-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathhackdeepwiki.env.example
More file actions
20 lines (19 loc) · 919 Bytes
/
Copy pathhackdeepwiki.env.example
File metadata and controls
20 lines (19 loc) · 919 Bytes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
# Copy this file to hackdeepwiki.env only when you want persistent overrides.
# Command-line flags passed to hackdeepwiki.sh take precedence.
OLLAMA_ENDPOINT=http://127.0.0.1:11434
# Optional: leave blank to discover all models and select defaults from Ollama.
OLLAMA_MODEL=
OLLAMA_EMBED_MODEL=
# Número de textos por petición de embeddings (reduce si el servidor tiene poca RAM).
OLLAMA_EMBED_BATCH_SIZE=32
# Segundos máximos por tanda de embeddings. La generación de texto no tiene timeout.
OLLAMA_REQUEST_TIMEOUT=1800
# El endpoint remoto puede tardar en responder mientras procesa otro modelo.
OLLAMA_HEALTH_TIMEOUT=60
# Recommended to avoid GitHub's anonymous API rate limit.
GITHUB_TOKEN=
HACKDEEPWIKI_API_PORT=8001
HACKDEEPWIKI_PROJECT_NAME=hackdeepwiki-local
# auto, host or bridge. "auto" selects host on native Linux and bridge on
# Docker Desktop, WSL, macOS and Windows.
HACKDEEPWIKI_NETWORK_MODE=auto