diff --git a/samples/torchvision/shufflenet_v2_x1_5/graph_hash.txt b/samples/torchvision/shufflenet_v2_x1_5/graph_hash.txt
new file mode 100644
index 000000000..86a736fcf
--- /dev/null
+++ b/samples/torchvision/shufflenet_v2_x1_5/graph_hash.txt
@@ -0,0 +1 @@
+20eed391b1be5e6cec6d31e0c192b31edea281a3b2b87741ffdade1d93740c79
\ No newline at end of file
diff --git a/samples/torchvision/shufflenet_v2_x1_5/graph_net.json b/samples/torchvision/shufflenet_v2_x1_5/graph_net.json
new file mode 100644
index 000000000..1373fe3b5
--- /dev/null
+++ b/samples/torchvision/shufflenet_v2_x1_5/graph_net.json
@@ -0,0 +1,5 @@
+{
+    "framework": "torch",
+    "num_devices_required": 1,
+    "num_nodes_required": 1
+}
\ No newline at end of file
diff --git a/samples/torchvision/shufflenet_v2_x1_5/input_meta.py b/samples/torchvision/shufflenet_v2_x1_5/input_meta.py
new file mode 100644
index 000000000..e69de29bb
diff --git a/samples/torchvision/shufflenet_v2_x1_5/input_tensor_constraints.py b/samples/torchvision/shufflenet_v2_x1_5/input_tensor_constraints.py
new file mode 100644
index 000000000..e69de29bb
diff --git a/samples/torchvision/shufflenet_v2_x1_5/model.py b/samples/torchvision/shufflenet_v2_x1_5/model.py
new file mode 100644
index 000000000..3c4669811
--- /dev/null
+++ b/samples/torchvision/shufflenet_v2_x1_5/model.py
@@ -0,0 +1,2743 @@
+import torch
+
+
+class GraphModule(torch.nn.Module):
+    def forward(
+        self,
+        L_self_modules_conv1_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_x_: torch.Tensor,
+        L_self_modules_conv1_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_conv1_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_conv1_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv1_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_: torch.Tensor,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_: torch.Tensor,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv5_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv5_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_conv5_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_conv5_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv5_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_fc_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_fc_parameters_bias_: torch.nn.parameter.Parameter,
+    ):
+        l_self_modules_conv1_modules_0_parameters_weight_ = (
+            L_self_modules_conv1_modules_0_parameters_weight_
+        )
+        l_x_ = L_x_
+        l_self_modules_conv1_modules_1_buffers_running_mean_ = (
+            L_self_modules_conv1_modules_1_buffers_running_mean_
+        )
+        l_self_modules_conv1_modules_1_buffers_running_var_ = (
+            L_self_modules_conv1_modules_1_buffers_running_var_
+        )
+        l_self_modules_conv1_modules_1_parameters_weight_ = (
+            L_self_modules_conv1_modules_1_parameters_weight_
+        )
+        l_self_modules_conv1_modules_1_parameters_bias_ = (
+            L_self_modules_conv1_modules_1_parameters_bias_
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_ = L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_
+        l_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_ = L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_
+        l_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_ = (
+            L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_ = L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_
+        l_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_ = L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_
+        l_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_ = (
+            L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_ = L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_
+        l_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_ = L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_
+        l_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_ = (
+            L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_ = L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_
+        l_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_ = L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_
+        l_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_ = (
+            L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_ = L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_
+        l_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_ = L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_
+        l_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_ = (
+            L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_ = L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_
+        l_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_ = L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_
+        l_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_ = L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_
+        l_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_ = L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_
+        l_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_ = L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_
+        l_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_ = L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_
+        l_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_
+        )
+        l_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_ = (
+            L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_
+        )
+        l_self_modules_conv5_modules_0_parameters_weight_ = (
+            L_self_modules_conv5_modules_0_parameters_weight_
+        )
+        l_self_modules_conv5_modules_1_buffers_running_mean_ = (
