Spaces:
Running
Running
File size: 1,553 Bytes
590a604 ee1a8a3 1fbc47b a18e93d 1fbc47b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 |
"""
FastAPI dependency providers for LexiMind.
Manages lazy initialization and caching of the inference pipeline.
Author: Oliver Perrin
Date: December 2025
"""
from __future__ import annotations
from functools import lru_cache
from pathlib import Path
from fastapi import HTTPException, status
from ..utils.logging import get_logger
logger = get_logger(__name__)
from ..inference.factory import create_inference_pipeline
from ..inference.pipeline import InferencePipeline
@lru_cache(maxsize=1)
def get_pipeline() -> InferencePipeline:
"""Lazily construct and cache the inference pipeline for the API."""
checkpoint = Path("checkpoints/best.pt")
labels = Path("artifacts/labels.json")
model_config = Path("configs/model/base.yaml")
try:
pipeline, _ = create_inference_pipeline(
checkpoint_path=checkpoint,
labels_path=labels,
model_config_path=model_config,
)
except FileNotFoundError as exc:
logger.exception("Pipeline initialization failed: missing artifact")
raise HTTPException(
status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
detail="Service temporarily unavailable",
) from exc
except Exception as exc: # noqa: BLE001 - surface initialization issues to the caller
logger.exception("Pipeline initialization failed")
raise HTTPException(
status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
detail="Service temporarily unavailable",
) from exc
return pipeline
|