类class
agentscope.modelParameters
class Parameters(BaseModel)源码Source
agentscope/model/_deepseek/_model.py:29–74参数用于 DeepSeek 聊天模型。位于 agentscope.model 包。
The parameters for the DeepSeek chat model. Declared in agentscope.model.
属性摘要Attribute Summary
| 属性Attribute | 类型Type | 说明Description | 值Value |
|---|---|---|---|
max_tokens | int | None | 最大令牌数字段。 | Field(default=None, title='Max Tokens', description='The maximum number of tokens for the LLM output.', gt=0) |
thinking_enable | bool | `thinking_enable` 字段。 | Field(default=False, title='Thinking', description='Whether to enable thinking mode. When enabled, the model outputs a chain-of-thought reasoning before the final answer via the reasoning_content field.') |
reasoning_effort | Literal['high', 'max'] | None | `reasoning_effort` 字段。 | Field(default=None, title='Reasoning Effort', description='Controls the depth of reasoning in thinking mode. Supported values: high (default), max. For compatibility, low/medium map to high, xhigh maps to max.') |
temperature | float | None | `temperature` 字段。 | Field(default=None, title='Temperature', description='The temperature for the LLM output.', ge=0, le=2) |
top_p | float | None | `top_p` 字段。 | Field(default=None, title='Top P', description='The top P value for the LLM output.', gt=0, le=1) |
属性详情Attribute Detail
max_tokens
int | None = Field(default=None, title='Max Tokens', description='The maximum number of tokens for the LLM output.', gt=0)最大令牌数字段。声明为 int | None。
The maximum number of tokens for the LLM output.
源码Source
agentscope/model/_deepseek/_model.py:32thinking_enable
bool = Field(default=False, title='Thinking', description='Whether to enable thinking mode. When enabled, the model outputs a chain-of-thought reasoning before the final answer via the reasoning_content field.')thinking_enable 字段。声明为 bool。
Whether to enable thinking mode.
源码Source
agentscope/model/_deepseek/_model.py:39reasoning_effort
Literal['high', 'max'] | None = Field(default=None, title='Reasoning Effort', description='Controls the depth of reasoning in thinking mode. Supported values: high (default), max. For compatibility, low/medium map to high, xhigh maps to max.')reasoning_effort 字段。声明为 Literal['high', 'max'] | None。
Controls the depth of reasoning in thinking mode.
源码Source
agentscope/model/_deepseek/_model.py:49temperature
float | None = Field(default=None, title='Temperature', description='The temperature for the LLM output.', ge=0, le=2)temperature 字段。声明为 float | None。
The temperature for the LLM output.
源码Source
agentscope/model/_deepseek/_model.py:60top_p
float | None = Field(default=None, title='Top P', description='The top P value for the LLM output.', gt=0, le=1)top_p 字段。声明为 float | None。
The top P value for the LLM output.
源码Source
agentscope/model/_deepseek/_model.py:68