ChemaVXandClaude Opus 4.8
9653898366
Build & Deploy ResearchOwl / build-and-push (push) Successful in 14s
Generación 3B→7B y embeddings a bge-m3 en README, CLAUDE.md, .env.example y k8s/deployment.yaml. Corregido el default buggy de config.py: ollama_embed_model era qwen2.5:3b (un modelo de chat) → bge-m3. Añadido OLLAMA_EMBED_MODEL donde faltaba. Nota del host público de ollama ahora tras Authentik. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
31 lines
817 B
Bash
31 lines
817 B
Bash
# ResearchOwl — Environment Variables
|
|
# Copy to .env and fill in values
|
|
|
|
# Required
|
|
TELEGRAM_BOT_TOKEN=your_bot_token_here
|
|
TELEGRAM_ALLOWED_USERS=123456789 # your Telegram user ID
|
|
|
|
# Ollama (default points to your existing instance)
|
|
OLLAMA_URL=http://ollama.chemavx.xyz
|
|
OLLAMA_MODEL=qwen2.5:7b
|
|
OLLAMA_EMBED_MODEL=bge-m3
|
|
|
|
# Claude fallback (optional, only for premium generation)
|
|
# ANTHROPIC_API_KEY=sk-ant-...
|
|
# CLAUDE_MODEL=claude-haiku-4-5
|
|
|
|
# Storage
|
|
DB_PATH=/data/researchowl.db
|
|
|
|
# Scraping tuning
|
|
MAX_DEPTH=3 # how deep to follow links (1-5)
|
|
MAX_SOURCES=150 # hard cap on total sources
|
|
MAX_PAGES_PER_SEARCH=5
|
|
REQUEST_DELAY=1.0 # seconds between requests (be polite)
|
|
MIN_CONTENT_LENGTH=200
|
|
|
|
# Processing
|
|
CHUNK_SIZE=800
|
|
CHUNK_OVERLAP=100
|
|
QUALITY_THRESHOLD=0.4 # 0-1, lower = more permissive
|