2.jpg", "text": "Monday mornings be like..."}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out.
The current time is 2020-12-24 19:32:06.Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out.\nThe current time is 2020-12-24 19:32:06.Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"generate_meme\", \"description\": \"Generate a meme with a specified image and text\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"image\": {\"type\": \"string\", \"description\": \"The URL or path of the image to be used\"}, \"text\": {\"type\": \"string\", \"description\": \"The text to be added to the meme\"}}, \"required\": [\"image\", \"text\"]}, \"required\": null}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Can you create two memes for me? Use this image (https://example.com/image1.jpg) for the first meme with the text \"When you realize it's Friday!\", and this image (https://example.com/image2.jpg) for the second meme with the text \"Monday mornings be like...\".",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"generate_meme\", \"arguments\": {\"image\": \"https://example.com/image1.jpg\", \"text\": \"When you realize it's Friday!\"}}, {\"name\": \"generate_meme\", \"arguments\": {\"image\": \"https://example.com/image2.jpg\", \"text\": \"Monday mornings be like...\"}}]",
"style": "rule"
}
|
{
"index": 12,
"split": "train"
}
|
[{"name": "is_prime", "description": "Checks if a number is prime.", "parameters": {"num": {"description": "The number to be checked.", "type": "int"}}}, {"name": "average", "description": "Calculates the arithmetic mean of a list of numbers.", "parameters": {"numbers": {"description": "The list of numbers.", "type": "List[float]"}}}, {"name": "density", "description": "Calculates the density of an object given its mass and volume.", "parameters": {"mass": {"description": "The mass of the object.", "type": "float"}, "volume": {"description": "The volume of the object.", "type": "float"}}}, {"name": "find_longest_palindromic_substring", "description": "Finds the longest palindromic substring in a string.", "parameters": {"s": {"description": "The input string.", "type": "str"}}}]
|
[
{
"from": "human",
"value": "Calculate the average of the numbers [1.5, 2.5, 3.5, 4.5]."
}
] |
[{"name": "average", "arguments": {"numbers": [1.5, 2.5, 3.5, 4.5]}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"is_prime\", \"description\": \"Checks if a number is prime.\", \"parameters\": {\"num\": {\"description\": \"The number to be checked.\", \"type\": \"int\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"average\", \"description\": \"Calculates the arithmetic mean of a list of numbers.\", \"parameters\": {\"numbers\": {\"description\": \"The list of numbers.\", \"type\": \"List[float]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"density\", \"description\": \"Calculates the density of an object given its mass and volume.\", \"parameters\": {\"mass\": {\"description\": \"The mass of the object.\", \"type\": \"float\"}, \"volume\": {\"description\": \"The volume of the object.\", \"type\": \"float\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"find_longest_palindromic_substring\", \"description\": \"Finds the longest palindromic substring in a string.\", \"parameters\": {\"s\": {\"description\": \"The input string.\", \"type\": \"str\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Calculate the average of the numbers [1.5, 2.5, 3.5, 4.5].",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"average\", \"arguments\": {\"numbers\": [1.5, 2.5, 3.5, 4.5]}}]",
"style": "rule"
}
|
{
"index": 13,
"split": "train"
}
|
[{"name": "stock_get_statistics", "description": "Retrieve statistical information for a given stock.", "parameters": {"is_id": {"description": "The ID of the stock. Use the market/auto-complete API to get the correct values.", "type": "str", "default": "aapl:us"}, "template": {"description": "Specifies the type of statistical template to use. Options include 'COMMODITY', 'CURRENCY', 'INDEX', 'INDEXFUTURE', 'RATE', and 'STOCK'. Defaults to 'STOCK'.", "type": "str, optional", "default": "STOCK"}}}, {"name": "account_transaction_history", "description": "Retrieves an array of all transactions associated with a specific address on the Optimism blockchain.", "parameters": {"address": {"description": "The blockchain address to query for transaction history.", "type": "str", "default": "0x0000000000000000000000000000000000000123"}, "network": {"description": "The network to query, either 'mainnet' or 'testnet'. Defaults to None.", "type": "str, optional", "default": ""}}}]
|
[
{
"from": "human",
"value": "I'd like to get the statistical information for the US 10-Year Treasury Bond and the transaction history of an address on the Optimism blockchain's mainnet."
}
] |
[{"name": "stock_get_statistics", "arguments": {"is_id": "DGS10", "template": "RATE"}}, {"name": "account_transaction_history", "arguments": {"address": "0x742d35Cc6634C0532925a3b844Bc454e4438f44e", "network": "mainnet"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"stock_get_statistics\", \"description\": \"Retrieve statistical information for a given stock.\", \"parameters\": {\"is_id\": {\"description\": \"The ID of the stock. Use the market/auto-complete API to get the correct values.\", \"type\": \"str\", \"default\": \"aapl:us\"}, \"template\": {\"description\": \"Specifies the type of statistical template to use. Options include 'COMMODITY', 'CURRENCY', 'INDEX', 'INDEXFUTURE', 'RATE', and 'STOCK'. Defaults to 'STOCK'.\", \"type\": \"str, optional\", \"default\": \"STOCK\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"account_transaction_history\", \"description\": \"Retrieves an array of all transactions associated with a specific address on the Optimism blockchain.\", \"parameters\": {\"address\": {\"description\": \"The blockchain address to query for transaction history.\", \"type\": \"str\", \"default\": \"0x0000000000000000000000000000000000000123\"}, \"network\": {\"description\": \"The network to query, either 'mainnet' or 'testnet'. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "I'd like to get the statistical information for the US 10-Year Treasury Bond and the transaction history of an address on the Optimism blockchain's mainnet.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"stock_get_statistics\", \"arguments\": {\"is_id\": \"DGS10\", \"template\": \"RATE\"}}, {\"name\": \"account_transaction_history\", \"arguments\": {\"address\": \"0x742d35Cc6634C0532925a3b844Bc454e4438f44e\", \"network\": \"mainnet\"}}]",
"style": "rule"
}
|
{
"index": 14,
"split": "train"
}
|
[{"name": "is_valid_parentheses", "description": "Checks if a string contains valid parentheses.", "parameters": {"s": {"description": "The input string.", "type": "str"}}}, {"name": "merge_dictionaries", "description": "Merges two dictionaries into a single dictionary.", "parameters": {"dict1": {"description": "The first dictionary.", "type": "Dict"}, "dict2": {"description": "The second dictionary.", "type": "Dict"}}}]
|
[
{
"from": "human",
"value": "Is the string '()()()' a valid parentheses sequence? Can you merge {'m': 13, 'n': 14} and {'n': 15, 'o': 16}?"
