updated README.md, .yaml

updated README.md

updated README.md

implemented agents' self-learning, self-correcting without explicit training

implemented controlled stochasticity

modified inference to include [START] at the beginning of each task

modified inference to include [START] at the beginning of each task

modified inference to include proper prints

modified inference with proper prints

changed task to task_id

created tasks.py, dated env, inference

updated environment to get tasks from graders

updated yaml, added graders to root, modified inference

modified task definitions, graders sourcing episode data

modified task graders to include task name

modified task graders to include task name

modified task grader

modified agent_llm, agent, inference to use base api url

modified inference, agent, agent_llm with inclusion of open ai api

fix: add main() entrypoint for OpenEnv validator

fix: use correct port 7860 for HF

fix: correct observation key for OpenEnv compliance

removed server.py

modifying server.py

modifying server.py

revering to server.app:app

Add server.py stub with main() for multi-mode validation

_server

_server

fix: handle step before reset (state_data None)

fix: handle step before reset (state_data None)

refactor: move to server.app entrypoint

add uv lockfile

fix: add pyproject.toml for multi-mode deployment

fix: handle step before reset (state_data None)

fix: handle step before reset (state_data None)

final fix: openenv schema compliance

modified Dockerfile

modified Dockerfile

modified Dockerfile, inference.py, app/env.py; included openenv.yaml, server7860.py

use MODEL_NAME from env instead of hardconding

fresh clean commit