Download tests/chat_tools_golden.json from Sariel00/Ling-3.0-tiny-RKNN: direct link, hf CLI and curl.
- Browser
- Download file 18.2 kB
-
https://huggingface.co/Sariel00/Ling-3.0-tiny-RKNN/resolve/main/tests/chat_tools_golden.json
- Command line
-
hf download hf://Sariel00/Ling-3.0-tiny-RKNN/tests/chat_tools_golden.json
-
curl -L -o chat_tools_golden.json https://huggingface.co/Sariel00/Ling-3.0-tiny-RKNN/resolve/main/tests/chat_tools_golden.json
18.2 kB
| [ | |
| { | |
| "name": "no_tools_history", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "system", | |
| "content": "请用中文回答" | |
| }, | |
| { | |
| "role": "user", | |
| "content": "你好" | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": "你好!" | |
| }, | |
| { | |
| "role": "user", | |
| "content": "继续" | |
| } | |
| ], | |
| "enable_thinking": false | |
| }, | |
| "prompt": "<role>SYSTEM</role>请用中文回答\ndetailed thinking off<|role_end|><role>HUMAN</role>你好<|role_end|><role>ASSISTANT</role>\n<think></think>你好!<|role_end|><role>HUMAN</role>继续<|role_end|><role>ASSISTANT</role>\n<think></think>" | |
| }, | |
| { | |
| "name": "tools_no_system", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "user", | |
| "content": "Reply with OK." | |
| } | |
| ], | |
| "enable_thinking": false, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "description": "Read a file's contents.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role># Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"read_file\", \"description\": \"Read a file's contents.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}}, \"required\": [\"path\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\ndetailed thinking off<|role_end|><role>HUMAN</role>Reply with OK.<|role_end|><role>ASSISTANT</role>\n<think></think>" | |
| }, | |
| { | |
| "name": "tools_system", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "system", | |
| "content": "You are terse." | |
| }, | |
| { | |
| "role": "user", | |
| "content": "Use the tools to read /tmp/scratch.txt, then patch it to change 'foo' to 'bar'." | |
| } | |
| ], | |
| "enable_thinking": false, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "description": "Read a file's contents.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path" | |
| ] | |
| } | |
| } | |
| }, | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "patch", | |
| "description": "Apply a small text patch to a file.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| }, | |
| "find": { | |
| "type": "string" | |
| }, | |
| "replace": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path", | |
| "find", | |
| "replace" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role>You are terse.\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"read_file\", \"description\": \"Read a file's contents.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}}, \"required\": [\"path\"]}}}\n{\"type\": \"function\", \"function\": {\"name\": \"patch\", \"description\": \"Apply a small text patch to a file.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}, \"find\": {\"type\": \"string\"}, \"replace\": {\"type\": \"string\"}}, \"required\": [\"path\", \"find\", \"replace\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\ndetailed thinking off<|role_end|><role>HUMAN</role>Use the tools to read /tmp/scratch.txt, then patch it to change 'foo' to 'bar'.<|role_end|><role>ASSISTANT</role>\n<think></think>" | |
| }, | |
| { | |
| "name": "tools_system_thinking_phrase", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "system", | |
| "content": "detailed thinking off\nBe brief." | |
| }, | |
| { | |
| "role": "user", | |
| "content": "Read /tmp/a." | |
| } | |
| ], | |
| "enable_thinking": false, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "description": "Read a file's contents.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role>detailed thinking off\nBe brief.\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"read_file\", \"description\": \"Read a file's contents.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}}, \"required\": [\"path\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\n<|role_end|><role>HUMAN</role>Read /tmp/a.<|role_end|><role>ASSISTANT</role>\n<think></think>" | |
| }, | |
| { | |
| "name": "tool_loop", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "system", | |
| "content": "You are terse." | |
| }, | |
| { | |
| "role": "user", | |
| "content": "Fix /tmp/a.txt." | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": null, | |
| "tool_calls": [ | |
| { | |
| "id": "call_1", | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "arguments": "{\"path\": \"/tmp/a.txt\"}" | |
| } | |
| }, | |
| { | |
| "id": "call_2", | |
| "type": "function", | |
| "function": { | |
| "name": "run", | |
| "arguments": "{\"command\": \"ls -l\", \"timeout\": 30, \"env\": {\"A\": \"1\", \"B\": [1, 2.5]}, \"dry\": false}" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "role": "tool", | |
| "tool_call_id": "call_1", | |
| "content": "foo\n" | |
| }, | |
| { | |
| "role": "tool", | |
| "tool_call_id": "call_2", | |
| "content": "total 0" | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": "Read it.", | |
| "tool_calls": [ | |
| { | |
| "id": "call_3", | |
| "type": "function", | |
| "function": { | |
| "name": "patch", | |
| "arguments": "{\"path\": \"/tmp/a.txt\", \"find\": \"foo\", \"replace\": \"bar\\n\"}" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "role": "tool", | |
| "tool_call_id": "call_3", | |
| "content": "ok" | |
| } | |
| ], | |
| "enable_thinking": false, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "description": "Read a file's contents.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path" | |
| ] | |
| } | |
| } | |
| }, | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "patch", | |
| "description": "Apply a small text patch to a file.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| }, | |
| "find": { | |
| "type": "string" | |
| }, | |
| "replace": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path", | |
| "find", | |
| "replace" | |
| ] | |
| } | |
| } | |
| }, | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "run", | |
| "description": "运行命令 — 可设超时。", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "command": { | |
| "type": "string" | |
| }, | |
| "timeout": { | |
| "type": "integer" | |
| }, | |
| "env": { | |
| "type": "object" | |
| }, | |
| "dry": { | |
| "type": "boolean" | |
| } | |
| }, | |
| "required": [ | |
| "command" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role>You are terse.\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"read_file\", \"description\": \"Read a file's contents.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}}, \"required\": [\"path\"]}}}\n{\"type\": \"function\", \"function\": {\"name\": \"patch\", \"description\": \"Apply a small text patch to a file.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}, \"find\": {\"type\": \"string\"}, \"replace\": {\"type\": \"string\"}}, \"required\": [\"path\", \"find\", \"replace\"]}}}\n{\"type\": \"function\", \"function\": {\"name\": \"run\", \"description\": \"运行命令 — 可设超时。\", \"parameters\": {\"type\": \"object\", \"properties\": {\"command\": {\"type\": \"string\"}, \"timeout\": {\"type\": \"integer\"}, \"env\": {\"type\": \"object\"}, \"dry\": {\"type\": \"boolean\"}}, \"required\": [\"command\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\ndetailed thinking off<|role_end|><role>HUMAN</role>Fix /tmp/a.txt.<|role_end|><role>ASSISTANT</role>\n<think></think><tool_call>read_file\n<arg_key>path</arg_key>\n<arg_value>/tmp/a.txt</arg_value>\n</tool_call>\n<tool_call>run\n<arg_key>command</arg_key>\n<arg_value>ls -l</arg_value><arg_key>timeout</arg_key>\n<arg_value>30</arg_value><arg_key>env</arg_key>\n<arg_value>{\"A\": \"1\", \"B\": [1, 2.5]}</arg_value><arg_key>dry</arg_key>\n<arg_value>false</arg_value>\n</tool_call><|role_end|><role>OBSERVATION</role>\n<tool_response>\nfoo\n\n</tool_response>\n<tool_response>\ntotal 0\n</tool_response><|role_end|><role>ASSISTANT</role>\n<think></think>Read it.\n<tool_call>patch\n<arg_key>path</arg_key>\n<arg_value>/tmp/a.txt</arg_value><arg_key>find</arg_key>\n<arg_value>foo</arg_value><arg_key>replace</arg_key>\n<arg_value>bar\n</arg_value>\n</tool_call><|role_end|><role>OBSERVATION</role>\n<tool_response>\nok\n</tool_response><|role_end|><role>ASSISTANT</role>\n<think></think>" | |
| }, | |
| { | |
| "name": "thinking_tool_loop", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "user", | |
| "content": "What is in /tmp/a?" | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": "", | |
| "reasoning_content": "\nI should read it.\n", | |
| "tool_calls": [ | |
| { | |