}
] |
[{"name": "is_valid_parentheses", "arguments": {"s": "()()()"}}, {"name": "merge_dictionaries", "arguments": {"dict1": {"m": 13, "n": 14}, "dict2": {"n": 15, "o": 16}}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"is_valid_parentheses\", \"description\": \"Checks if a string contains valid parentheses.\", \"parameters\": {\"s\": {\"description\": \"The input string.\", \"type\": \"str\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"merge_dictionaries\", \"description\": \"Merges two dictionaries into a single dictionary.\", \"parameters\": {\"dict1\": {\"description\": \"The first dictionary.\", \"type\": \"Dict\"}, \"dict2\": {\"description\": \"The second dictionary.\", \"type\": \"Dict\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Is the string '()()()' a valid parentheses sequence? Can you merge {'m': 13, 'n': 14} and {'n': 15, 'o': 16}?",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"is_valid_parentheses\", \"arguments\": {\"s\": \"()()()\"}}, {\"name\": \"merge_dictionaries\", \"arguments\": {\"dict1\": {\"m\": 13, \"n\": 14}, \"dict2\": {\"n\": 15, \"o\": 16}}}]",
"style": "rule"
}
|
{
"index": 15,
"split": "train"
}
|
[{"name": "v1_ownersearch", "description": "Performs a search on the owner name or postcode and returns details about the trademarks found for that owner. Duplicate owner results can be returned by this endpoint, especially when an owner had an address change.", "parameters": {"search_keyword": {"description": "The search keyword (e.g., 'ferr' will return results for 'ferrari'). Defaults to 'netflix'. Use literal 'null' if you don't want to use this parameter.", "type": "str", "default": "netflix"}, "postcode": {"description": "The postcode to filter the search results (e.g., '95066'). Use literal 'null' if you don't want to use this parameter. Note that attorneys might omit postcodes from filings, so those results cannot be returned as they don't have postcode attached as metadata.", "type": "str", "default": ""}}}, {"name": "search", "description": "This function performs a vehicle license plate lookup for a given plate number in the Netherlands using the RapidAPI service.", "parameters": {"plate": {"description": "The vehicle license plate number to look up. Defaults to '3XKF30'.", "type": "str", "default": "3XKF30"}, "function": {"description": "The function to use for the lookup. Defaults to 'getktypefornumplatenetherlands'.", "type": "str", "default": "getktypefornumplatenetherlands"}}}, {"name": "auto_complete", "description": "Fetches auto-complete suggestions for a given search term using the Zoopla API.", "parameters": {"search_term": {"description": "The term or phrase for which to fetch auto-complete suggestions.", "type": "str", "default": "greenwich"}, "search_type": {"description": "The type of search to perform. Acceptable values are 'listings' or 'properties'. Defaults to 'listings'.", "type": "str, optional", "default": "listings"}}}, {"name": "get_data_entity_uid", "description": "Fetches an entity record from the OFAC service using a unique identifier (UID).", "parameters": {"uid": {"description": "The unique identifier of the entity to retrieve.", "type": "int", "default": "7178"}}}]
|
[
{
"from": "human",
"value": "Get auto-complete suggestions for 'London properties' and retrieve data for entity UID 12345."
}
] |
[{"name": "auto_complete", "arguments": {"search_term": "London properties", "search_type": "properties"}}, {"name": "get_data_entity_uid", "arguments": {"uid": 12345}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"v1_ownersearch\", \"description\": \"Performs a search on the owner name or postcode and returns details about the trademarks found for that owner. Duplicate owner results can be returned by this endpoint, especially when an owner had an address change.\", \"parameters\": {\"search_keyword\": {\"description\": \"The search keyword (e.g., 'ferr' will return results for 'ferrari'). Defaults to 'netflix'. Use literal 'null' if you don't want to use this parameter.\", \"type\": \"str\", \"default\": \"netflix\"}, \"postcode\": {\"description\": \"The postcode to filter the search results (e.g., '95066'). Use literal 'null' if you don't want to use this parameter. Note that attorneys might omit postcodes from filings, so those results cannot be returned as they don't have postcode attached as metadata.\", \"type\": \"str\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"search\", \"description\": \"This function performs a vehicle license plate lookup for a given plate number in the Netherlands using the RapidAPI service.\", \"parameters\": {\"plate\": {\"description\": \"The vehicle license plate number to look up. Defaults to '3XKF30'.\", \"type\": \"str\", \"default\": \"3XKF30\"}, \"function\": {\"description\": \"The function to use for the lookup. Defaults to 'getktypefornumplatenetherlands'.\", \"type\": \"str\", \"default\": \"getktypefornumplatenetherlands\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"auto_complete\", \"description\": \"Fetches auto-complete suggestions for a given search term using the Zoopla API.\", \"parameters\": {\"search_term\": {\"description\": \"The term or phrase for which to fetch auto-complete suggestions.\", \"type\": \"str\", \"default\": \"greenwich\"}, \"search_type\": {\"description\": \"The type of search to perform. Acceptable values are 'listings' or 'properties'. Defaults to 'listings'.\", \"type\": \"str, optional\", \"default\": \"listings\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"get_data_entity_uid\", \"description\": \"Fetches an entity record from the OFAC service using a unique identifier (UID).\", \"parameters\": {\"uid\": {\"description\": \"The unique identifier of the entity to retrieve.\", \"type\": \"int\", \"default\": \"7178\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Get auto-complete suggestions for 'London properties' and retrieve data for entity UID 12345.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"auto_complete\", \"arguments\": {\"search_term\": \"London properties\", \"search_type\": \"properties\"}}, {\"name\": \"get_data_entity_uid\", \"arguments\": {\"uid\": 12345}}]",
"style": "rule"
}
|
{
"index": 16,
"split": "train"
}
|
[{"name": "get_6_month_historical_daily_prices", "description": "Retrieves 6 months of historical daily stock prices for the given ticker, including adjusted Open, High, Low, and Close prices, as well as Daily Volume, Dividend, and Split information.", "parameters": {"ticker": {"description": "The stock ticker symbol for which to retrieve the historical prices.", "type": "str", "default": "AAPL"}}}]
|
[
{
"from": "human",
"value": "Provide the daily price history for Pfizer (PFE) from 6 months ago."
}
] |
[{"name": "get_6_month_historical_daily_prices", "arguments": {"ticker": "PFE"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_6_month_historical_daily_prices\", \"description\": \"Retrieves 6 months of historical daily stock prices for the given ticker, including adjusted Open, High, Low, and Close prices, as well as Daily Volume, Dividend, and Split information.\", \"parameters\": {\"ticker\": {\"description\": \"The stock ticker symbol for which to retrieve the historical prices.\", \"type\": \"str\", \"default\": \"AAPL\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Provide the daily price history for Pfizer (PFE) from 6 months ago.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"get_6_month_historical_daily_prices\", \"arguments\": {\"ticker\": \"PFE\"}}]",
"style": "rule"
}
|
{
"index": 17,
"split": "train"
}
|
[{"name": "bacterial_growth", "description": "Calculates the bacterial population after a given time based on the initial population and growth rate.", "parameters": {"initial_population": {"description": "The initial bacterial population.", "type": "int", "default": 20}, "growth_rate": {"description": "The growth rate per unit time.", "type": "float", "default": 20}, "time": {"description": "The time elapsed.", "type": "float"}, "doubling_time": {"description": "The doubling time of the bacteria in minutes. Defaults to 20.", "type": "float, optional"}}}]
|
[
{
"from": "human",
"value": "What will be the bacterial population after 60 minutes if the initial population is 100, the growth rate is 0.02 per minute, and the doubling time is 20 minutes? Also, what will be the population after 45 minutes with the same initial population and growth rate but a doubling time of 30 minutes?"
}
] |
[{"name": "bacterial_growth", "arguments": {"initial_population": 100, "growth_rate": 0.02, "time": 60, "doubling_time": 20}}, {"name": "bacterial_growth", "arguments": {"initial_population": 100, "growth_rate": 0.02, "time": 45, "doubling_time": 30}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"bacterial_growth\", \"description\": \"Calculates the bacterial population after a given time based on the initial population and growth rate.\", \"parameters\": {\"initial_population\": {\"description\": \"The initial bacterial population.\", \"type\": \"int\", \"default\": 20}, \"growth_rate\": {\"description\": \"The growth rate per unit time.\", \"type\": \"float\", \"default\": 20}, \"time\": {\"description\": \"The time elapsed.\", \"type\": \"float\"}, \"doubling_time\": {\"description\": \"The doubling time of the bacteria in minutes. Defaults to 20.\", \"type\": \"float, optional\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "What will be the bacterial population after 60 minutes if the initial population is 100, the growth rate is 0.02 per minute, and the doubling time is 20 minutes? Also, what will be the population after 45 minutes with the same initial population and growth rate but a doubling time of 30 minutes?",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"bacterial_growth\", \"arguments\": {\"initial_population\": 100, \"growth_rate\": 0.02, \"time\": 60, \"doubling_time\": 20}}, {\"name\": \"bacterial_growth\", \"arguments\": {\"initial_population\": 100, \"growth_rate\": 0.02, \"time\": 45, \"doubling_time\": 30}}]",
"style": "rule"
}
|
{
"index": 18,
"split": "train"
}
|
[{"name": "get_company_jobs_count", "description": "Retrieves the number of job openings a specified company has posted on LinkedIn.", "parameters": {"company_id": {"description": "The unique identifier of the company whose job openings count is to be retrieved.", "type": "str", "default": "162479"}}}]
|
[
{
"from": "human",
"value": "Hey, can you check how many jobs are currently open at the company I work for? Its ID is 33333"
}
] |
[{"name": "get_company_jobs_count", "arguments": {"company_id": "33333"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_company_jobs_count\", \"description\": \"Retrieves the number of job openings a specified company has posted on LinkedIn.\", \"parameters\": {\"company_id\": {\"description\": \"The unique identifier of the company whose job openings count is to be retrieved.\", \"type\": \"str\", \"default\": \"162479\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Hey, can you check how many jobs are currently open at the company I work for? Its ID is 33333",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"get_company_jobs_count\", \"arguments\": {\"company_id\": \"33333\"}}]",
"style": "rule"
}
|
{
"index": 19,
"split": "train"
}
|
[{"name": "mean_confidence_interval", "description": "Calculates the confidence interval for a sample mean.", "parameters": {"std_dev": {"description": "The standard deviation of the sample.", "type": "float", "default": 0.95}, "sample_size": {"description": "The size of the sample.", "type": "int", "default": 0.95}, "mean": {"description": "The mean of the sample.", "type": "float"}, "confidence": {"description": "The desired confidence level. Default is 0.95.", "type": "float, optional"}}}, {"name": "calculate_standard_deviation", "description": "Calculates the standard deviation of a list of numbers.", "parameters": {"numbers": {"description": "The list of numbers.", "type": "List[Union[int, float]]"}}}, {"name": "greatest_common_divisor", "description": "Computes the greatest common divisor (GCD) of two non-negative integers.", "parameters": {"a": {"description": "The first non-negative integer.", "type": "int"}, "b": {"description": "The second non-negative integer.", "type": "int"}}}, {"name": "euclidean_distance", "description": "Computes the Euclidean distance between two points in a 2D plane.", "parameters": {"point_a": {"description": "The coordinates of the first point (x, y).", "type": "Tuple[float, float]"}, "point_b": {"description": "The coordinates of the second point (x, y).", "type": "Tuple[float, float]"}}}]
|
[
{
"from": "human",
"value": "What is the confidence interval for a sample with a standard deviation of 5.2, a sample size of 100, and a mean of 23.7?"
}
] |
[{"name": "mean_confidence_interval", "arguments": {"std_dev": 5.2, "sample_size": 100, "mean": 23.7}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"mean_confidence_interval\", \"description\": \"Calculates the confidence interval for a sample mean.\", \"parameters\": {\"std_dev\": {\"description\": \"The standard deviation of the sample.\", \"type\": \"float\", \"default\": 0.95}, \"sample_size\": {\"description\": \"The size of the sample.\", \"type\": \"int\", \"default\": 0.95}, \"mean\": {\"description\": \"The mean of the sample.\", \"type\": \"float\"}, \"confidence\": {\"description\": \"The desired confidence level. Default is 0.95.\", \"type\": \"float, optional\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"calculate_standard_deviation\", \"description\": \"Calculates the standard deviation of a list of numbers.\", \"parameters\": {\"numbers\": {\"description\": \"The list of numbers.\", \"type\": \"List[Union[int, float]]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"greatest_common_divisor\", \"description\": \"Computes the greatest common divisor (GCD) of two non-negative integers.\", \"parameters\": {\"a\": {\"description\": \"The first non-negative integer.\", \"type\": \"int\"}, \"b\": {\"description\": \"The second non-negative integer.\", \"type\": \"int\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"euclidean_distance\", \"description\": \"Computes the Euclidean distance between two points in a 2D plane.\", \"parameters\": {\"point_a\": {\"description\": \"The coordinates of the first point (x, y).\", \"type\": \"Tuple[float, float]\"}, \"point_b\": {\"description\": \"The coordinates of the second point (x, y).\", \"type\": \"Tuple[float, float]\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "What is the confidence interval for a sample with a standard deviation of 5.2, a sample size of 100, and a mean of 23.7?",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"mean_confidence_interval\", \"arguments\": {\"std_dev\": 5.2, \"sample_size\": 100, \"mean\": 23.7}}]",
"style": "rule"
}
|
{
"index": 20,
"split": "train"
}
|
[{"name": "fetch_movies", "description": "Fetch a list of movies from 2000 to 2019 based on the selected genre and year using the specified RapidAPI key.", "parameters": {"year": {"description": "The year for which to fetch movies. Should be between 2000 and 2019.", "type": "str", "default": "2018"}, "genre": {"description": "The genre of the movies to fetch.", "type": "str", "default": "Comedy"}}}, {"name": "ott_providers", "description": "Fetch information on OTT platforms supported in a given region.", "parameters": {"region": {"description": "The region to get OTT platform information for. Use 'US' for USA and 'IN' for India.", "type": "str", "default": "IN"}}}, {"name": "season_episodes", "description": "Fetch episodes for a given season from the Netflix RapidAPI.", "parameters": {"ids": {"description": "Season IDs (separated by commas) for which episodes are to be fetched.", "type": "str", "default": "80077209,80117715"}, "lang": {"description": "Language code for the episodes. Defaults to 'en'.", "type": "str, optional", "default": "en"}, "limit": {"description": "The maximum number of episodes to return. Defaults to 25.", "type": "int, optional", "default": "25"}, "offset": {"description": "The offset for pagination. Defaults to 0.", "type": "int, optional", "default": "0"}}}, {"name": "actors_random", "description": "Fetches a random list of actors from the Movies Database RapidAPI.", "parameters": {"limit": {"description": "The number of random actors to retrieve. Defaults to None, and the API will use its default value if not specified.", "type": "int, optional", "default": ""}}}]
|
[
{
"from": "human",
"value": "Fetch the episodes of the seasons with IDs 1234 and 5678 in German, limiting the results to 10 and starting from the 5th result, using the season_episodes function."
