Keep mx_obs after episode done, prevent reset_multi first error

Fix API_BASE_URL to correct HF inference /v1 endpoint

Fix HF Inference API URL to include model path /v1

Fix API_BASE_URL - remove duplicate /v1

Switch to HF Inference API + zephyr-7b (no provider needed)

Switch to Phi-3.5-mini-instruct (free HF router)

Switch to Llama-3.2-3B-Instruct (supported by HF router)

Add timeout+stderr logging to LLM calls, reduce max_tokens to 64

Log to stderr + gunicorn capture-output for HF diagnosis

Cache bust + debug env print for HF token diagnosis

Fix checkpoint path + default model to Qwen2.5-0.5B for HF API

Run 3: Qwen2.5-0.5B GRPO 500 steps | reward 0.96→1.75 | conflict 0.375→0.1875 | dashboard chart | agents wired

add /game route for game mode

fix: clean up auto_multi, set mx_obs=None on done

remove filter: let LLM produce natural conflicts for LLM vs LoRA comparison

debug max_steps + safety counter in runAllMulti

debug: log filtered proposals before step_multiagent

fix: enforce sensor ownership server-side, drop cross-agent proposals

fix: restrict LLM proposals to agent-type sensors, prevent S1 conflicts

add lora checkpoints and training updates

fix: rewards, conflict detection, training comparison panel

Fix: task difficulty counts, dashboard legend modal, training metrics, reward curve, /metrics/history endpoint

fix: implement Phase 1 core fixes (capability matrix, agent wiring, multi-agent inference, robust observation handling)

Merge branch 'main' into chathu01

feat: integrate trained LoRA adapters into inference and server, enabling real multi-agent learning deployment

Fix all audit issues: conflict schema, curriculum wiring, density_factor, eval path, dead files removed

Merge pull request #8 from smritis21/chathu01

feat: activate full multi-agent pipeline with negotiation, conflict handling, and GRPO training

feat: activate full multi-agent pipeline with negotiation, conflict handling, and GRPO training

Fix LLM multi-agent proposals: agent-type sensor filtering, prompt improvements, debug logging

Setup project environment, fixed Python version issues, added venv, installed dependencies, and configured server execution

error fixed

submission ready

feat: implement dashboard UI and integrate uv dependency management

feat: serve dashboard at / as well as /ui

feat: LLM agent, Leaflet map, batch step, reward fix, inference all tasks

improve: better prompt + fixed grader normalization, score 0.77

Round 1 submission — SentinelEnv OpenEnv