expand architecture tests to cover modelzoo format

tests/components/test_architectures.py

@@ -14,6 +14,7 @@
     CNNectomeUNetConfig,
     WrappedArchitectureConfig,
     ArchitectureConfig,
+    ModelZooConfig,
 )
 
 from pathlib import Path
@@ -22,6 +23,8 @@
 from dacapo.train import train_run
 from dacapo.validate import validate_run
 
+from bioimageio.spec.model.v0_5 import Author
+
 import pytest
 from pytest_lazy_fixtures import lf
 
@@ -38,7 +41,8 @@ def build_test_architecture_config(
     upsample: bool,
     use_attention: bool,
     padding: str,
-    wrapped: bool,
+    source: str,
+    tmp_path: Path,
 ) -> ArchitectureConfig:
     """
     Build the simplest architecture config given the parameters.
@@ -96,7 +100,9 @@ def build_test_architecture_config(
         padding=padding,
     )
 
-    if wrapped:
+    if source == "config":
+        return cnnectom_unet_config
+    elif source == "module":
         return WrappedArchitectureConfig(
             name="test_wrapped",
             module=cnnectom_unet_config.module(),
@@ -105,20 +111,30 @@ def build_test_architecture_config(
             input_shape=input_shape,
             scale=Coordinate(upsample_factors[0]) if upsample else None,
         )
-    else:
-        return cnnectom_unet_config
+    elif source == "bioimage_modelzoo":
+        run = RunConfig(
+            architecture_config=cnnectom_unet_config,
+            name="dacapo_modelzoo_test",
+        )
+        run.save_bioimage_io_model(
+            tmp_path / "dacapo_modelzoo_test.zip", authors=[Author(name="Test")]
+        )
+        return ModelZooConfig(
+            model_id = tmp_path / "dacapo_modelzoo_test.zip", name="test_model_zoo"
+        )
 
 
 # TODO: Move unet parameters that don't affect interaction with other modules
 # to a separate architcture test
+@pytest.mark.filterwarnings("ignore::FutureWarning")  # pytest treats this as an error but we don't care for now
 @pytest.mark.parametrize("data_dims", [2, 3])
 @pytest.mark.parametrize("channels", [True, False])
 @pytest.mark.parametrize("architecture_dims", [2, 3])
 @pytest.mark.parametrize("upsample", [True, False])
 @pytest.mark.parametrize("batch_norm", [True, False])
 @pytest.mark.parametrize("use_attention", [True, False])
 @pytest.mark.parametrize("padding", ["valid", "same"])
-@pytest.mark.parametrize("wrapped", [True, False])
+@pytest.mark.parametrize("source", ["config", "module", "bioimage_modelzoo"])
 def test_architectures(
     data_dims,
     channels,
@@ -127,7 +143,8 @@ def test_architectures(
     upsample,
     use_attention,
     padding,
-    wrapped,
+    source,
+    tmp_path,
 ):
     architecture_config = build_test_architecture_config(
         data_dims,
@@ -137,12 +154,18 @@ def test_architectures(
         upsample,
         use_attention,
         padding,
-        wrapped,
+        source,
+        tmp_path
     )
 
     in_data = torch.rand(
         (*(1, architecture_config.num_in_channels), *architecture_config.input_shape)
     )
     out_data = architecture_config.module()(in_data)
+    scale = architecture_config.scale(Coordinate((2,) * data_dims))
+    if upsample:
+        assert scale == Coordinate((1,) * data_dims)
+    else:
+        assert scale == Coordinate((2,) * data_dims)
 
     assert out_data.shape[1] == architecture_config.num_out_channels