Chore/fix tests in vscode

.vscode/settings.json

@@ -1,8 +1,6 @@
 {
-    "python.testing.pytestArgs": [
-        ".",
-        "--ignore=test_data"
-    ],
+    "python.testing.pytestArgs": [],
+    "python.testing.cwd": "${workspaceFolder}",
     "python.testing.unittestEnabled": false,
     "python.testing.pytestEnabled": true,
     "python.analysis.extraPaths": [

libs/datapipe-cvat/tests/test_simple_project_example.py

@@ -14,8 +14,7 @@
 from datapipe_cvat.cvat_step import create_cvat_client
 from PIL import Image
 
-from test_cvat_integration import _require_cvat
-
+from .test_cvat_integration import _require_cvat
 
 pytestmark = pytest.mark.cvat
 

libs/datapipe-label-studio/tests/test_pipeline.py

@@ -10,17 +10,17 @@
 from datapipe.step.datatable_transform import DatatableTransformStep
 from datapipe.store.database import TableStoreDB
 from datapipe.types import data_to_index
-from label_studio_sdk import LabelStudio
-from pytest_cases import parametrize, parametrize_with_cases
-from sqlalchemy.sql.schema import Column
-from sqlalchemy.sql.sqltypes import JSON, String
-
 from datapipe_label_studio.sdk_utils import get_project_by_title
 from datapipe_label_studio.upload_predictions_pipeline import (
     LabelStudioUploadPredictions,
 )
 from datapipe_label_studio.upload_tasks_pipeline import LabelStudioUploadTasks
-from tests.ls_test_helpers import (
+from label_studio_sdk import LabelStudio
+from pytest_cases import parametrize, parametrize_with_cases
+from sqlalchemy.sql.schema import Column
+from sqlalchemy.sql.sqltypes import JSON, String
+
+from .ls_test_helpers import (
     DELETE_UNANNOTATED_TASKS_ONLY_ON_UPDATE,
     INCLUDE_PARAMS,
     INCLUDE_PREDICTIONS,
@@ -30,7 +30,7 @@
     convert_to_ls_input_data,
     wrapped_partial,
 )
-from tests.util import get_project_id, get_project_tasks, wait_until_label_studio_is_up
+from .util import get_project_id, get_project_tasks, wait_until_label_studio_is_up
 
 
 def gen_data_df():

libs/datapipe-label-studio/tests/test_pipeline_many_projects.py

@@ -9,11 +9,6 @@
 from datapipe.step.batch_transform import BatchTransform
 from datapipe.step.datatable_transform import DatatableTransformStep
 from datapipe.store.database import TableStoreDB
-from label_studio_sdk import LabelStudio
-from pytest_cases import parametrize, parametrize_with_cases
-from sqlalchemy.sql.schema import Column
-from sqlalchemy.sql.sqltypes import JSON, String
-
 from datapipe_label_studio.create_projects_step import CreateLabelStudioProjects
 from datapipe_label_studio.sdk_utils import get_project_by_title
 from datapipe_label_studio.upload_predictions_pipeline import (
@@ -22,7 +17,12 @@
 from datapipe_label_studio.upload_tasks_pipeline import (
     LabelStudioUploadTasksToProjects,
 )
-from tests.ls_test_helpers import (
+from label_studio_sdk import LabelStudio
+from pytest_cases import parametrize, parametrize_with_cases
+from sqlalchemy.sql.schema import Column
+from sqlalchemy.sql.sqltypes import JSON, String
+
+from .ls_test_helpers import (
     DELETE_UNANNOTATED_TASKS_ONLY_ON_UPDATE,
     INCLUDE_PARAMS,
     INCLUDE_PREDICTIONS,
@@ -32,7 +32,7 @@
     convert_to_ls_input_data,
     wrapped_partial,
 )
-from tests.util import get_project_id, get_project_tasks, wait_until_label_studio_is_up
+from .util import get_project_id, get_project_tasks, wait_until_label_studio_is_up
 
 
 def gen_ls_project_setting():

libs/datapipe-ml/tests/conftest.py

@@ -2,7 +2,7 @@
 
 os.environ["SQLALCHEMY_WARN_20"] = "1"
 
-from tests.helpers.test_env import load_test_env
+from .helpers.test_env import load_test_env
 
 load_test_env()
 
@@ -13,8 +13,8 @@
 from sqlalchemy import Column, create_engine, text
 from sqlalchemy.sql.sqltypes import JSON, String
 
-from tests.fixtures.smoke_data import SmokeDataset, make_smoke_dataset
-from tests.helpers.dbconn import get_sqlite_dbconnstr
+from .fixtures.smoke_data import SmokeDataset, make_smoke_dataset
+from .helpers.dbconn import get_sqlite_dbconnstr
 
 
 @pytest.fixture(

libs/datapipe-ml/tests/helpers/checkpoint_kill9.py

@@ -6,7 +6,7 @@
 import time
 from pathlib import Path
 
-from tests.helpers.checkpoint_fixtures import write_corrupt_zip_checkpoint, write_valid_zip_checkpoint
+from .checkpoint_fixtures import write_corrupt_zip_checkpoint, write_valid_zip_checkpoint
 
 
 def _child_atomic_save_before_replace(checkpoint_path: str, ready_queue: mp.Queue) -> None:

libs/datapipe-ml/tests/helpers/cloud_smoke.py

@@ -5,7 +5,7 @@
 
 import pytest
 
-from tests.helpers.training_smoke import (
+from .training_smoke import (
     Workdir,
     classification_freeze_step,
     classification_train_step,

libs/datapipe-ml/tests/helpers/failure_injection.py

@@ -7,22 +7,23 @@
 from typing import TYPE_CHECKING
 
 import pytest
-
 from datapipe_ml.frameworks.yolo.checkpoint_sync import max_completed_epoch_from_run_dir
 from datapipe_ml.training.sync import manifest_path_for_run, read_checkpoint_manifest
 
-from tests.helpers.failure_injection_bootstrap import (
+from .failure_injection_bootstrap import (
     FAIL_AFTER_EPOCH_ENV,
     FAIL_MODE_ENV,
     checkpoint_for_epoch_exists,
     configured_fail_after_epoch,
     configured_fail_mode,
-    install_training_failure_hooks as _install_training_failure_hooks,
     install_training_failure_hooks_direct,
     maybe_fail_after_epoch,
     run_dir_for_pipe_death_poll,
     run_training_with_failure_hooks,
 )
+from .failure_injection_bootstrap import (
+    install_training_failure_hooks as _install_training_failure_hooks,
+)
 
 if TYPE_CHECKING:
     from _pytest.monkeypatch import MonkeyPatch
@@ -68,9 +69,7 @@ def _spawn_with_pipe_death(target, *args):  # noqa: ANN001
             except queue.Empty:
                 if not p.is_alive():
                     p.join()
-                    raise RuntimeError(
-                        f"Training subprocess exited before returning a result. exitcode={p.exitcode}"
-                    )
+                    raise RuntimeError(f"Training subprocess exited before returning a result. exitcode={p.exitcode}")
                 run_dir = run_dir_for_pipe_death_poll()
                 if fail_after is not None and run_dir and _pipe_death_checkpoint_ready(run_dir, fail_after):
                     if p.is_alive():

libs/datapipe-ml/tests/helpers/recovery_pipe_runner.py

@@ -28,15 +28,16 @@ def main(argv: list[str] | None = None) -> int:
     if args.phase == "train":
         _install_short_lease_status_manager()
 
-    from tests.helpers.failure_injection import install_pipe_death_hooks_direct
-    from tests.helpers.failure_injection_bootstrap import WORK_DIR_ENV
-    from tests.helpers.training_recovery import (
+    from datapipe.compute import Pipeline, build_compute, run_steps
+
+    from .failure_injection import install_pipe_death_hooks_direct
+    from .failure_injection_bootstrap import WORK_DIR_ENV
+    from .training_recovery import (
         configure_recovery_steps,
         make_recovery_runtime,
         recovery_case_by_id,
     )
-    from datapipe.compute import Pipeline, build_compute, run_steps
-    from tests.helpers.training_smoke import run_pipeline
+    from .training_smoke import run_pipeline
 
