linear_rag/.env.example at develop · shell-nlp/linear_rag · 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
# 大模型配置
LLM_API_KEY=your-api-key
LLM_BASE_URL=http://your-llm-host:8080/v1
LLM_MODEL_NAME=your-chat-model

# Embedding 配置
EMBEDDING_API_URL=http://your-embedding-host:8080/v1/embeddings
EMBEDDING_MODEL_NAME=your-embedding-model
EMBEDDING_DIM=1024

# 文档处理和索引配置
SPACY_MODEL=zh_core_web_md
SPACY_N_PROCESS=1
MAX_WORKERS=16
API_PORT=12125
INDEX_PROCESS_WORKERS=10
CHUNK_TOKEN_SIZE=1000
CHUNK_OVERLAP_TOKEN_SIZE=100
BATCH_SIZE=128
MAX_UPLOAD_BYTES=104857600
ENTITY_EXPANSION_ENABLED=true
ENTITY_EXPANSION_MAX_ENTITIES=20
ENTITY_EXPANSION_TOP_K=50
NEIGHBOR_EXPANSION_ENABLED=true
# 官方图计算路径与局部候选近似路径共用以下参数
LINEAR_MAX_ITERATIONS=3
LINEAR_TOP_K_SENTENCE=1
LINEAR_PASSAGE_RATIO=1.5
LINEAR_PASSAGE_NODE_WEIGHT=0.05
LINEAR_DAMPING=0.5
LINEAR_ITERATION_THRESHOLD=0.5
LINEAR_LOCAL_CANDIDATES=200
LINEAR_MAX_NODES=100000
LINEAR_EMBEDDING_BATCH_SIZE=128
LINEAR_SEED_ENTITIES=5
LINEAR_PASSAGES_PER_ENTITY=5
LINEAR_LOCAL_MAX_PASSAGES=250
LINEAR_LOCAL_MAX_SENTENCES=750
WORKING_DIR=./import_qwen_new

# 对象存储配置：local 或 minio
OBJECT_STORAGE_PROVIDER=local
LOCAL_STORAGE_ROOT=./.data/object-storage

# OBJECT_STORAGE_PROVIDER=minio 时必填
MINIO_ENDPOINT_URL=http://your-minio-host:9000
MINIO_ACCESS_KEY=your-minio-access-key
MINIO_SECRET_KEY=your-minio-secret-key

# Elasticsearch 配置
es_url=http://your-elasticsearch-host:9200
es_user=elastic
es_password=your-elasticsearch-password