You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
<|im_start|>system\nYou\'re a helpful assistant\n\n# 工具\n\n## 你拥有如下工具:\n\n### get_current_weather\n\nget_current_weather: Get the current weather in a given location 输入参数:{"type": "object", "properties": {"location": {"type": "string", "description": "The city and state, e.g. San Francisco, CA"}, "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}}, "required": ["location"]} 此工具的输入应为JSON对象。\n\n## 你可以在回复中插入零次、一次或多次以下命令以调用工具:\n\n✿FUNCTION✿: 工具名称,必须是[get_current_weather]之一。\n✿ARGS✿: 工具输入\n✿RESULT✿: 工具结果\n✿RETURN✿: 根据工具结果进行回复,需将图片用![](url)渲染出来<|im_end|>\n<|im_start|>user\nAudio 1: <|audio_bos|><|AUDIO|><|audio_eos|>\n上海今天天气如何<|im_end|>\n<|im_start|>assistant\n'
得到结果:
“上海今天天气晴朗,气温较低,建议出门携带外套。”
这里是因为audio模型没有经过function call微调导致的吗?
The text was updated successfully, but these errors were encountered:
Qwen2-Audio-7B-Instruct 是否支持Function Call功能?我按照Qwen2-7B-Instruct的function call prompt传入后,结果没有返回期望调用的函数。
<|im_start|>system\nYou\'re a helpful assistant\n\n# 工具\n\n## 你拥有如下工具:\n\n### get_current_weather\n\nget_current_weather: Get the current weather in a given location 输入参数:{"type": "object", "properties": {"location": {"type": "string", "description": "The city and state, e.g. San Francisco, CA"}, "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}}, "required": ["location"]} 此工具的输入应为JSON对象。\n\n## 你可以在回复中插入零次、一次或多次以下命令以调用工具:\n\n✿FUNCTION✿: 工具名称,必须是[get_current_weather]之一。\n✿ARGS✿: 工具输入\n✿RESULT✿: 工具结果\n✿RETURN✿: 根据工具结果进行回复,需将图片用![](url)渲染出来<|im_end|>\n<|im_start|>user\nAudio 1: <|audio_bos|><|AUDIO|><|audio_eos|>\n上海今天天气如何<|im_end|>\n<|im_start|>assistant\n'
得到结果:
“上海今天天气晴朗,气温较低,建议出门携带外套。”
这里是因为audio模型没有经过function call微调导致的吗?
The text was updated successfully, but these errors were encountered: