[Test] Update test for MoveFlattenPastAffine to use data layout

tests/transformation/test_move_flatten_past_affine.py

@@ -25,27 +25,40 @@
 # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
 # OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+import pytest
 
 import numpy as np
 from onnx import TensorProto, helper
 
 from finn.core.modelwrapper import ModelWrapper
 from finn.core.datatype import DataType
+import finn.core.data_layout as DataLayout
 from finn.util.basic import gen_finn_dt_tensor
 from finn.transformation.infer_shapes import InferShapes
 from finn.transformation.infer_datatypes import InferDataTypes
+from finn.transformation.infer_data_layouts import InferDataLayouts
 from finn.transformation.general import GiveUniqueNodeNames, GiveReadableTensorNames
 from finn.transformation.streamline.reorder import MoveFlattenPastAffine
 import finn.core.onnx_exec as oxe
 
+# data layout
+@pytest.mark.parametrize("data_layout", [DataLayout.NHWC, DataLayout.NCHW])
+# batch size
+@pytest.mark.parametrize("batch_size", [1, 2])
+def test_move_flatten_past_affine(data_layout, batch_size):
+    if data_layout == DataLayout.NHWC:
+        ishape = [batch_size, 1, 1, 1024]
+        oshape = [batch_size, 1000]
+    else:
+        ishape = [batch_size, 1024, 1, 1]
+        oshape = [batch_size, 1000]
 
-def test_move_flatten_past_affine():
-    inp = helper.make_tensor_value_info("inp", TensorProto.FLOAT, [1, 1, 1, 1024])
+    inp = helper.make_tensor_value_info("inp", TensorProto.FLOAT, ishape)
     a0 = helper.make_tensor_value_info("a1", TensorProto.FLOAT, [1024, 1000])
     a1 = helper.make_tensor_value_info("a2", TensorProto.FLOAT, [])
     a2 = helper.make_tensor_value_info("a3", TensorProto.FLOAT, [1000])
+    outp = helper.make_tensor_value_info("outp", TensorProto.FLOAT, oshape)
 
-    outp = helper.make_tensor_value_info("outp", TensorProto.FLOAT, [1, 1000])
     flatten_node = helper.make_node("Flatten", ["inp"], ["flatten_out"])
     matmul_node = helper.make_node("MatMul", ["flatten_out", "a0"], ["matmul_out"])
     mul_node = helper.make_node("Mul", ["matmul_out", "a1"], ["mul_out"])
@@ -71,18 +84,23 @@ def test_move_flatten_past_affine():
     model.set_initializer("a2", a2_values)
 
     model.set_tensor_datatype("inp", DataType.INT2)
-    model.set_tensor_datatype("flatten_out", DataType.INT2)
+    model.set_tensor_layout("inp", data_layout)
     model = model.transform(InferShapes())
     model = model.transform(InferDataTypes())
+    model = model.transform(InferDataLayouts())
     model = model.transform(GiveUniqueNodeNames())
     model = model.transform(GiveReadableTensorNames())
 
     # compare execution before and after transformation
-    inp_values = gen_finn_dt_tensor(DataType.INT2, [1, 1, 1, 1024])
+    inp_values = gen_finn_dt_tensor(DataType.INT2, ishape)
     idict = {"inp": inp_values}
     model_transformed = model.transform(MoveFlattenPastAffine())
     assert oxe.compare_execution(model, model_transformed, idict)
 
-    # check if nodes have new order in transformed graph
-    assert model.graph != model_transformed.graph
-    assert model_transformed.graph.node[-1].op_type == "Flatten"
+    # depending on data layout check if graph is transformed or not
+    if data_layout == DataLayout.NHWC:
+        # check if nodes have new order in transformed graph
+        assert model.graph != model_transformed.graph
+        assert model_transformed.graph.node[-1].op_type == "Flatten"
+    else:
+        assert model.graph == model_transformed.graph