7 Commits

Author	SHA1	Message	Date
Greg Hendrickson	c6fd06369f	fix: Upgrade torch via pip instead of new base image ... - Use known-working base image (2.4.0) - Upgrade torch to 2.6+ via pip - Previous base image was too slow to pull	2026-02-03 06:45:27 +00:00
Greg Hendrickson	9b66b5fd14	fix: Upgrade to torch 2.6+ for CVE-2025-32434 compliance ... - Use runpod/pytorch:1.0.3-cu1290-torch260-ubuntu2204 base image - Torch 2.6.0 required by transformers for secure model loading - CUDA 12.9 compatible	2026-02-03 06:02:38 +00:00
Greg Hendrickson	7a7b737cc0	fix: Pin transformers version for torch compatibility ... - Force reinstall torch 2.4.0 in Dockerfile - Pin transformers==4.47.1 (known working with bitsandbytes) - Fix set_submodule AttributeError	2026-02-03 05:20:31 +00:00
Greg Hendrickson	67a1095100	feat: Add 176K real DrugBank DDI samples with drug names ... - Downloaded 191K DDI pairs from TDC DrugBank - Fetched 1,634 drug names from PubChem API (96% hit rate) - Created complete training dataset with: - Real drug names (not just IDs) - 86 interaction type descriptions - Severity labels (minor/moderate/major/contraindicated) - Bundled 34MB data file in Docker image - Handler loads real data instead of curated samples	2026-02-03 04:34:54 +00:00
Greg Hendrickson	b086239c52	Fix Python/transformers version compatibility	2026-02-03 01:11:20 +00:00
Greg Hendrickson	222be0fb68	Use Tailscale endpoints, add RunPod Docker build files	2026-02-03 00:23:16 +00:00
Greg Hendrickson	9ca3d6c195	Add DDI training pipeline with RunPod serverless GPU support	2026-02-02 23:56:05 +00:00