codescout 0.14.0

High-performance coding agent toolkit MCP server
Documentation 
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
73
74
75
76
77
78
79
80

name: codescout-retrieval-matrix

# Parallel embedder containers for the chunk×model benchmark matrix.
# Run alongside docker-compose.yml's `qdrant` service; this file does NOT define qdrant.
#
#   docker compose -f docker-compose.yml -f docker-compose.matrix.yml up -d \
#     qdrant embedder-sparse embedder-jina-base embedder-bge-small embedder-jina-small
#
# CodeRankEmbed is NOT in this file — it runs on the AMD card via host llama-server
# at http://localhost:43300 (managed externally by ~/agents/llm).

services:
  embedder-sparse:
    image: ghcr.io/huggingface/text-embeddings-inference:86-1.8
    container_name: codescout-matrix-sparse
    restart: unless-stopped
    command: ["--model-id", "naver/splade-cocondenser-ensembledistil", "--pooling", "splade", "--dtype", "float16", "--auto-truncate"]
    ports:
      - "127.0.0.1:8091:80"
    volumes:
      - model_cache:/data
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    networks: [retrieval_net]

  embedder-jina-base:
    image: ghcr.io/huggingface/text-embeddings-inference:86-1.8
    container_name: codescout-matrix-jina-base
    restart: unless-stopped
    command: ["--model-id", "jinaai/jina-embeddings-v2-base-code", "--dtype", "float16", "--auto-truncate"]
    ports:
      - "127.0.0.1:8090:80"
    shm_size: 2g
    volumes:
      - model_cache:/data
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    networks: [retrieval_net]

  embedder-bge-small:
    image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.6
    container_name: codescout-matrix-bge-small
    restart: unless-stopped
    command: ["--model-id", "BAAI/bge-small-en-v1.5", "--dtype", "float32", "--auto-truncate"]
    ports:
      - "127.0.0.1:8092:80"
    volumes:
      - model_cache:/data
    networks: [retrieval_net]

  embedder-jina-small:
    image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.6
    container_name: codescout-matrix-jina-small
    restart: unless-stopped
    command: ["--model-id", "jinaai/jina-embeddings-v2-small-en", "--dtype", "float32", "--auto-truncate"]
    ports:
      - "127.0.0.1:8093:80"
    volumes:
      - model_cache:/data
    networks: [retrieval_net]

volumes:
  model_cache:
    external: true
    name: codescout-retrieval_model_cache

networks:
  retrieval_net:
    external: true
    name: codescout-retrieval_retrieval_net