+            L_self_modules_conv5_modules_1_buffers_running_mean_
+        )
+        l_self_modules_conv5_modules_1_buffers_running_var_ = (
+            L_self_modules_conv5_modules_1_buffers_running_var_
+        )
+        l_self_modules_conv5_modules_1_parameters_weight_ = (
+            L_self_modules_conv5_modules_1_parameters_weight_
+        )
+        l_self_modules_conv5_modules_1_parameters_bias_ = (
+            L_self_modules_conv5_modules_1_parameters_bias_
+        )
+        l_self_modules_fc_parameters_weight_ = L_self_modules_fc_parameters_weight_
+        l_self_modules_fc_parameters_bias_ = L_self_modules_fc_parameters_bias_
+        input_1 = torch.conv2d(
+            l_x_,
+            l_self_modules_conv1_modules_0_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        l_x_ = l_self_modules_conv1_modules_0_parameters_weight_ = None
+        input_2 = torch.nn.functional.batch_norm(
+            input_1,
+            l_self_modules_conv1_modules_1_buffers_running_mean_,
+            l_self_modules_conv1_modules_1_buffers_running_var_,
+            l_self_modules_conv1_modules_1_parameters_weight_,
+            l_self_modules_conv1_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_1 = (
+            l_self_modules_conv1_modules_1_buffers_running_mean_
+        ) = (
+            l_self_modules_conv1_modules_1_buffers_running_var_
+        ) = (
+            l_self_modules_conv1_modules_1_parameters_weight_
+        ) = l_self_modules_conv1_modules_1_parameters_bias_ = None
+        input_3 = torch.nn.functional.relu(input_2, inplace=True)
+        input_2 = None
+        x = torch.nn.functional.max_pool2d(
+            input_3, 3, 2, 1, 1, ceil_mode=False, return_indices=False
+        )
+        input_3 = None
+        input_4 = torch.conv2d(
+            x,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            24,
+        )
+        l_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_ = (
+            None
+        )
+        input_5 = torch.nn.functional.batch_norm(
+            input_4,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_4 = l_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_ = l_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_ = (
+            l_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_
+        ) = None
+        input_6 = torch.conv2d(
+            input_5,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_5 = (
+            l_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_
+        ) = None
+        input_7 = torch.nn.functional.batch_norm(
+            input_6,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_,
+            l_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_6 = l_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_ = l_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_ = (
+            l_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_
+        ) = None
+        input_8 = torch.nn.functional.relu(input_7, inplace=True)
+        input_7 = None
+        input_9 = torch.conv2d(
+            x,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_10 = torch.nn.functional.batch_norm(
+            input_9,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_9 = l_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_11 = torch.nn.functional.relu(input_10, inplace=True)
+        input_10 = None
+        input_12 = torch.conv2d(
+            input_11,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            88,
+        )
+        input_11 = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_13 = torch.nn.functional.batch_norm(
+            input_12,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_12 = l_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_14 = torch.conv2d(
+            input_13,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_13 = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_15 = torch.nn.functional.batch_norm(
+            input_14,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_14 = l_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_16 = torch.nn.functional.relu(input_15, inplace=True)
+        input_15 = None
+        out = torch.cat((input_8, input_16), dim=1)
+        input_8 = input_16 = None
+        x_1 = out.view(1, 2, 88, 28, 28)
+        out = None
+        transpose = torch.transpose(x_1, 1, 2)
+        x_1 = None
+        x_2 = transpose.contiguous()
+        transpose = None
+        x_3 = x_2.view(1, 176, 28, 28)
+        x_2 = None
+        chunk = x_3.chunk(2, dim=1)
+        x_3 = None
+        x1 = chunk[0]
+        x2 = chunk[1]
+        chunk = None
+        input_17 = torch.conv2d(
+            x2,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2 = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_18 = torch.nn.functional.batch_norm(
+            input_17,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_17 = l_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_19 = torch.nn.functional.relu(input_18, inplace=True)
+        input_18 = None
+        input_20 = torch.conv2d(
+            input_19,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            88,
+        )
+        input_19 = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_21 = torch.nn.functional.batch_norm(
+            input_20,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_20 = l_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_22 = torch.conv2d(
+            input_21,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_21 = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_23 = torch.nn.functional.batch_norm(
+            input_22,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_22 = l_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_24 = torch.nn.functional.relu(input_23, inplace=True)
