Spaces:
Sleeping
Sleeping
Download server/app.py from SamanPandey/meta: direct link, hf CLI and curl.
- Browser
- Download file 2.83 kB
-
https://huggingface.co/spaces/SamanPandey/meta/resolve/main/server/app.py
- Command line
-
hf download hf://spaces/SamanPandey/meta/server/app.py
-
curl -L -o app.py https://huggingface.co/spaces/SamanPandey/meta/resolve/main/server/app.py
2.83 kB
| # server/app.py | |
| import os, sys, uuid | |
| sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..")) | |
| from fastapi import FastAPI | |
| from fastapi.responses import RedirectResponse, JSONResponse | |
| from pydantic import BaseModel | |
| from typing import Optional | |
| import uvicorn | |
| from server.environment import PunjabPrecisionIrrigationEnvironment | |
| from server.graders import grade_task | |
| from models import IrrigationAction | |
| app = FastAPI(title="Punjab Precision Irrigation") | |
| ENV_STORE = {} | |
| class ResetRequest(BaseModel): | |
| task_name: str = "easy" | |
| seed: Optional[int] = None | |
| episode_id: Optional[str] = None | |
| class StepRequest(BaseModel): | |
| episode_id: str | |
| action: IrrigationAction | |
| def _obs_to_dict(obs, reward=None, done=False): | |
| return {"observation": obs.model_dump(), "reward": reward, "done": done, "info": {}} | |
| def root(): return RedirectResponse(url="/docs") | |
| def reset(req: ResetRequest = None): | |
| if req is None: req = ResetRequest() | |
| env = PunjabPrecisionIrrigationEnvironment() | |
| obs = env.reset(task_name=req.task_name, seed=req.seed, episode_id=req.episode_id) | |
| ENV_STORE[obs.episode_id] = {"env": env, "rewards": [], "steps": 0} | |
| return _obs_to_dict(obs) | |
| def step(req: StepRequest): | |
| entry = ENV_STORE.get(req.episode_id) | |
| if not entry: return JSONResponse(status_code=404, content={"error": "Not found"}) | |
| obs = entry["env"].step(req.action) | |
| entry["rewards"].append(obs.reward) | |
| entry["steps"] += 1 | |
| if obs.done: | |
| res = _obs_to_dict(obs, obs.reward, True) | |
| res["score"] = grade_task(entry["env"].state.task_name, res["observation"], entry["rewards"], entry["steps"]) | |
| ENV_STORE.pop(req.episode_id) | |
| return res | |
| return _obs_to_dict(obs, obs.reward, False) | |
| def state(episode_id: str): | |
| entry = ENV_STORE.get(episode_id) | |
| if not entry: return JSONResponse(status_code=404, content={"error": "Not found"}) | |
| return entry["env"].state.model_dump() | |
| def health(): return {"status": "healthy"} | |
| def metadata(): | |
| return { | |
| "name": "punjab-precision-irrigation", | |
| "description": "Punjab Rice Farm Precision Irrigation", | |
| "tasks": [ | |
| "easy", "easy_1", "easy_2", "easy_3", | |
| "medium", "medium_1", "medium_2", "medium_3", | |
| "hard", "hard_1", "hard_2", "hard_3" | |
| ] | |
| } | |
| def schema(): | |
| from models import IrrigationAction, IrrigationObservation, IrrigationState | |
| return { | |
| "action": IrrigationAction.model_json_schema(), | |
| "observation": IrrigationObservation.model_json_schema(), | |
| "state": IrrigationState.model_json_schema() | |
| } | |
| def main(): | |
| uvicorn.run(app, host="0.0.0.0", port=7860) | |
| if __name__ == "__main__": | |
| main() | |