Fix: Graceful fallback on HF API timeout/error instead of 502 Bad Gateway

Made adapyters work independently

Resolved api errors

Made some more change to UI/UX

Made the landing HTML more detailed

Made landing page more responsive and visual friendly

feat: Integrate HF Inference API for real model evaluation

Resrted training from scratch

Changed training parameters

Changed dataset configurations for better training

Resolved errors in training by resolving endpoints

feat: add cyberpunk SOC dashboard frontend at root /

fix: port 7860 for HF Spaces + lazy dataset streaming for fast startup

REsolved errors and made deployment lightweight

resolved training errors

Making files ready for Training

Added openenv.yaml and updated REDAME with placeholders for post completion video and documentation according to judging criteria

FastAPI server with CORS, /info endpoint, Docker Compose multi-service stack

Vectorized OpenEnv state machine with 32-process parallelization and telemetry

Dual-grader reward system with independent sub-functions and anti-hacking verifier

Created streaming payload generator with async prefetch and dynamic obfuscation

Implemented adaptive curriculum scheduler and dynamic embedding generation

Defined Pydantic V2 contracts for observations, actions, and rewards