+        input_23 = None
+        out_1 = torch.cat((x1, input_24), dim=1)
+        x1 = input_24 = None
+        x_4 = out_1.view(1, 2, 88, 28, 28)
+        out_1 = None
+        transpose_1 = torch.transpose(x_4, 1, 2)
+        x_4 = None
+        x_5 = transpose_1.contiguous()
+        transpose_1 = None
+        x_6 = x_5.view(1, 176, 28, 28)
+        x_5 = None
+        chunk_1 = x_6.chunk(2, dim=1)
+        x_6 = None
+        x1_1 = chunk_1[0]
+        x2_1 = chunk_1[1]
+        chunk_1 = None
+        input_25 = torch.conv2d(
+            x2_1,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_1 = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_26 = torch.nn.functional.batch_norm(
+            input_25,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_25 = l_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_27 = torch.nn.functional.relu(input_26, inplace=True)
+        input_26 = None
+        input_28 = torch.conv2d(
+            input_27,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            88,
+        )
+        input_27 = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_29 = torch.nn.functional.batch_norm(
+            input_28,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_28 = l_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_30 = torch.conv2d(
+            input_29,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_29 = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_31 = torch.nn.functional.batch_norm(
+            input_30,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_30 = l_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_32 = torch.nn.functional.relu(input_31, inplace=True)
+        input_31 = None
+        out_2 = torch.cat((x1_1, input_32), dim=1)
+        x1_1 = input_32 = None
+        x_7 = out_2.view(1, 2, 88, 28, 28)
+        out_2 = None
+        transpose_2 = torch.transpose(x_7, 1, 2)
+        x_7 = None
+        x_8 = transpose_2.contiguous()
+        transpose_2 = None
+        x_9 = x_8.view(1, 176, 28, 28)
+        x_8 = None
+        chunk_2 = x_9.chunk(2, dim=1)
+        x_9 = None
+        x1_2 = chunk_2[0]
+        x2_2 = chunk_2[1]
+        chunk_2 = None
+        input_33 = torch.conv2d(
+            x2_2,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_2 = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_34 = torch.nn.functional.batch_norm(
+            input_33,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_33 = l_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_35 = torch.nn.functional.relu(input_34, inplace=True)
+        input_34 = None
+        input_36 = torch.conv2d(
+            input_35,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            88,
+        )
+        input_35 = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_37 = torch.nn.functional.batch_norm(
+            input_36,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_36 = l_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_38 = torch.conv2d(
+            input_37,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_37 = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_39 = torch.nn.functional.batch_norm(
+            input_38,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_38 = l_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_40 = torch.nn.functional.relu(input_39, inplace=True)
+        input_39 = None
+        out_3 = torch.cat((x1_2, input_40), dim=1)
+        x1_2 = input_40 = None
+        x_10 = out_3.view(1, 2, 88, 28, 28)
+        out_3 = None
+        transpose_3 = torch.transpose(x_10, 1, 2)
+        x_10 = None
+        x_11 = transpose_3.contiguous()
+        transpose_3 = None
+        x_12 = x_11.view(1, 176, 28, 28)
+        x_11 = None
+        input_41 = torch.conv2d(
+            x_12,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        l_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_ = (
+            None
+        )
+        input_42 = torch.nn.functional.batch_norm(
+            input_41,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_41 = l_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_
+        ) = None
+        input_43 = torch.conv2d(
+            input_42,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_42 = (
+            l_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_
+        ) = None
+        input_44 = torch.nn.functional.batch_norm(
+            input_43,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_,
+            l_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_43 = l_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_ = l_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_ = (
+            l_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_
+        ) = None
+        input_45 = torch.nn.functional.relu(input_44, inplace=True)
+        input_44 = None
+        input_46 = torch.conv2d(
+            x_12,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x_12 = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_47 = torch.nn.functional.batch_norm(
+            input_46,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_46 = l_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_48 = torch.nn.functional.relu(input_47, inplace=True)
+        input_47 = None
+        input_49 = torch.conv2d(
+            input_48,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_48 = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_50 = torch.nn.functional.batch_norm(
+            input_49,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_49 = l_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_51 = torch.conv2d(
