cabele/spamBERT
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
2 commits 1 branch 0 tags 43 KiB
Commit graph

2 commits

This branch
This branch
All branches
Author SHA1 Message Date
Carsten Abele
f05320a8cb Add project README and reason/quote fields to classifier response 
- README.md: full project overview with setup, training, API, and RSpamd integration docs
- server.py: add reason (human-readable explanation) and quote (suspicious snippet) to response
- spamllm.lua: pass reason and quote through to RSpamd symbol description for logs/UI

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-19 22:32:54 +01:00
Carsten Abele
38efd20b4d Initial commit: SpamLLM - DistilBERT spam classifier for RSpamd 
Multilingual spam classifier (DE/EN) with language detection.
Non-DE/EN mails receive an additional spam score bonus.

- train.py: Fine-tune distilbert-base-multilingual-cased on spam/ham data
- server.py: FastAPI service with langdetect integration
- rspamd/: Lua plugin and config for RSpamd integration
- export_rspamd_data.py: Export Maildir folders to CSV training data
- test_classify.py: Local model validation with DE/EN/foreign test cases

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-19 22:27:05 +01:00