Spaces:
Running
Running
Commit History
-
feat: cache reranker snapshot at build time via huggingface_hub
AkshayMdownload_model.py now downloads BAAI/bge-reranker-base raw weight files during build using snapshot_download() — no model load into RAM, no OOM risk. Dockerfile sets HF_HOME to the cached dir so CrossEncoder finds it at runtime without network download.e61df26
-
perf: revert to all-MiniLM-L6-v2, query embed cache, fire-and-forget Cloudinary, pre-warm reranker, reduce top-k
AkshayM1476572
-
fix: skip reranker build-time cache (1.1 GB OOMs builder), lazy-load at runtime
AkshayMa42957b -
fix: free memory between model downloads, graceful fallback if reranker OOMs
AkshayM41f063a -
perf: upgrade embedding to all-mpnet-base-v2, add cross-encoder reranker, LRU cache, enable PII by default
AkshayM9147527
-
production hardening: circuit breakers, graceful degradation, health checks, parallel uploads, HF Spaces deploy config
AkshayMfc2fe9f
-
fix upload pipeline: await add_document, pre-load model eagerly, add verbose UPLOAD/VectorStore logging; remove unused deployment configs
AkshayM90c440b
-
fix models.py: match Neon schema varchar sizes, add email unique constraint
AkshayM167c070