+            input_50,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_50 = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_52 = torch.nn.functional.batch_norm(
+            input_51,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_51 = l_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_53 = torch.nn.functional.relu(input_52, inplace=True)
+        input_52 = None
+        out_4 = torch.cat((input_45, input_53), dim=1)
+        input_45 = input_53 = None
+        x_13 = out_4.view(1, 2, 176, 14, 14)
+        out_4 = None
+        transpose_4 = torch.transpose(x_13, 1, 2)
+        x_13 = None
+        x_14 = transpose_4.contiguous()
+        transpose_4 = None
+        x_15 = x_14.view(1, 352, 14, 14)
+        x_14 = None
+        chunk_3 = x_15.chunk(2, dim=1)
+        x_15 = None
+        x1_3 = chunk_3[0]
+        x2_3 = chunk_3[1]
+        chunk_3 = None
+        input_54 = torch.conv2d(
+            x2_3,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_3 = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_55 = torch.nn.functional.batch_norm(
+            input_54,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_54 = l_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_56 = torch.nn.functional.relu(input_55, inplace=True)
+        input_55 = None
+        input_57 = torch.conv2d(
+            input_56,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_56 = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_58 = torch.nn.functional.batch_norm(
+            input_57,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_57 = l_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_59 = torch.conv2d(
+            input_58,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_58 = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_60 = torch.nn.functional.batch_norm(
+            input_59,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_59 = l_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_61 = torch.nn.functional.relu(input_60, inplace=True)
+        input_60 = None
+        out_5 = torch.cat((x1_3, input_61), dim=1)
+        x1_3 = input_61 = None
+        x_16 = out_5.view(1, 2, 176, 14, 14)
+        out_5 = None
+        transpose_5 = torch.transpose(x_16, 1, 2)
+        x_16 = None
+        x_17 = transpose_5.contiguous()
+        transpose_5 = None
+        x_18 = x_17.view(1, 352, 14, 14)
+        x_17 = None
+        chunk_4 = x_18.chunk(2, dim=1)
+        x_18 = None
+        x1_4 = chunk_4[0]
+        x2_4 = chunk_4[1]
+        chunk_4 = None
+        input_62 = torch.conv2d(
+            x2_4,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_4 = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_63 = torch.nn.functional.batch_norm(
+            input_62,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_62 = l_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_64 = torch.nn.functional.relu(input_63, inplace=True)
+        input_63 = None
+        input_65 = torch.conv2d(
+            input_64,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_64 = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_66 = torch.nn.functional.batch_norm(
+            input_65,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_65 = l_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_67 = torch.conv2d(
+            input_66,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_66 = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_68 = torch.nn.functional.batch_norm(
+            input_67,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_67 = l_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_69 = torch.nn.functional.relu(input_68, inplace=True)
+        input_68 = None
+        out_6 = torch.cat((x1_4, input_69), dim=1)
+        x1_4 = input_69 = None
+        x_19 = out_6.view(1, 2, 176, 14, 14)
+        out_6 = None
+        transpose_6 = torch.transpose(x_19, 1, 2)
+        x_19 = None
+        x_20 = transpose_6.contiguous()
+        transpose_6 = None
+        x_21 = x_20.view(1, 352, 14, 14)
+        x_20 = None
+        chunk_5 = x_21.chunk(2, dim=1)
+        x_21 = None
+        x1_5 = chunk_5[0]
+        x2_5 = chunk_5[1]
+        chunk_5 = None
+        input_70 = torch.conv2d(
+            x2_5,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_5 = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_71 = torch.nn.functional.batch_norm(
+            input_70,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_70 = l_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_72 = torch.nn.functional.relu(input_71, inplace=True)
+        input_71 = None
+        input_73 = torch.conv2d(
+            input_72,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_72 = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_74 = torch.nn.functional.batch_norm(
+            input_73,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_73 = l_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_75 = torch.conv2d(
+            input_74,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_74 = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_76 = torch.nn.functional.batch_norm(
+            input_75,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_75 = l_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_77 = torch.nn.functional.relu(input_76, inplace=True)
+        input_76 = None
+        out_7 = torch.cat((x1_5, input_77), dim=1)
+        x1_5 = input_77 = None
+        x_22 = out_7.view(1, 2, 176, 14, 14)
+        out_7 = None
+        transpose_7 = torch.transpose(x_22, 1, 2)
+        x_22 = None
+        x_23 = transpose_7.contiguous()
+        transpose_7 = None
+        x_24 = x_23.view(1, 352, 14, 14)
+        x_23 = None
+        chunk_6 = x_24.chunk(2, dim=1)
+        x_24 = None
+        x1_6 = chunk_6[0]
