[Test] add test_convert_to_hls_layers_tfc_w1a2

tests/fpgadataflow/test_convert_to_hls_layers.py

@@ -27,8 +27,8 @@ export_onnx_path = "test_output_tfc.onnx"
 
 
 def test_convert_to_hls_layers_tfc_w1a1():
-    lfc = get_test_model_trained("TFC", 1, 1)
-    bo.export_finn_onnx(lfc, (1, 1, 28, 28), export_onnx_path)
+    tfc = get_test_model_trained("TFC", 1, 1)
+    bo.export_finn_onnx(tfc, (1, 1, 28, 28), export_onnx_path)
     model = ModelWrapper(export_onnx_path)
     model = model.transform(InferShapes())
     model = model.transform(FoldConstants())
@@ -91,7 +91,7 @@ def test_convert_to_hls_layers_tfc_w1a1():
     input_tensor = torch.from_numpy(nph.to_array(input_tensor)).float()
     assert input_tensor.shape == (1, 1, 28, 28)
     # do forward pass in PyTorch/Brevitas
-    expected = lfc.forward(input_tensor).detach().numpy()
+    expected = tfc.forward(input_tensor).detach().numpy()
     assert np.isclose(produced, expected, atol=1e-3).all()
 
 
@@ -129,38 +129,34 @@ def test_convert_to_hls_layers_tfc_w1a2():
     assert fc3.op_type == "StreamingFCLayer_Batch"
     assert model.get_tensor_shape(fc3.input[0]) == [1, 64]
     assert model.get_tensor_shape(fc3.input[1]) == [64, 10]
-    os.remove(export_onnx_path)
-
     fc0w = getCustomOp(fc0)
     fc0w.set_nodeattr("SIMD", 784)
     fc0w.set_nodeattr("PE", 16)
-
     fc1w = getCustomOp(fc1)
     fc1w.set_nodeattr("SIMD", 16)
     fc1w.set_nodeattr("PE", 16)
-
     fc2w = getCustomOp(fc2)
     fc2w.set_nodeattr("SIMD", 16)
     fc2w.set_nodeattr("PE", 16)
-
     fc3w = getCustomOp(fc3)
     fc3w.set_nodeattr("SIMD", 16)
     fc3w.set_nodeattr("PE", 10)
-
     model = model.transform(CodeGen_npysim())
     model = model.transform(Compile())
     model = model.transform(SetExecMode("npysim"))
-    # model.save("tfc.onnx")
-
     raw_i = get_data("finn", "data/onnx/mnist-conv/test_data_set_0/input_0.pb")
     input_tensor = onnx.load_tensor_from_string(raw_i)
     # run using FINN-based execution
     input_dict = {"global_in": nph.to_array(input_tensor)}
-    output_dict = oxe.execute_onnx(model, input_dict)
-    produced = output_dict[list(output_dict.keys())[0]]
-    # run using PyTorch/Brevitas
-    input_tensor = torch.from_numpy(nph.to_array(input_tensor)).float()
-    assert input_tensor.shape == (1, 1, 28, 28)
-    # do forward pass in PyTorch/Brevitas
-    expected = tfc.forward(input_tensor).detach().numpy()
+    output_dict = oxe.execute_onnx(model, input_dict, True)
+    produced = output_dict[model.graph.output[0].name]
+    model = ModelWrapper(export_onnx_path)
+    model = model.transform(InferShapes())
+    model = model.transform(FoldConstants())
+    model = model.transform(GiveUniqueNodeNames())
+    model = model.transform(GiveReadableTensorNames())
+    model = model.transform(Streamline())
+    golden_output_dict = oxe.execute_onnx(model, input_dict, True)
+    expected = golden_output_dict[model.graph.output[0].name]
     assert np.isclose(produced, expected, atol=1e-3).all()
+    os.remove(export_onnx_path)