-
fix: production hardening (mlflow optional + 403 forbidden bypass)
Biswajeet Arukha
-
v2.0: Production hardening + Premium betting dashboard
Biswajeet Arukha
- NEW: Betting analytics engine (1xbet/4rabet-style odds, prop bets, over/under)
- NEW: Centralized config.py with Pydantic Settings
- NEW: Premium unified dashboard with glassmorphism design
- FIX: JSON.parse bug in WebSocket handler
- FIX: train_efficient.py import paths and 30-match limit removed
- FIX: MLflow context scope covering all training
- FIX: Graceful degradation for Redis/DB/ML
- FIX: Lazy agent swarm initialization
- FIX: WebSocket exponential backoff reconnection
- CLEAN: Removed ~500MB unused deps from requirements
- CLEAN: Slimmed Dockerfile (removed playwright/opencv/tesseract)
- ADD: .env.example with all config vars documented
-
Critical bug fixes: TransformerModel, HTTP status codes, hardcoded paths, lifespan, CORS, labels, WebSocket cleanup, bowling team extraction
Biswajeet Arukha
-
Deep Prediction Upgrade: Transformer, MLflow, and MiroFish Multi-Agent Simulation
Biswajeet Arukha
-
Initial commit (v1.0 Cleaned)
Biswajeet Arukha