+        x2_6 = chunk_6[1]
+        chunk_6 = None
+        input_78 = torch.conv2d(
+            x2_6,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_6 = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_79 = torch.nn.functional.batch_norm(
+            input_78,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_78 = l_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_80 = torch.nn.functional.relu(input_79, inplace=True)
+        input_79 = None
+        input_81 = torch.conv2d(
+            input_80,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_80 = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_82 = torch.nn.functional.batch_norm(
+            input_81,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_81 = l_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_83 = torch.conv2d(
+            input_82,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_82 = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_84 = torch.nn.functional.batch_norm(
+            input_83,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_83 = l_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_85 = torch.nn.functional.relu(input_84, inplace=True)
+        input_84 = None
+        out_8 = torch.cat((x1_6, input_85), dim=1)
+        x1_6 = input_85 = None
+        x_25 = out_8.view(1, 2, 176, 14, 14)
+        out_8 = None
+        transpose_8 = torch.transpose(x_25, 1, 2)
+        x_25 = None
+        x_26 = transpose_8.contiguous()
+        transpose_8 = None
+        x_27 = x_26.view(1, 352, 14, 14)
+        x_26 = None
+        chunk_7 = x_27.chunk(2, dim=1)
+        x_27 = None
+        x1_7 = chunk_7[0]
+        x2_7 = chunk_7[1]
+        chunk_7 = None
+        input_86 = torch.conv2d(
+            x2_7,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_7 = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_87 = torch.nn.functional.batch_norm(
+            input_86,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_86 = l_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_88 = torch.nn.functional.relu(input_87, inplace=True)
+        input_87 = None
+        input_89 = torch.conv2d(
+            input_88,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_88 = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_90 = torch.nn.functional.batch_norm(
+            input_89,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_89 = l_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_91 = torch.conv2d(
+            input_90,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_90 = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_92 = torch.nn.functional.batch_norm(
+            input_91,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_91 = l_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_93 = torch.nn.functional.relu(input_92, inplace=True)
+        input_92 = None
+        out_9 = torch.cat((x1_7, input_93), dim=1)
+        x1_7 = input_93 = None
+        x_28 = out_9.view(1, 2, 176, 14, 14)
+        out_9 = None
+        transpose_9 = torch.transpose(x_28, 1, 2)
+        x_28 = None
+        x_29 = transpose_9.contiguous()
+        transpose_9 = None
+        x_30 = x_29.view(1, 352, 14, 14)
+        x_29 = None
+        chunk_8 = x_30.chunk(2, dim=1)
+        x_30 = None
+        x1_8 = chunk_8[0]
+        x2_8 = chunk_8[1]
+        chunk_8 = None
+        input_94 = torch.conv2d(
+            x2_8,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_8 = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_95 = torch.nn.functional.batch_norm(
+            input_94,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_94 = l_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_96 = torch.nn.functional.relu(input_95, inplace=True)
+        input_95 = None
+        input_97 = torch.conv2d(
+            input_96,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_96 = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_98 = torch.nn.functional.batch_norm(
+            input_97,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_97 = l_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_99 = torch.conv2d(
+            input_98,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_98 = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_100 = torch.nn.functional.batch_norm(
+            input_99,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_99 = l_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_101 = torch.nn.functional.relu(input_100, inplace=True)
+        input_100 = None
+        out_10 = torch.cat((x1_8, input_101), dim=1)
+        x1_8 = input_101 = None
+        x_31 = out_10.view(1, 2, 176, 14, 14)
+        out_10 = None
+        transpose_10 = torch.transpose(x_31, 1, 2)
+        x_31 = None
+        x_32 = transpose_10.contiguous()
+        transpose_10 = None
+        x_33 = x_32.view(1, 352, 14, 14)
+        x_32 = None
+        chunk_9 = x_33.chunk(2, dim=1)
+        x_33 = None
+        x1_9 = chunk_9[0]
+        x2_9 = chunk_9[1]
+        chunk_9 = None
+        input_102 = torch.conv2d(
+            x2_9,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_9 = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_103 = torch.nn.functional.batch_norm(
+            input_102,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_102 = l_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_104 = torch.nn.functional.relu(input_103, inplace=True)
+        input_103 = None
+        input_105 = torch.conv2d(
+            input_104,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            176,
+        )
+        input_104 = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_106 = torch.nn.functional.batch_norm(
