文本生成
请求方法
POST
请求地址
https://api.lanrui.co/v1/chat/completions
CURL 示例
curl -X POST https://api.lanrui.co/v1/chat/completions \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer {填入API Key,删除花括号}' \
--data '{
"stream": true,
"model": "{填入推理服务所对应的模型名称,具体格式为model_name/model_version,删除花括号}",
"messages": [
{
"role": "user",
"content": "Hello!"
}
]
}'
请求 Body 参数(application/json)
参数名 | 参数类型 | 取值范围 | 参数释义 | 是否必填 |
---|---|---|---|---|
model | string | 要使用的模型的ID | 是 | |
message | list<dict> | 调用语言模型时,将当前对话信息列表作为提示输入给模型。其中的元素按照 {"role": "user", "content": "Hello!"} 的形式传参,role可选择system 、user 、assistant 其中之一; content为字符串类型 | 是 | |
temperature | float | 默认值: 1 取值范围: [0.0, 2.0] | 要使用的采样温度(temperature),介于 0 和 2 之间。较高的值(如 0.8)会使输出更随机,而较低的值(如 0.2)则会使其更加专注和确定。建议只修改此参数或 top_p 一个参数 | 否 |
top_p | float | 默认值: 1 取值范围: (0.0, 1.0] | 一种称为“核心采样”的采样替代方法,其中模型考虑概率质量值在前 top_p 的标记的结果。因此,0.1 意味着仅考虑概率质量值前 10% 的标记。 | 否 |
stream | boolean | 默认值: false | 如果设置为 true,则模型输出以流模式返回 | 否 |
max_tokens | integer | 默认值: inf | 要生成的最大token数 | 否 |
presence_penalty | float | 默认值: 0 取值范围: [-2.0, 2.0] | 正值会根据新标记在迄今为止的文本中出现的频率惩罚新标记,增加模型谈论新话题的可能性 | 否 |
frequency_penalty | float | 默认值: 0 取值范围: [-2.0, 2.0] | 正值会根据其在文本中的现有频率惩罚新标记,从而减少模型重复同一行的可能性 | 否 |
非流式调用响应
{
"id": "chatcmpl-123",
"object": "chat.completion",
"created": 1677652288,
"model": "meta-llama/Llama-2-7b-chat-hf",
"system_fingerprint": "fp_44709d6fcb",
"choices": [{
"index": 0,
"message": {
"role": "assistant",
"content": "\n\nHello there, how may I assist you today?",
},
"logprobs": null,
"finish_reason": "stop"
}],
"usage": {
"prompt_tokens": 9,
"completion_tokens": 12,
"total_tokens": 21
}
}
流式调用响应
{"id":"chatcmpl-123","object":"chat.completion.chunk","created":1694268190,"model":"meta-llama/Llama-2-7b-chat-hf", "system_fingerprint": "fp_44709d6fcb", "choices":[{"index":0,"delta":{"role":"assistant","content":""},"logprobs":null,"finish_reason":null}]}
{"id":"chatcmpl-123","object":"chat.completion.chunk","created":1694268190,"model":"meta-llama/Llama-2-7b-chat-hf", "system_fingerprint": "fp_44709d6fcb", "choices":[{"index":0,"delta":{"content":"Hello"},"logprobs":null,"finish_reason":null}]}
....
{"id":"chatcmpl-123","object":"chat.completion.chunk","created":1694268190,"model":"meta-llama/Llama-2-7b-chat-hf", "system_fingerprint": "fp_44709d6fcb", "choices":[{"index":0,"delta":{},"logprobs":null,"finish_reason":"stop"}]}