| "id": "call_1", | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "arguments": "{\"path\": \"/tmp/a\"}" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "role": "tool", | |
| "tool_call_id": "call_1", | |
| "content": "foo" | |
| } | |
| ], | |
| "enable_thinking": true, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "description": "Read a file's contents.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role># Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"read_file\", \"description\": \"Read a file's contents.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}}, \"required\": [\"path\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\ndetailed thinking on<|role_end|><role>HUMAN</role>What is in /tmp/a?<|role_end|><role>ASSISTANT</role>\n<think>I should read it.</think><tool_call>read_file\n<arg_key>path</arg_key>\n<arg_value>/tmp/a</arg_value>\n</tool_call><|role_end|><role>OBSERVATION</role>\n<tool_response>\nfoo\n</tool_response><|role_end|><role>ASSISTANT</role>\n<think>" | |
| }, | |
| { | |
| "name": "think_in_content", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "user", | |
| "content": "Hi" | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": "<think>\nplan\n</think>\n\nHello." | |
| }, | |
| { | |
| "role": "user", | |
| "content": "读取 /tmp/文件 ✓" | |
| } | |
| ], | |
| "enable_thinking": true, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "read_file", | |
| "description": "Read a file's contents.", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "path": { | |
| "type": "string" | |
| } | |
| }, | |
| "required": [ | |
| "path" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role># Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"read_file\", \"description\": \"Read a file's contents.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"path\": {\"type\": \"string\"}}, \"required\": [\"path\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\ndetailed thinking on<|role_end|><role>HUMAN</role>Hi<|role_end|><role>ASSISTANT</role>\n<think>plan</think>Hello.<|role_end|><role>HUMAN</role>读取 /tmp/文件 ✓<|role_end|><role>ASSISTANT</role>\n<think>" | |
| }, | |
| { | |
| "name": "unicode_call", | |
| "request": { | |
| "model": "mindnano-ling3-tiny", | |
| "messages": [ | |
| { | |
| "role": "user", | |
| "content": "打招呼" | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": null, | |
| "tool_calls": [ | |
| { | |
| "id": "call_1", | |
| "type": "function", | |
| "function": { | |
| "name": "run", | |
| "arguments": "{\"command\": \"echo 你好\"}" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "role": "tool", | |
| "tool_call_id": "call_1", | |
| "content": "你好" | |
| } | |
| ], | |
| "enable_thinking": false, | |
| "tools": [ | |
| { | |
| "type": "function", | |
| "function": { | |
| "name": "run", | |
| "description": "运行命令 — 可设超时。", | |
| "parameters": { | |
| "type": "object", | |
| "properties": { | |
| "command": { | |
| "type": "string" | |
| }, | |
| "timeout": { | |
| "type": "integer" | |
| }, | |
| "env": { | |
| "type": "object" | |
| }, | |
| "dry": { | |
| "type": "boolean" | |
| } | |
| }, | |
| "required": [ | |
| "command" | |
| ] | |
| } | |
| } | |
| } | |
| ] | |
| }, | |
| "prompt": "<role>SYSTEM</role># Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>\n{\"type\": \"function\", \"function\": {\"name\": \"run\", \"description\": \"运行命令 — 可设超时。\", \"parameters\": {\"type\": \"object\", \"properties\": {\"command\": {\"type\": \"string\"}, \"timeout\": {\"type\": \"integer\"}, \"env\": {\"type\": \"object\"}, \"dry\": {\"type\": \"boolean\"}}, \"required\": [\"command\"]}}}\n</tools>\n\nIf none of the functions can be used, point it out. If the given question lacks the parameters required by the function, also point it out.\nIf you need to use a function, for each function call, output the function name and arguments within the following XML format:\n<tool_call>{function-name}\n<arg_key>{arg-key-1}</arg_key>\n<arg_value>{arg-value-1}</arg_value>\n<arg_key>{arg-key-2}</arg_key>\n<arg_value>{arg-value-2}</arg_value>\n...\n</tool_call>\ndetailed thinking off<|role_end|><role>HUMAN</role>打招呼<|role_end|><role>ASSISTANT</role>\n<think></think><tool_call>run\n<arg_key>command</arg_key>\n<arg_value>echo 你好</arg_value>\n</tool_call><|role_end|><role>OBSERVATION</role>\n<tool_response>\n你好\n</tool_response><|role_end|><role>ASSISTANT</role>\n<think></think>" | |
| } | |
| ] | |