From ee387abcd4f23740c38b800b36e06fc4c7a83af0 Mon Sep 17 00:00:00 2001 From: m17hr1l Date: Wed, 20 May 2026 23:55:47 +0200 Subject: [PATCH] stage-21: swap inference server to psyc-v5 adapter MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit v5 trained on 598 ex/task (20× v4's 30), with --defang-frac 0.5 over the new ThreatFox + MalwareBazaar + OTX corpus. Final train_loss 0.3225 vs v4's 0.7397 (56% reduction), 60m20s wall clock on a 3090. Live eval before swap: - severity (botnet, ONLINE): v4 high / v5 high — tied, both correct - ioc_extraction with defanged input (hxxps://, [.], (.), [dot]): v4 kept hxxps:// in output (failed canonicalization) v5 returned canonical https:// — defang training paid off - ioc_extraction on real OTX-style prose (never trained on this shape): v5 cleanly extracted 2 domains + 1 IP + 1 SHA256 + 1 CVE Cockpit /api/inference-status confirms the swap: {"online":true,"adapter":"/data/adapters/psyc-v5/final"} Co-Authored-By: Claude Opus 4.7 --- docker-compose.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docker-compose.yml b/docker-compose.yml index 7476d03..d034f38 100644 --- a/docker-compose.yml +++ b/docker-compose.yml @@ -51,7 +51,7 @@ services: # `--profile gpu`. Uses the psyc-trainer image (built from Dockerfile.train). inference: image: psyc-trainer - command: ["/scripts/serve_model.py", "--adapter", "/data/adapters/psyc-v4/final", "--host", "0.0.0.0", "--port", "8771"] + command: ["/scripts/serve_model.py", "--adapter", "/data/adapters/psyc-v5/final", "--host", "0.0.0.0", "--port", "8771"] volumes: - ./data:/data - ./scripts:/scripts