+            input_105,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_105 = l_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_107 = torch.conv2d(
+            input_106,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_106 = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_108 = torch.nn.functional.batch_norm(
+            input_107,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_107 = l_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_109 = torch.nn.functional.relu(input_108, inplace=True)
+        input_108 = None
+        out_11 = torch.cat((x1_9, input_109), dim=1)
+        x1_9 = input_109 = None
+        x_34 = out_11.view(1, 2, 176, 14, 14)
+        out_11 = None
+        transpose_11 = torch.transpose(x_34, 1, 2)
+        x_34 = None
+        x_35 = transpose_11.contiguous()
+        transpose_11 = None
+        x_36 = x_35.view(1, 352, 14, 14)
+        x_35 = None
+        input_110 = torch.conv2d(
+            x_36,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            352,
+        )
+        l_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_ = (
+            None
+        )
+        input_111 = torch.nn.functional.batch_norm(
+            input_110,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_110 = l_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_ = l_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_ = (
+            l_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_
+        ) = None
+        input_112 = torch.conv2d(
+            input_111,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_111 = (
+            l_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_
+        ) = None
+        input_113 = torch.nn.functional.batch_norm(
+            input_112,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_,
+            l_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_112 = l_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_ = l_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_ = (
+            l_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_
+        ) = None
+        input_114 = torch.nn.functional.relu(input_113, inplace=True)
+        input_113 = None
+        input_115 = torch.conv2d(
+            x_36,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x_36 = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_116 = torch.nn.functional.batch_norm(
+            input_115,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_115 = l_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_117 = torch.nn.functional.relu(input_116, inplace=True)
+        input_116 = None
+        input_118 = torch.conv2d(
+            input_117,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (2, 2),
+            (1, 1),
+            (1, 1),
+            352,
+        )
+        input_117 = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_119 = torch.nn.functional.batch_norm(
+            input_118,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_118 = l_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_120 = torch.conv2d(
+            input_119,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_119 = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_121 = torch.nn.functional.batch_norm(
+            input_120,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_120 = l_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_122 = torch.nn.functional.relu(input_121, inplace=True)
+        input_121 = None
+        out_12 = torch.cat((input_114, input_122), dim=1)
+        input_114 = input_122 = None
+        x_37 = out_12.view(1, 2, 352, 7, 7)
+        out_12 = None
+        transpose_12 = torch.transpose(x_37, 1, 2)
+        x_37 = None
+        x_38 = transpose_12.contiguous()
+        transpose_12 = None
+        x_39 = x_38.view(1, 704, 7, 7)
+        x_38 = None
+        chunk_10 = x_39.chunk(2, dim=1)
+        x_39 = None
+        x1_10 = chunk_10[0]
+        x2_10 = chunk_10[1]
+        chunk_10 = None
+        input_123 = torch.conv2d(
+            x2_10,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_10 = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_124 = torch.nn.functional.batch_norm(
+            input_123,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_123 = l_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_125 = torch.nn.functional.relu(input_124, inplace=True)
+        input_124 = None
+        input_126 = torch.conv2d(
+            input_125,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            352,
+        )
+        input_125 = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_127 = torch.nn.functional.batch_norm(
+            input_126,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_126 = l_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_128 = torch.conv2d(
+            input_127,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_127 = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_129 = torch.nn.functional.batch_norm(
+            input_128,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_128 = l_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_130 = torch.nn.functional.relu(input_129, inplace=True)
+        input_129 = None
+        out_13 = torch.cat((x1_10, input_130), dim=1)
+        x1_10 = input_130 = None
+        x_40 = out_13.view(1, 2, 352, 7, 7)
+        out_13 = None
+        transpose_13 = torch.transpose(x_40, 1, 2)
+        x_40 = None
+        x_41 = transpose_13.contiguous()
+        transpose_13 = None
+        x_42 = x_41.view(1, 704, 7, 7)
+        x_41 = None
+        chunk_11 = x_42.chunk(2, dim=1)
+        x_42 = None
+        x1_11 = chunk_11[0]
+        x2_11 = chunk_11[1]
+        chunk_11 = None
+        input_131 = torch.conv2d(
