semantic-router/e2e/profiles/rag-hybrid-search/values.yaml at 169fb47ffdd0b34eabc48b9a3399d336c3ab3410 · vllm-project/semantic-router · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
config:
  version: v0.3
  listeners: []
  providers:
    defaults:
      default_model: general-expert
      reasoning_families:
        qwen3:
          type: chat_template_kwargs
          parameter: enable_thinking
    models:
      - name: base-model
        reasoning_family: qwen3
        backend_refs:
          - name: local-vllm
            endpoint: vllm-llama3-8b-instruct.default.svc.cluster.local:8000
            weight: 1
  routing:
    decisions:
      - name: other_decision
        description: General knowledge and miscellaneous topics
        priority: 1
        rules:
          operator: OR
          conditions:
            - type: domain
              name: other
        modelRefs:
          - model: base-model
            lora_name: general-expert
            use_reasoning: false
    signals:
      domains:
        - name: other
          description: General knowledge and miscellaneous topics
    modelCards:
      - name: base-model
        loras:
          - name: general-expert
            description: General-purpose adapter
      - name: general-expert
  global:
    model_catalog:
      kbs: []
    router:
      strategy: priority
    services: {}
    stores:
      vector_store:
        enabled: true
        backend_type: llama_stack
        file_storage_dir: /tmp/vsr-data
        max_file_size_mb: 50
        embedding_model: mmbert
        embedding_dimension: 384
        ingestion_workers: 2
        ingestion_batch_size: 64
        ingestion_drain_timeout_seconds: 30
        supported_formats:
          - .txt
          - .md
        llama_stack:
          endpoint: http://llama-stack.llama-stack-system.svc.cluster.local:8321
          embedding_model: sentence-transformers/all-MiniLM-L6-v2
          request_timeout_seconds: 120
          search_type: hybrid
      semantic_cache:
        enabled: false
        embedding_model: mmbert
      memory:
        embedding_model: mmbert
    integrations: {}