Add dataset QA and custom YOLO training flow

backend/tests/test_catalog.py

@@ -7,9 +7,9 @@ def test_catalog_contains_required_capabilities():
     for task in [
         "dataset.video_frames",
         "segmodel.train",
+        "yolo.train_custom",
         "yolo.predict",
         "mmseg.flops_fps",
         "analysis.all",
     ]:
         assert task in TASK_TYPES
-

backend/tests/test_dataset_service.py

@@ -1,4 +1,7 @@
-from app.modules.dataset.service import create_dataset, describe_dataset
+import cv2
+import numpy as np
+
+from app.modules.dataset.service import create_dataset, describe_dataset, generate_yolo_dataset_yaml, validate_dataset
 
 
 def test_create_dataset_layout(tmp_path, monkeypatch):
@@ -11,3 +14,29 @@ def test_create_dataset_layout(tmp_path, monkeypatch):
     assert created["counts"] == {"images": 0, "labels": 0, "masks": 0}
     described = describe_dataset("case_01")
     assert described["layout"]["images"].endswith("images")
+
+
+def test_validate_dataset_and_generate_yolo_yaml(tmp_path, monkeypatch):
+    from types import SimpleNamespace
+    from app.modules.dataset import service
+
+    monkeypatch.setattr(service, "settings", SimpleNamespace(project_root=tmp_path))
+    create_dataset("case yolo", "demo")
+    root = service.dataset_dir("case_yolo")
+    image = np.zeros((32, 32, 3), dtype=np.uint8)
+    mask = np.zeros((32, 32), dtype=np.uint8)
+    cv2.imwrite(str(root / "images" / "sample.png"), image)
+    cv2.imwrite(str(root / "masks" / "sample.png"), mask)
+    (root / "labels" / "sample.txt").write_text("0 0.25 0.25 0.75 0.25 0.75 0.75 0.25 0.75\n", encoding="utf-8")
+
+    validation = validate_dataset("case_yolo")
+
+    assert validation["ready"] == {"yolo": True, "mask": True, "any": True}
+    assert validation["pairs"]["image_label"] == 1
+    assert validation["classes"] == [0]
+
+    generated = generate_yolo_dataset_yaml("case_yolo", ["tool"])
+
+    assert generated["relative_path"] == "var/uploads/datasets/case_yolo/dataset.yaml"
+    assert "nc: 1" in generated["content"]
+    assert "0: tool" in generated["content"]