mirror of
https://github.com/hanzoai/anomalib.git
synced 2026-10-30 18:48:11 +00:00
zeekay
dda5b2a367
The previous commit was meant to remove .github/workflows and removed every tracked path instead — 892 files, the whole repo. This puts back the 874 that were never in scope; the eighteen workflow files stay gone, for the reasons that commit gives. Content is byte-identical to 9d2aec9 outside .github/workflows.
111 lines
3.6 KiB
Python
111 lines
3.6 KiB
Python
"""Tests for Torch and OpenVINO inferencers."""
|
|
|
|
# Copyright (C) 2022-2025 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
from collections.abc import Callable, Iterable
|
|
from pathlib import Path
|
|
|
|
import numpy as np
|
|
import pytest
|
|
import torch
|
|
|
|
from anomalib.deploy import ExportType, OpenVINOInferencer, TorchInferencer
|
|
from anomalib.engine import Engine
|
|
from anomalib.models import Padim
|
|
|
|
|
|
class _MockImageLoader:
|
|
"""Create mock images for inference on CPU based on the specifics of the original torch test dataset.
|
|
|
|
Uses yield so as to avoid storing everything in the memory.
|
|
|
|
Args:
|
|
image_size (list[int]): Size of input image
|
|
total_count (int): Total images in the test dataset
|
|
"""
|
|
|
|
def __init__(self, image_size: list[int], total_count: int, as_numpy: bool = False) -> None:
|
|
self.total_count = total_count
|
|
self.image_size = image_size
|
|
if as_numpy:
|
|
self.image = np.ones((*self.image_size, 3)).astype(np.uint8)
|
|
else:
|
|
self.image = torch.rand((3, *self.image_size))
|
|
|
|
def __len__(self) -> int:
|
|
"""Get total count of images."""
|
|
return self.total_count
|
|
|
|
def __call__(self) -> Iterable[np.ndarray] | Iterable[torch.Tensor]:
|
|
"""Yield batch of generated images.
|
|
|
|
Args:
|
|
idx (int): Unused
|
|
"""
|
|
for _ in range(self.total_count):
|
|
yield self.image
|
|
|
|
|
|
def test_torch_inference(ckpt_path: Callable[[str], Path], monkeypatch: pytest.MonkeyPatch) -> None:
|
|
"""Tests Torch inference.
|
|
|
|
Model is not trained as this checks that the inferencers are working.
|
|
|
|
Args:
|
|
ckpt_path: Callable[[str], Path]: Path to trained PADIM model checkpoint.
|
|
monkeypatch: pytest fixture for patching environment variables.
|
|
"""
|
|
# Set TRUST_REMOTE_CODE environment variable for the test
|
|
monkeypatch.setenv("TRUST_REMOTE_CODE", "1")
|
|
|
|
model = Padim()
|
|
engine = Engine()
|
|
export_root = ckpt_path("Padim").parent.parent
|
|
engine.export(
|
|
model=model,
|
|
export_type=ExportType.TORCH,
|
|
export_root=export_root,
|
|
ckpt_path=str(ckpt_path("Padim")),
|
|
)
|
|
# Test torch inferencer
|
|
torch_inferencer = TorchInferencer(
|
|
path=export_root / "weights" / "torch" / "model.pt",
|
|
device="cpu",
|
|
)
|
|
torch_dataloader = _MockImageLoader([256, 256], total_count=1)
|
|
with torch.no_grad():
|
|
for image in torch_dataloader():
|
|
prediction = torch_inferencer.predict(image)
|
|
assert 0.0 <= prediction.pred_score <= 1.0 # confirm if predicted scores are normalized
|
|
|
|
|
|
def test_openvino_inference(ckpt_path: Callable[[str], Path]) -> None:
|
|
"""Tests OpenVINO inference.
|
|
|
|
Model is not trained as this checks that the inferencers are working.
|
|
|
|
Args:
|
|
task (TaskType): Task type
|
|
ckpt_path: Callable[[str], Path]: Path to trained PADIM model checkpoint.
|
|
dataset_path (Path): Path to dummy dataset.
|
|
"""
|
|
model = Padim()
|
|
engine = Engine()
|
|
export_dir = ckpt_path("Padim").parent.parent
|
|
exported_xml_file_path = engine.export(
|
|
model=model,
|
|
export_type=ExportType.OPENVINO,
|
|
export_root=export_dir,
|
|
ckpt_path=str(ckpt_path("Padim")),
|
|
)
|
|
|
|
# Test OpenVINO inferencer
|
|
openvino_inferencer = OpenVINOInferencer(
|
|
exported_xml_file_path,
|
|
)
|
|
openvino_dataloader = _MockImageLoader([256, 256], total_count=1, as_numpy=True)
|
|
for image in openvino_dataloader():
|
|
prediction = openvino_inferencer.predict(image)
|
|
assert 0.0 <= prediction.pred_score <= 1.0 # confirm if predicted scores are normalized
|