+            x2_11,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_11 = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_132 = torch.nn.functional.batch_norm(
+            input_131,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_131 = l_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_133 = torch.nn.functional.relu(input_132, inplace=True)
+        input_132 = None
+        input_134 = torch.conv2d(
+            input_133,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            352,
+        )
+        input_133 = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_135 = torch.nn.functional.batch_norm(
+            input_134,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_134 = l_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_136 = torch.conv2d(
+            input_135,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_135 = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_137 = torch.nn.functional.batch_norm(
+            input_136,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_136 = l_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_138 = torch.nn.functional.relu(input_137, inplace=True)
+        input_137 = None
+        out_14 = torch.cat((x1_11, input_138), dim=1)
+        x1_11 = input_138 = None
+        x_43 = out_14.view(1, 2, 352, 7, 7)
+        out_14 = None
+        transpose_14 = torch.transpose(x_43, 1, 2)
+        x_43 = None
+        x_44 = transpose_14.contiguous()
+        transpose_14 = None
+        x_45 = x_44.view(1, 704, 7, 7)
+        x_44 = None
+        chunk_12 = x_45.chunk(2, dim=1)
+        x_45 = None
+        x1_12 = chunk_12[0]
+        x2_12 = chunk_12[1]
+        chunk_12 = None
+        input_139 = torch.conv2d(
+            x2_12,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x2_12 = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_
+        ) = None
+        input_140 = torch.nn.functional.batch_norm(
+            input_139,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_139 = l_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_ = l_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_ = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_
+        ) = None
+        input_141 = torch.nn.functional.relu(input_140, inplace=True)
+        input_140 = None
+        input_142 = torch.conv2d(
+            input_141,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_,
+            None,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            352,
+        )
+        input_141 = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_
+        ) = None
+        input_143 = torch.nn.functional.batch_norm(
+            input_142,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_142 = l_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_ = l_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_ = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_
+        ) = None
+        input_144 = torch.conv2d(
+            input_143,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_143 = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_
+        ) = None
+        input_145 = torch.nn.functional.batch_norm(
+            input_144,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_,
+            l_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_144 = l_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_ = l_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_ = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_
+        ) = (
+            l_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_
+        ) = None
+        input_146 = torch.nn.functional.relu(input_145, inplace=True)
+        input_145 = None
+        out_15 = torch.cat((x1_12, input_146), dim=1)
+        x1_12 = input_146 = None
+        x_46 = out_15.view(1, 2, 352, 7, 7)
+        out_15 = None
+        transpose_15 = torch.transpose(x_46, 1, 2)
+        x_46 = None
+        x_47 = transpose_15.contiguous()
+        transpose_15 = None
+        x_48 = x_47.view(1, 704, 7, 7)
+        x_47 = None
+        input_147 = torch.conv2d(
+            x_48,
+            l_self_modules_conv5_modules_0_parameters_weight_,
+            None,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        x_48 = l_self_modules_conv5_modules_0_parameters_weight_ = None
+        input_148 = torch.nn.functional.batch_norm(
+            input_147,
+            l_self_modules_conv5_modules_1_buffers_running_mean_,
+            l_self_modules_conv5_modules_1_buffers_running_var_,
+            l_self_modules_conv5_modules_1_parameters_weight_,
+            l_self_modules_conv5_modules_1_parameters_bias_,
+            False,
+            0.1,
+            1e-05,
+        )
+        input_147 = (
+            l_self_modules_conv5_modules_1_buffers_running_mean_
+        ) = (
+            l_self_modules_conv5_modules_1_buffers_running_var_
+        ) = (
+            l_self_modules_conv5_modules_1_parameters_weight_
+        ) = l_self_modules_conv5_modules_1_parameters_bias_ = None
+        input_149 = torch.nn.functional.relu(input_148, inplace=True)
+        input_148 = None
+        x_49 = input_149.mean([2, 3])
+        input_149 = None
+        x_50 = torch._C._nn.linear(
+            x_49,
+            l_self_modules_fc_parameters_weight_,
+            l_self_modules_fc_parameters_bias_,
+        )
+        x_49 = (
+            l_self_modules_fc_parameters_weight_
+        ) = l_self_modules_fc_parameters_bias_ = None
+        return (x_50,)
diff --git a/samples/torchvision/shufflenet_v2_x1_5/weight_meta.py b/samples/torchvision/shufflenet_v2_x1_5/weight_meta.py
new file mode 100644
index 000000000..08937c7a3
--- /dev/null
+++ b/samples/torchvision/shufflenet_v2_x1_5/weight_meta.py
@@ -0,0 +1,3152 @@
+class Program_weight_tensor_meta_L_self_modules_conv1_modules_0_parameters_weight_:
