# ============================================================ # 多专业小模型 + 路由模型系统 — 配置 # 默认全 mock 模式(零依赖、离线可跑)。要接入真实模型,把对应 # 后端 type 改为 hf / api 即可(见 README)。 # ============================================================ system: name: multi-expert-router version: 0.1.0 router: low_confidence_threshold: 0.60 # 分类置信度低于此值 -> 直接走最后处理者 judge_fallback_threshold: 0.70 # Judge 质量分低于此值 -> 升级最后处理者 default_temperature: 0.2 # 专家系统内核执行配置(L0 默认零参数) execution: mode: rule # rule(默认,L0 零参数)| hybrid planner: rule # rule(规则拆解)| hf(可选小模型拆解) expert_backend: rule # rule(规则执行器)| hf | api(本地小模型按需加载) model_level: L0 # L0 纯规则 | L1 分类/Planner增强 | L2 领域生成 max_plan_depth: 3 # 任务拆解深度上限 classifier: type: rule # rule(零依赖)| hf(transformers) model: Qwen/Qwen3-0.6B confidence_floor: 0.55 domains: - code - math - legal - medical - finance - life - education - general # 两级路由:大领域分组(用户接口指定 group → 组内路由模型 → 组内专业小模型) # 组内路由模型只识别本组领域,体积约为统一路由模型的 1/4 domain_groups: tech: [code, math] professional: [legal, medical, finance] lifestyle: [life, education] general: [general] experts: code: { type: mock, model: Qwen/Qwen2.5-Coder-7B-Instruct } math: { type: mock, model: Qwen/Qwen3-4B-Instruct } legal: { type: mock, model: Qwen/Qwen3-4B-Instruct } medical: { type: mock, model: Qwen/Qwen3-4B-Instruct } finance: { type: mock, model: Qwen/Qwen3-4B-Instruct } life: { type: mock, model: Qwen/Qwen3-1.7B-Instruct } education: { type: mock, model: Qwen/Qwen3-1.7B-Instruct } general: { type: mock, model: Qwen/Qwen3-1.7B-Instruct } fallback: type: mock # none(降级模板)| mock | local(本地≤8B 按需加载)| api model: deepseek-chat base_url: https://api.deepseek.com/v1 api_key_env: DEEPSEEK_API_KEY judge: type: rule # rule(零依赖)| llm model: Qwen/Qwen3-1.7B-Instruct cache: enabled: true semantic_enabled: true # 语义缓存(字符 n-gram 相似度,零依赖) similarity_threshold: 0.88 promote_frequency: 5 # 命中 N 次后提升为精确缓存 # ============================================================ # v2:端云协同 LLM 协作系统(《实现方案_v2》)配置段 # v1 段(上方)保留,供 legacy 路由(POST /chat/legacy)使用。 # ============================================================ runtime: llama_server: binary: bin/llama-server.exe # 捆绑上游 release,不改源码(D1) model: models/qwen3.5-4b-q4_k_m.gguf port: 8901 hw_profile: auto # auto | gpu12 | gpu8 | cpu extra_args: ["-fa", "-ctk", "q8_0", "-ctv", "q8_0", "--cache-reuse", "256"] tiers: # 三档硬件模板(保守默认,可手动覆盖) gpu12: {ngl: 99, ctx: 32768} gpu8: {ngl: 14, ctx: 16384} cpu: {ngl: 0, ctx: 8192} architect: # 大模型(API) model: deepseek-chat base_url: https://api.deepseek.com/v1 api_key_env: DEEPSEEK_API_KEY temperature: 0.2 timeout_s: 60 worker: # 小模型(本地) backend: llama_server temperature: 0.3 max_fix_attempts: 2 per_step_timeout_s: 300 pipeline: fast_path: true rounds_cap: 6 api_token_cap: 8000 breach_policy: architect_do # architect_do | local_only review: queue_db: data/review.sqlite3 sample_rate: 0.10 # 随机抽样送审 force_tags: [safety] # brief.tags 命中即强制送审