Upload models/router_lightgbm.pkl with huggingface_hub

Upload models/router_mlp.pkl with huggingface_hub

Upload models/router_logistic.pkl with huggingface_hub

Upload models/router_nomic_lightgbm.pkl with huggingface_hub

Trigger rebuild for BGE/MiniLM (6-row)

Deploy 9-Row Benchmark (via API)

Deploy 9-Row Benchmark (via API)

Deploy 9-Row Benchmark (via API)

Fix: Update router.py and vector_db.py to support multi-cluster prediction

Upload models/router_mlp.pkl with huggingface_hub

Upload models/router_lightgbm.pkl with huggingface_hub

Upload models/router_logistic.pkl with huggingface_hub

Fix: Update UI, Shards (16), and Router Logic

Fix: Use MiniLM-L6-v2 to match Qdrant vector dimension (384)

Fix: Pin Gradio version and simplify interactions

Fix: Disable queue completely to resolve ProtocolError

Fix: Replace streaming with chained events to resolve ProtocolError

Fix: Disable queue and add yield delay to resolve ProtocolError

Fix: Add try-except block to catch runtime errors and fix CSS

UI: Fix header, logo, layout and add debug logs

Fix: Add get_embedding function and caching

UI: Complete overhaul to match benchmark HTML reference

UI: Add logo, queue, and reference benchmarks

Fix: Remove deprecated style arg from gr.Row

UI: Single screen redesign with benchmarking table

Upload models/router_v1.pkl with huggingface_hub

Fix: Prevent data wipe in initialize and set correct port

UI: Major overhaul with structured layout and premium styling

Fix: Add HF Space Metadata

Deploy dashVectorspace v1 (Full)

Upload models/router_v1.pkl with huggingface_hub

initial commit