     os.environ[WORK_DIR_ENV] = str(workdir)
 

libs/datapipe-ml/tests/helpers/training_recovery.py

@@ -13,16 +13,16 @@
 import pytest
 from datapipe.compute import Pipeline, PipelineStep, build_compute
 from datapipe.types import IndexDF
-
+from datapipe_ml.frameworks.yolo.checkpoint_sync import max_completed_epoch_from_run_dir
 from datapipe_ml.training.runs import active_lease
 from datapipe_ml.training.specs import TrainingResumeConfig, TrainingSyncConfig
-from datapipe_ml.frameworks.yolo.checkpoint_sync import max_completed_epoch_from_run_dir
 from datapipe_ml.training.sync import (
     manifest_path_for_run,
     read_checkpoint_manifest,
     verify_manifest_checkpoint,
 )
-from tests.helpers.training_smoke import (
+
+from .training_smoke import (
     SmokeRuntime,
     Workdir,
     detection_freeze_step,
@@ -230,7 +230,7 @@ def real_recovery_cases() -> list:
 
 def recovery_case_by_id(case_id: str) -> RealRecoveryCase:
     if case_id in TENSORFLOW_RECOVERY_CASE_IDS:
-        from tests.helpers.training_recovery_tensorflow import recovery_tensorflow_case_by_id
+        from .training_recovery_tensorflow import recovery_tensorflow_case_by_id
 
         return recovery_tensorflow_case_by_id(case_id)
     for param in real_recovery_torch_cases():
@@ -242,7 +242,7 @@ def recovery_case_by_id(case_id: str) -> RealRecoveryCase:
 
 def recovery_extra_step_configure(case: RealRecoveryCase) -> dict[type, StepConfigureFn] | None:
     if case.id in TENSORFLOW_RECOVERY_CASE_IDS:
-        from tests.helpers.training_recovery_tensorflow import tensorflow_step_configure
+        from .training_recovery_tensorflow import tensorflow_step_configure
 
         return tensorflow_step_configure()
     return None
@@ -330,7 +330,7 @@ def make_recovery_runtime(
     tmp_path: Path, case: RealRecoveryCase, *, working_dir: Workdir | None = None
 ) -> tuple[SmokeRuntime, list]:
     if case.id in TENSORFLOW_RECOVERY_CASE_IDS:
-        from tests.helpers.training_recovery_tensorflow import make_recovery_runtime as make_tf_recovery_runtime
+        from .training_recovery_tensorflow import make_recovery_runtime as make_tf_recovery_runtime
 
         return make_tf_recovery_runtime(tmp_path, case, working_dir=working_dir)
     workdir = working_dir if working_dir is not None else tmp_path
@@ -499,7 +499,7 @@ def invoke_real_train_callable_for_backfill(
     step: RecoveryTrainStep,
 ) -> None:
     if case.id in TENSORFLOW_RECOVERY_CASE_IDS:
-        from tests.helpers.training_recovery_tensorflow import (
+        from .training_recovery_tensorflow import (
             invoke_real_train_callable_for_backfill as invoke_tf_backfill,
         )
 

libs/datapipe-ml/tests/helpers/training_recovery_tensorflow.py

@@ -7,7 +7,7 @@
 import pytest
 from datapipe.types import IndexDF
 
-from tests.helpers.training_recovery import (
+from .training_recovery import (
     TENSORFLOW_RECOVERY_CASE_IDS,
     RealRecoveryCase,
     RecoveryTrainStep,
@@ -19,12 +19,13 @@
     direct_train_kwargs,
     make_runtime,
 )
-from tests.helpers.training_smoke import Workdir, classification_freeze_step, classification_train_step
+from .training_smoke import Workdir, classification_freeze_step, classification_train_step
 
 if TYPE_CHECKING:
     from datapipe_ml.frameworks.tensorflow.classification_runner import TF_ClassificationTrainingConfig
     from datapipe_ml.tasks.classification.train.tensorflow import Train_Tensorflow_ClassificationModel
 