}
] |
[{"name": "season_episodes", "arguments": {"ids": "1234,5678", "lang": "de", "limit": 10, "offset": 5}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"fetch_movies\", \"description\": \"Fetch a list of movies from 2000 to 2019 based on the selected genre and year using the specified RapidAPI key.\", \"parameters\": {\"year\": {\"description\": \"The year for which to fetch movies. Should be between 2000 and 2019.\", \"type\": \"str\", \"default\": \"2018\"}, \"genre\": {\"description\": \"The genre of the movies to fetch.\", \"type\": \"str\", \"default\": \"Comedy\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"ott_providers\", \"description\": \"Fetch information on OTT platforms supported in a given region.\", \"parameters\": {\"region\": {\"description\": \"The region to get OTT platform information for. Use 'US' for USA and 'IN' for India.\", \"type\": \"str\", \"default\": \"IN\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"season_episodes\", \"description\": \"Fetch episodes for a given season from the Netflix RapidAPI.\", \"parameters\": {\"ids\": {\"description\": \"Season IDs (separated by commas) for which episodes are to be fetched.\", \"type\": \"str\", \"default\": \"80077209,80117715\"}, \"lang\": {\"description\": \"Language code for the episodes. Defaults to 'en'.\", \"type\": \"str, optional\", \"default\": \"en\"}, \"limit\": {\"description\": \"The maximum number of episodes to return. Defaults to 25.\", \"type\": \"int, optional\", \"default\": \"25\"}, \"offset\": {\"description\": \"The offset for pagination. Defaults to 0.\", \"type\": \"int, optional\", \"default\": \"0\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"actors_random\", \"description\": \"Fetches a random list of actors from the Movies Database RapidAPI.\", \"parameters\": {\"limit\": {\"description\": \"The number of random actors to retrieve. Defaults to None, and the API will use its default value if not specified.\", \"type\": \"int, optional\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Fetch the episodes of the seasons with IDs 1234 and 5678 in German, limiting the results to 10 and starting from the 5th result, using the season_episodes function.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"season_episodes\", \"arguments\": {\"ids\": \"1234,5678\", \"lang\": \"de\", \"limit\": 10, \"offset\": 5}}]",
"style": "rule"
}
|
{
"index": 21,
"split": "train"
}
|
[{"name": "word_frequency", "description": "Calculates the frequency of each word in a text.", "parameters": {"text": {"description": "The input text.", "type": "str"}}}, {"name": "count_occurrences", "description": "Counts the occurrences of each element in a list and returns a dictionary with the counts.", "parameters": {"lst": {"description": "The input list.", "type": "List"}}}, {"name": "is_armstrong_number", "description": "Checks if a number is an Armstrong number.", "parameters": {"num": {"description": "The number to check.", "type": "int"}}}, {"name": "binary_search", "description": "Performs binary search on a sorted list to find the index of a target value.", "parameters": {"arr": {"description": "The sorted list of integers.", "type": "List[int]"}, "target": {"description": "The target value to search for.", "type": "int"}}}]
|
[
{
"from": "human",
"value": "Calculate word frequencies in the paragraph: 'The quick brown fox jumps over the lazy dog.'"
}
] |
[{"name": "word_frequency", "arguments": {"text": "The quick brown fox jumps over the lazy dog."}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"word_frequency\", \"description\": \"Calculates the frequency of each word in a text.\", \"parameters\": {\"text\": {\"description\": \"The input text.\", \"type\": \"str\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"count_occurrences\", \"description\": \"Counts the occurrences of each element in a list and returns a dictionary with the counts.\", \"parameters\": {\"lst\": {\"description\": \"The input list.\", \"type\": \"List\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"is_armstrong_number\", \"description\": \"Checks if a number is an Armstrong number.\", \"parameters\": {\"num\": {\"description\": \"The number to check.\", \"type\": \"int\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"binary_search\", \"description\": \"Performs binary search on a sorted list to find the index of a target value.\", \"parameters\": {\"arr\": {\"description\": \"The sorted list of integers.\", \"type\": \"List[int]\"}, \"target\": {\"description\": \"The target value to search for.\", \"type\": \"int\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Calculate word frequencies in the paragraph: 'The quick brown fox jumps over the lazy dog.'",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"word_frequency\", \"arguments\": {\"text\": \"The quick brown fox jumps over the lazy dog.\"}}]",
"style": "rule"
}
|
{
"index": 22,
"split": "train"
}
|
[{"name": "hijos", "description": "Retrieves the list of children (hijos) from the Colegio Santa Ana API.", "parameters": {"authorization": {"description": "The authorization token required for accessing the API.", "type": "str", "default": "1234567890"}}}, {"name": "v1_babynames", "description": "Fetches a list of baby names from the API Ninjas Baby Names API endpoint.", "parameters": {"popular_only": {"description": "Whether to only return popular (top 10%) of names. Must be either 'true' or 'false'. Defaults to 'true' if not set.", "type": "str, optional", "default": ""}, "gender": {"description": "The gender for the baby names. Must be one of the following: 'boy', 'girl', or 'neutral'.", "type": "str, optional", "default": ""}}}, {"name": "topic", "description": "Fetches topics from the Quotes Diffusion API with optional pagination and search functionality.", "parameters": {"page": {"description": "Page number for pagination. Defaults to None.", "type": "int, optional", "default": ""}, "search": {"description": "Custom search term to filter topics. Defaults to None.", "type": "str, optional", "default": ""}}}, {"name": "mensajes", "description": "This function fetches messages from the Colegio Santa Ana API.", "parameters": {"authorization": {"description": "The authorization token required for the API request.", "type": "str", "default": "1234567890"}}}, {"name": "sign", "description": "Fetch horoscope information for a given astrological sign using the Horoscope Astrology API.", "parameters": {"s": {"description": "The astrological sign to fetch information for. Valid options include 'aries', 'taurus', 'gemini', 'cancer', 'leo', 'virgo', 'libra', 'scorpio', 'sagittarius', 'capricorn', 'aquarius', and 'pisces'.", "type": "str", "default": "libra"}}}, {"name": "tsv2", "description": "Fetches values from a specified data subset within a dataset using the ODAM RapidAPI.", "parameters": {"subset": {"description": "Identifier for the data subset.", "type": "str", "default": "samples"}, "dataset": {"description": "Identifier for the dataset.", "type": "str", "default": "frim1"}, "limit": {"description": "Limits the number of lines in the output table. Defaults to 10. A value of 0 means no limit.", "type": "int, optional", "default": 10}, "debug": {"description": "Debug status. When set to a non-zero value, debug mode is enabled. Defaults to 0.", "type": "int, optional", "default": 0}, "format": {"description": "Output format. Defaults to 'tsv'.", "type": "str, optional", "default": "tsv"}}}, {"name": "file_digest", "description": "Calculates the digest/checksum for a given file by making a request to the pragmavantapi via RapidAPI.", "parameters": {"source": {"description": "The URL of the file for which the digest/checksum is to be calculated.", "type": "str", "default": "http://google.com"}, "cache_key": {"description": "An optional cache key for quicker lookup. Defaults to None.", "type": "str, optional", "default": ""}}}]
|
[
{
"from": "human",
"value": "Fetch topics from the Quotes Diffusion API related to 'inspiration' on page 2."
