-
Notifications
You must be signed in to change notification settings - Fork 0
/
.env.example
128 lines (103 loc) · 4.48 KB
/
.env.example
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
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
ENVIRONMENT='LOCAL'
# Key to be sent as x-api-key header to authenticate requests
# WARNING: Auth disabled if not provided
PERSONAFLOW_API_KEY=<your_personaflow_api_key>
# If using openai
OPENAI_API_KEY=<your_openai_api_key>
# Required if you intend to use reranking functionality
COHERE_API_KEY=<your_cohere_api_key>
# Azure OpenAI deployment details (Optional)
AZURE_OPENAI_DEPLOYMENT_NAME=<your_azure_openai_deployment_name>
AZURE_OPENAI_API_BASE=<your_azure_openai_deployment_base_url>
AZURE_OPENAI_API_KEY=<your_azure_openai_deployment_key
AZURE_OPENAI_API_VERSION=<your_azure_openai_deployment_version>
# Postgres Database environment variables (Required)
INTERNAL_DATABASE_USER="postgres"
INTERNAL_DATABASE_PASSWORD="postgres"
INTERNAL_DATABASE_HOST="localhost"
INTERNAL_DATABASE_PORT=5432
INTERNAL_DATABASE_DATABASE="internal"
INTERNAL_DATABASE_SCHEMA="personaflow"
# Vector DB environment variables (Required)
VECTOR_DB_HOST="localhost"
VECTOR_DB_PORT=6333
VECTOR_DB_NAME="qdrant"
VECTOR_DB_API_KEY="123456789"
# default collection name - defaults to "documents", uncomment to change
VECTOR_DB_COLLECTION_NAME="documents"
# This is the default encoder - options are "openai", "azure_openai", "cohere", "huggingface", or "ollama"
VECTOR_DB_ENCODER_NAME="openai"
# vector embedding dimension size - defaults to 1536, uncomment to change
VECTOR_DB_ENCODER_DIMENSIONS=1536
VECTOR_DB_ENCODER_MODEL="text-embedding-3-small"
# This is used when there is no namespace specified for filtering when ingesting or querying documents
VECTOR_DB_DEFAULT_NAMESPACE="default"
# Number of iterations assistant is allowed to run to
# accomplish task or improve response (Required)
LANGGRAPH_RECURSION_LIMIT=5
# Max size of file uploads (Optional)
# Default is 25MB
# MAX_FILE_UPLOAD_SIZE=25000000
# Uncomment if using Azure App Insights monitoring (Optional)
# APPLICATIONINSIGHTS_CONNECTION_STRING=""
# Used for processing of unstructured documents to be ingested into vector db
# "semantic" splitting method will not work without these (required)
UNSTRUCTURED_API_KEY=""
UNSTRUCTURED_BASE_URL="http://localhost:8000"
# Default unstructured configuration (Optional)
# DEFAULT_PARTITION_STRATEGY="auto"
# DEFAULT_HI_RES_MODEL_NAME="detectron2_onnx"
# PROCESS_UNSTRUCTURED_TABLES="false"
# Tavily external search service api key - used for assistant external search tool (Optional)
# TAVILY_API_KEY=""
# Creates a separate collection that contains document summaries.
# If a query involves summarization, the summary collection will be used instead. (Optional)
# Note: this is overriden by the `summarize` parameter in the ingest request.
# Default is "true" (Optional)
# CREATE_SUMMARY_COLLECTION="true"
# Default splitter configuration (Optional)
# DEFAULT_CHUNKING_STRATEGY="semantic"
# DEFAULT_SEMANTIC_CHUNK_MIN_TOKENS=30
# DEFAULT_SEMANTIC_CHUNK_MAX_TOKENS=800
# SEMANTIC_ROLLING_WINDOW_SIZE=1 # Only for `semantic` method, cumulative window size for comparing similarity between elements
# PREFIX_TITLES="true"
# PREFIX_SUMMARY="true"
# If true, semantic lookup results will be automatically reranked by cohere for every request,
# unless overridden by the `rerank` parameter in the ingest request body.
# Requires COHERE_API_KEY to be set. Default is false (Optional)
# ENABLE_RERANK_BY_DEFAULT="false"
# Default number of results to return from a vector db lookup (Optional)
# Default is 5
# MAX_QUERY_TOP_K=5
# ================ Tracing ================
# For Langsmith tracing (Optional)
ENABLE_LANGSMITH_TRACING="false"
# LANGCHAIN_API_KEY=<your_langchain_api_key>
# LANGCHAIN_ENDPOINT="https://api.smith.langchain.com"
# LANGCHAIN_TRACING_V2=true
# LANGSMITH_PROJECT_NAME="<project_name_for_langsmith>"
# For Langfuse tracing (Optional)
ENABLE_LANGFUSE_TRACING="false"
# LANGFUSE_SECRET_KEY="sk-lf-..."
# LANGFUSE_PUBLIC_KEY="pk-lf-..."
# LANGFUSE_HOST="http://localhost:3000"
# For Phoenix tracing (Optional)
ENABLE_PHOENIX_TRACING="false"
# Auth
JWT_ISSUER="personaflow"
JWT_ALGORITHM="HS256"
TOKEN_EXPIRY_HOURS=24
# For setting up authentication, see: docs/auth_guide.md
AUTH_SECRET_KEY=<See auth.guide.md on how to generate a secure one>
# Required for specifying Redirect URI
FRONTEND_HOSTNAME=http://localhost:3000
# Google OAuth
GOOGLE_CLIENT_ID=<GOOGLE_CLIENT_ID>
GOOGLE_CLIENT_SECRET=<GOOGLE_CLIENT_SECRET>
# OpenID Connect
OIDC_CLIENT_ID=<OIDC_CLIENT_ID>
OIDC_CLIENT_SECRET=<OIDC_CLIENT_SECRET>
OIDC_WELL_KNOWN_ENDPOINT=<OIDC_WELL_KNOWN_ENDPOINT>
# For Anthropic Claude API
ANTHROPIC_MODEL_NAME="claude-3-5-sonnet-20240620"
ANTHROPIC_API_KEY=""