+
 def _configure_tf_configs(configs: list[TF_ClassificationTrainingConfig], epochs: int) -> None:
     for config in configs:
         config.epochs = epochs

libs/datapipe-ml/tests/helpers/training_smoke.py

@@ -8,7 +8,7 @@
 from pathlib import Path, PurePosixPath
 from typing import Iterable
 
-from tests.helpers.cloud_storage import (
+from .cloud_storage import (
     assert_model_path_under_working_dir,
     assert_url_exists,
     is_cloud_url,
@@ -18,8 +18,8 @@
 
 Workdir = str | Path
 
-import pandas as pd
 import fsspec
+import pandas as pd
 from datapipe.compute import (
     Catalog,
     Pipeline,
@@ -33,6 +33,7 @@
 from sklearn.model_selection import train_test_split
 from sqlalchemy import Column
 from sqlalchemy.sql.sqltypes import JSON, String
+
 from .dbconn import get_sqlite_dbconnstr
 
 TESTS_DIR = Path(__file__).parents[1]
@@ -232,7 +233,7 @@ def make_runtime(
 
 
 def make_cloud_runtime(tmp_path: Path, suffix: str, **kwargs) -> tuple[SmokeRuntime, str]:
-    from tests.helpers.cloud_storage import cloud_working_dir
+    from .cloud_storage import cloud_working_dir
 
     workdir = str(cloud_working_dir(suffix))
     runtime = make_runtime(tmp_path, working_dir=workdir, **kwargs)
@@ -455,7 +456,9 @@ def detection_freeze_step(workdir: Workdir):
     )
 
 
-def detection_train_step(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def detection_train_step(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.detection.train.yolov8 import (
         Train_YoloV8_DetectionModel,
         YoloV8_TrainingConfig,
@@ -548,7 +551,9 @@ def detection_train_step_with_local_checkpoint(
     )
 
 
-def detection_train_step_with_augmentations(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def detection_train_step_with_augmentations(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.detection.train.yolov8 import (
         Train_YoloV8_DetectionModel,
         YoloV8_TrainingConfig,
@@ -600,7 +605,9 @@ def detection_train_step_with_augmentations(workdir: Workdir, *, local_scratch:
     )
 
 
-def detection_yolov5_train_step(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def detection_yolov5_train_step(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.detection.train.yolov5 import (
         Train_YoloV5_DetectionModel,
         YoloV5_TrainingConfig,
@@ -642,7 +649,9 @@ def detection_yolov5_train_step(workdir: Workdir, *, local_scratch: Path | None
     )
 
 
-def detection_yolov5_train_step_with_augmentations(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def detection_yolov5_train_step_with_augmentations(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.detection.train.yolov5 import (
         Train_YoloV5_DetectionModel,
         YoloV5_TrainingConfig,
@@ -738,7 +747,9 @@ def segmentation_freeze_step(workdir: Workdir):
     )
 
 
-def segmentation_train_step(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def segmentation_train_step(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.segmentation.train.yolov8 import (
         Train_YoloV8_SegmentationModel,
         YoloV8_TrainingConfig,
@@ -847,7 +858,9 @@ def keypoints_freeze_step(workdir: Workdir):
     )
 
 
-def keypoints_train_step(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def keypoints_train_step(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.keypoints.train.yolov8 import (
         Train_YoloV8_KeypointsModel,
         YoloV8_TrainingConfig,
@@ -974,7 +987,9 @@ def classification_freeze_step(workdir: Workdir):
     )
 
 
-def classification_train_step(workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None):
+def classification_train_step(
+    workdir: Workdir, *, local_scratch: Path | None = None, filedir_fsspec_kwargs: dict | None = None
+):
     from datapipe_ml.tasks.classification.train.tensorflow import (
         TF_ClassificationTrainingConfig,
         Train_Tensorflow_ClassificationModel,

libs/datapipe-ml/tests/test_app_smoke.py

@@ -5,7 +5,7 @@
 from sqlalchemy import Column
 from sqlalchemy.sql.sqltypes import JSON, String
 
-from tests.fixtures.smoke_data import (
+from .fixtures.smoke_data import (
     SmokeDataset,
     define_ground_truth_for_classification,
     generate_smoke_data,