}
] |
[{"name": "topic", "arguments": {"search": "inspiration", "page": 2}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"hijos\", \"description\": \"Retrieves the list of children (hijos) from the Colegio Santa Ana API.\", \"parameters\": {\"authorization\": {\"description\": \"The authorization token required for accessing the API.\", \"type\": \"str\", \"default\": \"1234567890\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"v1_babynames\", \"description\": \"Fetches a list of baby names from the API Ninjas Baby Names API endpoint.\", \"parameters\": {\"popular_only\": {\"description\": \"Whether to only return popular (top 10%) of names. Must be either 'true' or 'false'. Defaults to 'true' if not set.\", \"type\": \"str, optional\", \"default\": \"\"}, \"gender\": {\"description\": \"The gender for the baby names. Must be one of the following: 'boy', 'girl', or 'neutral'.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"topic\", \"description\": \"Fetches topics from the Quotes Diffusion API with optional pagination and search functionality.\", \"parameters\": {\"page\": {\"description\": \"Page number for pagination. Defaults to None.\", \"type\": \"int, optional\", \"default\": \"\"}, \"search\": {\"description\": \"Custom search term to filter topics. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"mensajes\", \"description\": \"This function fetches messages from the Colegio Santa Ana API.\", \"parameters\": {\"authorization\": {\"description\": \"The authorization token required for the API request.\", \"type\": \"str\", \"default\": \"1234567890\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"sign\", \"description\": \"Fetch horoscope information for a given astrological sign using the Horoscope Astrology API.\", \"parameters\": {\"s\": {\"description\": \"The astrological sign to fetch information for. Valid options include 'aries', 'taurus', 'gemini', 'cancer', 'leo', 'virgo', 'libra', 'scorpio', 'sagittarius', 'capricorn', 'aquarius', and 'pisces'.\", \"type\": \"str\", \"default\": \"libra\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"tsv2\", \"description\": \"Fetches values from a specified data subset within a dataset using the ODAM RapidAPI.\", \"parameters\": {\"subset\": {\"description\": \"Identifier for the data subset.\", \"type\": \"str\", \"default\": \"samples\"}, \"dataset\": {\"description\": \"Identifier for the dataset.\", \"type\": \"str\", \"default\": \"frim1\"}, \"limit\": {\"description\": \"Limits the number of lines in the output table. Defaults to 10. A value of 0 means no limit.\", \"type\": \"int, optional\", \"default\": 10}, \"debug\": {\"description\": \"Debug status. When set to a non-zero value, debug mode is enabled. Defaults to 0.\", \"type\": \"int, optional\", \"default\": 0}, \"format\": {\"description\": \"Output format. Defaults to 'tsv'.\", \"type\": \"str, optional\", \"default\": \"tsv\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"file_digest\", \"description\": \"Calculates the digest/checksum for a given file by making a request to the pragmavantapi via RapidAPI.\", \"parameters\": {\"source\": {\"description\": \"The URL of the file for which the digest/checksum is to be calculated.\", \"type\": \"str\", \"default\": \"http://google.com\"}, \"cache_key\": {\"description\": \"An optional cache key for quicker lookup. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Fetch topics from the Quotes Diffusion API related to 'inspiration' on page 2.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"topic\", \"arguments\": {\"search\": \"inspiration\", \"page\": 2}}]",
"style": "rule"
}
|
{
"index": 23,
"split": "train"
}
|
[{"name": "get_products_in_category", "description": "Fetches products in a specified category from a remote catalog using the RapidAPI service.", "parameters": {"category": {"description": "The category from which to fetch products.", "type": "str", "default": ""}, "limit": {"description": "The number of products to retrieve.", "type": "int", "default": ""}, "skip": {"description": "The number of products to skip before starting to retrieve.", "type": "int", "default": ""}}}, {"name": "specificcolorset", "description": "Fetch a set of colors based on the input color name using the ColorMe API.", "parameters": {"colorname": {"description": "The name of the base color to generate a set of colors from (e.g., \"yellow\", \"red\").", "type": "str", "default": "yellow"}}}]
|
[
{
"from": "human",
"value": "Fetch 7 products from the books category, skipping the first 5 products, and generate a color set based on the color 'purple'."
}
] |
[{"name": "get_products_in_category", "arguments": {"category": "books", "limit": 7, "skip": 5}}, {"name": "specificcolorset", "arguments": {"colorname": "purple"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_products_in_category\", \"description\": \"Fetches products in a specified category from a remote catalog using the RapidAPI service.\", \"parameters\": {\"category\": {\"description\": \"The category from which to fetch products.\", \"type\": \"str\", \"default\": \"\"}, \"limit\": {\"description\": \"The number of products to retrieve.\", \"type\": \"int\", \"default\": \"\"}, \"skip\": {\"description\": \"The number of products to skip before starting to retrieve.\", \"type\": \"int\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"specificcolorset\", \"description\": \"Fetch a set of colors based on the input color name using the ColorMe API.\", \"parameters\": {\"colorname\": {\"description\": \"The name of the base color to generate a set of colors from (e.g., \\\"yellow\\\", \\\"red\\\").\", \"type\": \"str\", \"default\": \"yellow\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Fetch 7 products from the books category, skipping the first 5 products, and generate a color set based on the color 'purple'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"get_products_in_category\", \"arguments\": {\"category\": \"books\", \"limit\": 7, \"skip\": 5}}, {\"name\": \"specificcolorset\", \"arguments\": {\"colorname\": \"purple\"}}]",
"style": "rule"
}
|
{
"index": 24,
"split": "train"
}
|
[{"name": "title_cast_crew", "description": "Fetches and returns the cast and crew information of a given title from the Watchmode API.", "parameters": {"is_id": {"description": "The ID of the title. This can be the Watchmode ID, IMDB ID, or a TMDB type and ID combination.", "type": "str", "default": "3173903"}}}]
|
[
{
"from": "human",
"value": "Fetch the cast and crew details for a movie with a custom API key 'abcdefghijklmnopqrstuvwxyz123456'."
}
] |
[{"name": "title_cast_crew", "arguments": {"is_id": "tt0111161"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"title_cast_crew\", \"description\": \"Fetches and returns the cast and crew information of a given title from the Watchmode API.\", \"parameters\": {\"is_id\": {\"description\": \"The ID of the title. This can be the Watchmode ID, IMDB ID, or a TMDB type and ID combination.\", \"type\": \"str\", \"default\": \"3173903\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Fetch the cast and crew details for a movie with a custom API key 'abcdefghijklmnopqrstuvwxyz123456'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"title_cast_crew\", \"arguments\": {\"is_id\": \"tt0111161\"}}]",
"style": "rule"
}
|
{
"index": 25,
"split": "train"
}
|
[{"name": "software_assets", "description": "Fetch details of SaaS/Software tools based on specified feature tags or software name.", "parameters": {"q": {"description": "Query string representing the list of feature tags or the name of the software for which details are to be fetched.", "type": "str", "default": "email-marketing, landing-pages"}}}, {"name": "get_campaign_lead_from_id", "description": "Retrieve lead information from a campaign using its ID via the Fomo API.", "parameters": {"is_id": {"description": "The ID of the campaign to retrieve the lead information from.", "type": "str", "default": "Lead"}}}, {"name": "loginuser", "description": "Logs in a user by sending their credentials to a specified endpoint.", "parameters": {"username": {"description": "The user's username for login.", "type": "str", "default": ""}, "password": {"description": "The user's password for login in clear text.", "type": "str", "default": ""}}}, {"name": "loginuser", "description": "Logs in a user to the specified toolbench API by making a GET request to the login endpoint with the provided username and password.", "parameters": {"username": {"description": "The username for login.", "type": "str", "default": ""}, "password": {"description": "The password for login in clear text.", "type": "str", "default": ""}}}]
|
[
{
"from": "human",
"value": "Identify SaaS tools offering 'ticketing' and 'event management' features and gather lead information for campaigns with IDs 'ccccc' and 'ddddd' from Fomo API. Moreover, log in users 'johnthompson' and 'janethompson' to my_api_12345 and freeplanhardlimitexternal APIs using their password 'secr3tpa$$w'."
