跳转至

配置

cg_rag_config 负责读取 CG_RAG HTTP/MCP 路径、生成模型、超时、候选条文数量和检索缓存参数。环境变量优先使用 CG_RAG_*，并保留 QUESTION_GENERATED_APP_* 兼容名称。

cg_rag_config

CgRagConfig dataclass

源代码位于： eval_rag_results/scripts/question_generated_app/cg_rag_config.py 
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
@dataclass(frozen=True)
class CgRagConfig:
    http_prefix: str = "/cg-rag"
    mcp_path: str = "/cg-rag/mcp"
    generation_endpoint: str = ""
    generation_model: str = ""
    api_key: str = ""
    generation_response_format_json: bool = True
    generation_enable_thinking: bool = False
    timeout_seconds: float = DEFAULT_TIMEOUT_SECONDS
    max_context_docs: int = DEFAULT_MAX_CONTEXT_DOCS
    default_max_items: int = DEFAULT_MAX_ITEMS
    cache_max_entries: int = DEFAULT_CACHE_MAX_ENTRIES
    cache_ttl_seconds: float = DEFAULT_CACHE_TTL_SECONDS

    @property
    def generation_configured(self) -> bool:
        return bool(self.generation_endpoint and self.generation_model)

load_cg_rag_config() -> CgRagConfig

源代码位于： eval_rag_results/scripts/question_generated_app/cg_rag_config.py 
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
def load_cg_rag_config() -> CgRagConfig:
    load_env_file()
    return CgRagConfig(
        http_prefix=_clean_path(
            _env_str("CG_RAG_HTTP_PREFIX", "QUESTION_GENERATED_APP_HTTP_PREFIX", "/cg-rag"),
            "/cg-rag",
        ),
        mcp_path=_clean_path(
            _env_str("CG_RAG_MCP_PATH", "QUESTION_GENERATED_APP_MCP_PATH", "/cg-rag/mcp"),
            "/cg-rag/mcp",
        ),
        generation_endpoint=_env_str(
            "CG_RAG_GENERATION_ENDPOINT",
            "QUESTION_GENERATED_APP_GENERATION_ENDPOINT",
            "",
        ),
        generation_model=_env_str(
            "CG_RAG_GENERATION_MODEL",
            "QUESTION_GENERATED_APP_GENERATION_MODEL",
            "",
        ),
        api_key=_env_str("CG_RAG_API_KEY", "QUESTION_GENERATED_APP_API_KEY", ""),
        generation_response_format_json=_env_bool(
            "CG_RAG_GENERATION_RESPONSE_FORMAT_JSON",
            "QUESTION_GENERATED_APP_GENERATION_RESPONSE_FORMAT_JSON",
            True,
        ),
        generation_enable_thinking=_env_bool(
            "CG_RAG_GENERATION_ENABLE_THINKING",
            "QUESTION_GENERATED_APP_GENERATION_ENABLE_THINKING",
            False,
        ),
        timeout_seconds=_env_float(
            "CG_RAG_TIMEOUT_SECONDS",
            "QUESTION_GENERATED_APP_TIMEOUT_SECONDS",
            DEFAULT_TIMEOUT_SECONDS,
        ),
        max_context_docs=_env_positive_int(
            "CG_RAG_MAX_CONTEXT_DOCS",
            "QUESTION_GENERATED_APP_MAX_CONTEXT_DOCS",
            DEFAULT_MAX_CONTEXT_DOCS,
        ),
        default_max_items=_env_positive_int(
            "CG_RAG_DEFAULT_MAX_ITEMS",
            "QUESTION_GENERATED_APP_DEFAULT_MAX_ITEMS",
            DEFAULT_MAX_ITEMS,
        ),
        cache_max_entries=_env_non_negative_int(
            "CG_RAG_RETRIEVE_CACHE_MAX_ENTRIES",
            "QUESTION_GENERATED_APP_RETRIEVE_CACHE_MAX_ENTRIES",
            DEFAULT_CACHE_MAX_ENTRIES,
        ),
        cache_ttl_seconds=_env_non_negative_float(
            "CG_RAG_RETRIEVE_CACHE_TTL_SECONDS",
            "QUESTION_GENERATED_APP_RETRIEVE_CACHE_TTL_SECONDS",
            DEFAULT_CACHE_TTL_SECONDS,
        ),
    )