OFPBadWord / config /config.yaml
BladeSzaSza's picture
added dataset
589ab6a
# OFP Bad Word Sentinel Configuration
sentinel:
# Sentinel identification
speaker_uri: 'tag:sentinel.ofpbadword.service,2025:badword-01'
service_url: 'https://bladeszasza-ofpbadword.hf.space/ofp'
# Convener endpoints (update with actual convener details)
convener_uri: 'tag:convener.service,2025:default'
convener_url: 'https://convener-service.com/ofp'
profanity:
# Use default better-profanity word list
use_default: true
# Path to custom word list (optional)
# One word per line, lines starting with # are comments
custom_wordlist: 'config/wordlist.txt'
# Kaggle dataset configuration
use_kaggle_dataset: true
kaggle_dataset_path: 'data/'
# Languages to load from Kaggle dataset
# Available: ar, cs, da, de, en, eo, es, fa, fi, fil, fr, fr-CA-u-sd-caqc,
# hi, hu, it, ja, kab, ko, nl, no, pl, pt, ru, sv, th, tlh, tr, zh
# Use 'all' to enable all languages (140KB total, includes Hungarian)
languages:
- all
# Whitelist words that should not be flagged (false positives)
whitelist:
- scunthorpe
- arsenal
- pussycat
- classic
# Alert on these severity levels
alert_on_severity:
- low
- medium
- high
monitoring:
# Monitoring check interval (seconds)
check_interval: 30
# Auto-start monitoring on launch
auto_start: true
# Maximum activity log entries to keep
max_log_entries: 100
dashboard:
# Auto-refresh interval (seconds)
refresh_interval: 5
# Show test panel by default
show_test_panel: true
# Theme
theme: 'soft' # Options: soft, glass, monochrome