}
] |
[{"name": "software_assets", "arguments": {"q": "ticketing, event management"}}, {"name": "get_campaign_lead_from_id", "arguments": {"is_id": "ccccc"}}, {"name": "get_campaign_lead_from_id", "arguments": {"is_id": "ddddd"}}, {"name": "loginuser", "arguments": {"username": "johnthompson", "password": "secr3tpa$$w"}}, {"name": "loginuser", "arguments": {"username": "janethompson", "password": "secr3tpa$$w"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"software_assets\", \"description\": \"Fetch details of SaaS/Software tools based on specified feature tags or software name.\", \"parameters\": {\"q\": {\"description\": \"Query string representing the list of feature tags or the name of the software for which details are to be fetched.\", \"type\": \"str\", \"default\": \"email-marketing, landing-pages\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"get_campaign_lead_from_id\", \"description\": \"Retrieve lead information from a campaign using its ID via the Fomo API.\", \"parameters\": {\"is_id\": {\"description\": \"The ID of the campaign to retrieve the lead information from.\", \"type\": \"str\", \"default\": \"Lead\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"loginuser\", \"description\": \"Logs in a user by sending their credentials to a specified endpoint.\", \"parameters\": {\"username\": {\"description\": \"The user's username for login.\", \"type\": \"str\", \"default\": \"\"}, \"password\": {\"description\": \"The user's password for login in clear text.\", \"type\": \"str\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"loginuser\", \"description\": \"Logs in a user to the specified toolbench API by making a GET request to the login endpoint with the provided username and password.\", \"parameters\": {\"username\": {\"description\": \"The username for login.\", \"type\": \"str\", \"default\": \"\"}, \"password\": {\"description\": \"The password for login in clear text.\", \"type\": \"str\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Identify SaaS tools offering 'ticketing' and 'event management' features and gather lead information for campaigns with IDs 'ccccc' and 'ddddd' from Fomo API. Moreover, log in users 'johnthompson' and 'janethompson' to my_api_12345 and freeplanhardlimitexternal APIs using their password 'secr3tpa$$w'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"software_assets\", \"arguments\": {\"q\": \"ticketing, event management\"}}, {\"name\": \"get_campaign_lead_from_id\", \"arguments\": {\"is_id\": \"ccccc\"}}, {\"name\": \"get_campaign_lead_from_id\", \"arguments\": {\"is_id\": \"ddddd\"}}, {\"name\": \"loginuser\", \"arguments\": {\"username\": \"johnthompson\", \"password\": \"secr3tpa$$w\"}}, {\"name\": \"loginuser\", \"arguments\": {\"username\": \"janethompson\", \"password\": \"secr3tpa$$w\"}}]",
"style": "rule"
}
|
{
"index": 26,
"split": "train"
}
|
[{"name": "Quote of The Day", "description": "Returns a quote of the day, including the quote text and author.", "parameters": {"type": "dict", "properties": {"category": {"description": "Optional category filter for the quote (e.g., inspirational, humorous, etc.).", "type": "string"}}, "required": ["category"]}, "required": null}, {"name": "PL Car Averages", "description": "Retrieve average data for cars by year, make, and model", "parameters": {"type": "dict", "properties": {"year": {"description": "The year of the car", "type": "int"}, "make": {"description": "The make of the car", "type": "string"}, "model": {"description": "The model of the car", "type": "string"}}, "required": ["year", "make", "model"]}, "required": null}, {"name": "Get Glyph Information", "description": "Retrieve information about a specific Unicode codepoint, including the corresponding glyph.", "parameters": {"type": "dict", "properties": {"codepoint": {"description": "The hexadecimal codepoint of the Unicode character to retrieve information about.", "type": "string", "default": "0A2F"}}, "required": ["codepoint"]}, "required": null}, {"name": "Get Messages", "description": "Retrieve a list of messages based on the specified category", "parameters": {"type": "dict", "properties": {"category": {"description": "Select a category to get unlimited messages", "type": "string", "default": "love"}}, "required": ["category"]}, "required": null}, {"name": "Wayback Machine Captures", "description": "Captures a web page or snapshot using the Wayback Machine API", "parameters": {"type": "dict", "properties": {"url": {"description": "The URL of the web page to capture", "type": "string"}, "string": {"description": "The string to capture the web page (YYYY-MM-DD format)", "type": "string", "default": "2023-04-15"}}, "required": ["url"]}, "required": null}, {"name": "Get Brewery", "description": "Retrieve information about a specific brewery by its ID", "parameters": {"type": "dict", "properties": {"id": {"description": "ID of the brewery", "type": "int"}}, "required": ["id"]}, "required": null}]
|
[
{
"from": "human",
"value": "I am working on a report and need some specific information. Could you help me with the following:\n1. Get the average data for a 2018 Ford Mustang.\n2. Retrieve a quote of the day in the \"motivational\" category.\n3. Get messages in the \"friendship\" category.\n4. Capture the current state of the URL \"https://www.example.com\" using the Wayback Machine.\n5. Retrieve information about the Unicode codepoint \"1F600\"."
}
] |
[{"name": "PL Car Averages", "arguments": {"year": 2018, "make": "Ford", "model": "Mustang"}}, {"name": "Quote of The Day", "arguments": {"category": "motivational"}}, {"name": "Get Messages", "arguments": {"category": "friendship"}}, {"name": "Wayback Machine Captures", "arguments": {"url": "https://www.example.com"}}, {"name": "Get Glyph Information", "arguments": {"codepoint": "1F600"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out.
The current time is 2026-01-24 08:54:52.Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out.\nThe current time is 2026-01-24 08:54:52.Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"Quote of The Day\", \"description\": \"Returns a quote of the day, including the quote text and author.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"category\": {\"description\": \"Optional category filter for the quote (e.g., inspirational, humorous, etc.).\", \"type\": \"string\"}}, \"required\": [\"category\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"PL Car Averages\", \"description\": \"Retrieve average data for cars by year, make, and model\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"year\": {\"description\": \"The year of the car\", \"type\": \"int\"}, \"make\": {\"description\": \"The make of the car\", \"type\": \"string\"}, \"model\": {\"description\": \"The model of the car\", \"type\": \"string\"}}, \"required\": [\"year\", \"make\", \"model\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Get Glyph Information\", \"description\": \"Retrieve information about a specific Unicode codepoint, including the corresponding glyph.\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"codepoint\": {\"description\": \"The hexadecimal codepoint of the Unicode character to retrieve information about.\", \"type\": \"string\", \"default\": \"0A2F\"}}, \"required\": [\"codepoint\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Get Messages\", \"description\": \"Retrieve a list of messages based on the specified category\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"category\": {\"description\": \"Select a category to get unlimited messages\", \"type\": \"string\", \"default\": \"love\"}}, \"required\": [\"category\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Wayback Machine Captures\", \"description\": \"Captures a web page or snapshot using the Wayback Machine API\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"url\": {\"description\": \"The URL of the web page to capture\", \"type\": \"string\"}, \"string\": {\"description\": \"The string to capture the web page (YYYY-MM-DD format)\", \"type\": \"string\", \"default\": \"2023-04-15\"}}, \"required\": [\"url\"]}, \"required\": null}}\n{\"type\": \"function\", \"function\": {\"name\": \"Get Brewery\", \"description\": \"Retrieve information about a specific brewery by its ID\", \"parameters\": {\"type\": \"dict\", \"properties\": {\"id\": {\"description\": \"ID of the brewery\", \"type\": \"int\"}}, \"required\": [\"id\"]}, \"required\": null}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "I am working on a report and need some specific information. Could you help me with the following:\n1. Get the average data for a 2018 Ford Mustang.\n2. Retrieve a quote of the day in the \"motivational\" category.\n3. Get messages in the \"friendship\" category.\n4. Capture the current state of the URL \"https://www.example.com\" using the Wayback Machine.\n5. Retrieve information about the Unicode codepoint \"1F600\".",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"PL Car Averages\", \"arguments\": {\"year\": 2018, \"make\": \"Ford\", \"model\": \"Mustang\"}}, {\"name\": \"Quote of The Day\", \"arguments\": {\"category\": \"motivational\"}}, {\"name\": \"Get Messages\", \"arguments\": {\"category\": \"friendship\"}}, {\"name\": \"Wayback Machine Captures\", \"arguments\": {\"url\": \"https://www.example.com\"}}, {\"name\": \"Get Glyph Information\", \"arguments\": {\"codepoint\": \"1F600\"}}]",
"style": "rule"
}
|
{
"index": 27,
"split": "train"
}
|
[{"name": "get_referral_domains", "description": "Retrieves a list of referral domains for a given domain using the Toolbench RapidAPI.", "parameters": {"domain": {"description": "The domain to retrieve referral domains for.", "type": "str", "default": "getecz.com"}}}]
|
[
{
"from": "human",
"value": "What are the referral domains for 'twitter.com'?"
}
] |
[{"name": "get_referral_domains", "arguments": {"domain": "twitter.com"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"get_referral_domains\", \"description\": \"Retrieves a list of referral domains for a given domain using the Toolbench RapidAPI.\", \"parameters\": {\"domain\": {\"description\": \"The domain to retrieve referral domains for.\", \"type\": \"str\", \"default\": \"getecz.com\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "What are the referral domains for 'twitter.com'?",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"get_referral_domains\", \"arguments\": {\"domain\": \"twitter.com\"}}]",
"style": "rule"
}
|
{
"index": 28,
"split": "train"
}
|
[{"name": "baseball_match_statistics", "description": "Fetches and returns statistics for a specified baseball match.", "parameters": {"match_id": {"description": "The unique identifier for the baseball match to retrieve statistics for.", "type": "str", "default": "9863520"}}}]
|
[
{
"from": "human",
"value": "Retrieve details for the baseball game with ID 'T778899'."
}
] |
[{"name": "baseball_match_statistics", "arguments": {"match_id": "T778899"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"baseball_match_statistics\", \"description\": \"Fetches and returns statistics for a specified baseball match.\", \"parameters\": {\"match_id\": {\"description\": \"The unique identifier for the baseball match to retrieve statistics for.\", \"type\": \"str\", \"default\": \"9863520\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Retrieve details for the baseball game with ID 'T778899'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"baseball_match_statistics\", \"arguments\": {\"match_id\": \"T778899\"}}]",
"style": "rule"
}
|
{
"index": 29,
"split": "train"
}
|
[{"name": "getmusiclenses", "description": "Fetches music lenses from the Snapchat API.", "parameters": {"cursor": {"description": "A string to navigate through paginated results. Defaults to None.", "type": "str, optional", "default": ""}}}, {"name": "gettrendinglenses", "description": "Fetches the trending lenses from Snapchat using the RapidAPI service.", "parameters": {"cursor": {"description": "A cursor for pagination. Defaults to None.", "type": "str, optional", "default": ""}}}, {"name": "sign_info", "description": "Fetches and returns the sign information from the OnlyFans API using the provided user agent and RapidAPI key.", "parameters": {"useragent": {"description": "The user agent string to be used in the API request.", "type": "str", "default": ""}}}]
|
[
{
"from": "human",
"value": "I'm interested in seeing the sign information for a user on OnlyFans. Their user agent is 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3'. Can you help me with that?"
}
] |
[{"name": "sign_info", "arguments": {"useragent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"getmusiclenses\", \"description\": \"Fetches music lenses from the Snapchat API.\", \"parameters\": {\"cursor\": {\"description\": \"A string to navigate through paginated results. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"gettrendinglenses\", \"description\": \"Fetches the trending lenses from Snapchat using the RapidAPI service.\", \"parameters\": {\"cursor\": {\"description\": \"A cursor for pagination. Defaults to None.\", \"type\": \"str, optional\", \"default\": \"\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"sign_info\", \"description\": \"Fetches and returns the sign information from the OnlyFans API using the provided user agent and RapidAPI key.\", \"parameters\": {\"useragent\": {\"description\": \"The user agent string to be used in the API request.\", \"type\": \"str\", \"default\": \"\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "I'm interested in seeing the sign information for a user on OnlyFans. Their user agent is 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3'. Can you help me with that?",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"sign_info\", \"arguments\": {\"useragent\": \"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3\"}}]",
"style": "rule"
}
|
{
"index": 30,
"split": "train"
}
|
[{"name": "games_esports", "description": "Fetches information about esports games for a given event ID using the SofaSport API.", "parameters": {"event_id": {"description": "The ID of the event to retrieve esports games for.", "type": "int", "default": "10289011"}}}, {"name": "shotmap", "description": "Fetches the shotmap data for a given event from the SofaSport API.", "parameters": {"event_id": {"description": "The unique identifier for the event.", "type": "int", "default": "10230638"}}}, {"name": "getdivision", "description": "Retrieve information about a specific NHL division by its ID via the NHL Stats and Live Data API.", "parameters": {"is_id": {"description": "The ID of the division to retrieve information for.", "type": "int", "default": 15}}}, {"name": "results", "description": "Fetch soccer match results for a specified date from the Msport API.", "parameters": {"date": {"description": "The date of the soccer matches to retrieve results for (format: 'YYYY-MM-DD').", "type": "str", "default": "2023-01-30"}, "page": {"description": "The page number of results to retrieve. Defaults to 1.", "type": "int, optional", "default": "1"}}}]
|
[
{
"from": "human",
"value": "Get the shotmap data for event ID 11223 and the esports games for event ID 44556."
}
] |
[{"name": "shotmap", "arguments": {"event_id": 11223}}, {"name": "games_esports", "arguments": {"event_id": 44556}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"games_esports\", \"description\": \"Fetches information about esports games for a given event ID using the SofaSport API.\", \"parameters\": {\"event_id\": {\"description\": \"The ID of the event to retrieve esports games for.\", \"type\": \"int\", \"default\": \"10289011\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"shotmap\", \"description\": \"Fetches the shotmap data for a given event from the SofaSport API.\", \"parameters\": {\"event_id\": {\"description\": \"The unique identifier for the event.\", \"type\": \"int\", \"default\": \"10230638\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"getdivision\", \"description\": \"Retrieve information about a specific NHL division by its ID via the NHL Stats and Live Data API.\", \"parameters\": {\"is_id\": {\"description\": \"The ID of the division to retrieve information for.\", \"type\": \"int\", \"default\": 15}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"results\", \"description\": \"Fetch soccer match results for a specified date from the Msport API.\", \"parameters\": {\"date\": {\"description\": \"The date of the soccer matches to retrieve results for (format: 'YYYY-MM-DD').\", \"type\": \"str\", \"default\": \"2023-01-30\"}, \"page\": {\"description\": \"The page number of results to retrieve. Defaults to 1.\", \"type\": \"int, optional\", \"default\": \"1\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Get the shotmap data for event ID 11223 and the esports games for event ID 44556.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"shotmap\", \"arguments\": {\"event_id\": 11223}}, {\"name\": \"games_esports\", \"arguments\": {\"event_id\": 44556}}]",
"style": "rule"
}
|
{
"index": 31,
"split": "train"
}
|
[{"name": "calculate_distance", "description": "Calculates the Euclidean distance between two points in n-dimensional space.", "parameters": {"point1": {"description": "The coordinates of the first point.", "type": "List[float]"}, "point2": {"description": "The coordinates of the second point.", "type": "List[float]"}}}]
|
[
{
"from": "human",
"value": "Calculate the Euclidean distance between the points (0, 0, 0, 0, 0) and (1, 1, 1, 1, 1), and also between the points (2, 2, 2, 2, 2) and (3, 3, 3, 3, 3) using the 'calculate_distance' function."
}
] |
[{"name": "calculate_distance", "arguments": {"point1": [0, 0, 0, 0, 0], "point2": [1, 1, 1, 1, 1]}}, {"name": "calculate_distance", "arguments": {"point1": [2, 2, 2, 2, 2], "point2": [3, 3, 3, 3, 3]}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"calculate_distance\", \"description\": \"Calculates the Euclidean distance between two points in n-dimensional space.\", \"parameters\": {\"point1\": {\"description\": \"The coordinates of the first point.\", \"type\": \"List[float]\"}, \"point2\": {\"description\": \"The coordinates of the second point.\", \"type\": \"List[float]\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Calculate the Euclidean distance between the points (0, 0, 0, 0, 0) and (1, 1, 1, 1, 1), and also between the points (2, 2, 2, 2, 2) and (3, 3, 3, 3, 3) using the 'calculate_distance' function.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"calculate_distance\", \"arguments\": {\"point1\": [0, 0, 0, 0, 0], \"point2\": [1, 1, 1, 1, 1]}}, {\"name\": \"calculate_distance\", \"arguments\": {\"point1\": [2, 2, 2, 2, 2], \"point2\": [3, 3, 3, 3, 3]}}]",
"style": "rule"
}
|
{
"index": 32,
"split": "train"
}
|
[{"name": "polygon_area_shoelace", "description": "Calculates the area of a polygon using the shoelace formula.", "parameters": {"vertices": {"description": "A list of polygon vertices represented as tuples (x, y).", "type": "List[Tuple[float, float]]"}}}, {"name": "euclidean_distance", "description": "Computes the Euclidean distance between two points in a 2D plane.", "parameters": {"point_a": {"description": "The coordinates of the first point (x, y).", "type": "Tuple[float, float]"}, "point_b": {"description": "The coordinates of the second point (x, y).", "type": "Tuple[float, float]"}}}]
|
[
{
"from": "human",
"value": "Determine the distance between the points (-1,-1) and (1,1)."
}
] |
[{"name": "euclidean_distance", "arguments": {"point_a": [-1, -1], "point_b": [1, 1]}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"polygon_area_shoelace\", \"description\": \"Calculates the area of a polygon using the shoelace formula.\", \"parameters\": {\"vertices\": {\"description\": \"A list of polygon vertices represented as tuples (x, y).\", \"type\": \"List[Tuple[float, float]]\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"euclidean_distance\", \"description\": \"Computes the Euclidean distance between two points in a 2D plane.\", \"parameters\": {\"point_a\": {\"description\": \"The coordinates of the first point (x, y).\", \"type\": \"Tuple[float, float]\"}, \"point_b\": {\"description\": \"The coordinates of the second point (x, y).\", \"type\": \"Tuple[float, float]\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Determine the distance between the points (-1,-1) and (1,1).",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"euclidean_distance\", \"arguments\": {\"point_a\": [-1, -1], \"point_b\": [1, 1]}}]",
"style": "rule"
}
|
{
"index": 33,
"split": "train"
}
|
[{"name": "runner_up", "description": "Fetches the runner-up team data for the given identifying parameter from the NCAA Final Four API.", "parameters": {"identifier": {"description": "The identifier (e.g., year) to fetch the runner-up team data for.", "type": "str", "default": "1959"}}}, {"name": "basketball_search", "description": "Search for basketball players, teams, and tournaments using the provided term.", "parameters": {"term": {"description": "The search term used to look for players, teams, and tournaments.", "type": "str", "default": "kevin"}}}]
|
[
{
"from": "human",
"value": "Who were the runner-up teams in the NCAA Final Four in the years 2000, 2005, and 2010? Also, search for basketball players named 'LeBron James'."
}
] |
[{"name": "runner_up", "arguments": {"identifier": "2000"}}, {"name": "runner_up", "arguments": {"identifier": "2005"}}, {"name": "runner_up", "arguments": {"identifier": "2010"}}, {"name": "basketball_search", "arguments": {"term": "LeBron James"}}]
|
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:
|
toolcall
|
[
{
"content": "You are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections. Here is a list of functions in JSON format that you can invoke:\n# Tool\n\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"runner_up\", \"description\": \"Fetches the runner-up team data for the given identifying parameter from the NCAA Final Four API.\", \"parameters\": {\"identifier\": {\"description\": \"The identifier (e.g., year) to fetch the runner-up team data for.\", \"type\": \"str\", \"default\": \"1959\"}}}}\n{\"type\": \"function\", \"function\": {\"name\": \"basketball_search\", \"description\": \"Search for basketball players, teams, and tournaments using the provided term.\", \"parameters\": {\"term\": {\"description\": \"The search term used to look for players, teams, and tournaments.\", \"type\": \"str\", \"default\": \"kevin\"}}}}\n</tools>\n\nIn each action step, you MUST: \n1. Think about the reasoning process in the mind and enclosed your reasoning within <think> </think> XML tags.\n2. Then, provide a json object with function names and arguments within <tool_call></tool_call> XML tags. i.e., <tool_call>[{\"name\": <function-name>, \"arguments\": <args-json-object>}, {\"name\": <function-name2>, \"arguments\": <args-json-object2>}, ...]</tool_call>\n3. Make sure both the reasoning and the tool call steps are included together in one single reply.\nA complete reply example is: <think>To address the query, I need to send the email to Bob and then buy the banana through walmart. </think> <tool_call> [{\"name\": \"email\", \"arguments\": {\"receiver\": \"Bob\", \"content\": \"I will bug banana through walmart\"}}, {\"name\": \"walmart\", \"arguments\": {\"input\": \"banana\"}}]</tool_call>. Please make sure the type of the arguments is correct. \n",
"role": "system"
},
{
"content": "Who were the runner-up teams in the NCAA Final Four in the years 2000, 2005, and 2010? Also, search for basketball players named 'LeBron James'.",
"role": "user"
}
] |
toolcall
|
{
"ground_truth": "[{\"name\": \"runner_up\", \"arguments\": {\"identifier\": \"2000\"}}, {\"name\": \"runner_up\", \"arguments\": {\"identifier\": \"2005\"}}, {\"name\": \"runner_up\", \"arguments\": {\"identifier\": \"2010\"}}, {\"name\": \"basketball_search\", \"arguments\": {\"term\": \"LeBron James\"}}]",
"style": "rule"
}
|
{
"index": 34,
"split": "train"
}
|
End of preview. Expand
in Data Studio
README.md exists but content is empty.
- Downloads last month
- 17