+    name = "L_self_modules_conv1_modules_0_parameters_weight_"
+    shape = [24, 3, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.006
+    std = 0.793
+    data = None
+
+
+class Program_weight_tensor_meta_L_x_:
+    name = "L_x_"
+    shape = [1, 3, 224, 224]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.227
+    std = 1.286
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv1_modules_1_buffers_running_mean_:
+    name = "L_self_modules_conv1_modules_1_buffers_running_mean_"
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.024
+    std = 0.407
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv1_modules_1_buffers_running_var_:
+    name = "L_self_modules_conv1_modules_1_buffers_running_var_"
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 13.317
+    std = 21.986
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv1_modules_1_parameters_weight_:
+    name = "L_self_modules_conv1_modules_1_parameters_weight_"
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 5.296
+    std = 2.182
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv1_modules_1_parameters_bias_:
+    name = "L_self_modules_conv1_modules_1_parameters_bias_"
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 3.813
+    std = 5.516
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch1_modules_0_parameters_weight_"
+    )
+    shape = [24, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.016
+    std = 0.443
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_mean_"
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -2.349
+    std = 9.589
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch1_modules_1_buffers_running_var_"
+    )
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 55.150
+    std = 81.187
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_weight_"
+    )
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 3.235
+    std = 1.509
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_:
+    name = "L_self_modules_stage2_modules_0_modules_branch1_modules_1_parameters_bias_"
+    shape = [24]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch1_modules_2_parameters_weight_"
+    )
+    shape = [88, 24, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.003
+    std = 0.206
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch1_modules_3_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 21.674
+    std = 19.420
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.199
+    std = 0.961
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_:
+    name = "L_self_modules_stage2_modules_0_modules_branch1_modules_3_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.031
+    std = 1.677
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [88, 24, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.031
+    std = 0.319
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -4.308
+    std = 17.214
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 124.169
+    std = 88.641
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 3.201
+    std = 1.330
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage2_modules_0_modules_branch2_modules_1_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.121
+    std = 2.188
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [88, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.004
+    std = 0.309
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.125
+    std = 1.240
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.324
+    std = 2.087
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 3.094
+    std = 0.755
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage2_modules_0_modules_branch2_modules_4_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.001
+    std = 0.157
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 36.440
+    std = 34.688
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.095
+    std = 0.827
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage2_modules_0_modules_branch2_modules_6_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.347
+    std = 3.130
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.131
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.179
+    std = 2.755
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 4.607
+    std = 4.569
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.759
+    std = 0.554
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage2_modules_1_modules_branch2_modules_1_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.059
+    std = 1.779
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [88, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.003
+    std = 0.327
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.022
+    std = 0.658
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.036
+    std = 1.133
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.515
+    std = 0.523
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage2_modules_1_modules_branch2_modules_4_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.140
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 17.218
+    std = 9.564
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.397
+    std = 0.721
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage2_modules_1_modules_branch2_modules_6_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.354
+    std = 2.092
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.010
+    std = 0.131
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.570
+    std = 2.384
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 4.678
+    std = 2.554
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.872
+    std = 0.499
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage2_modules_2_modules_branch2_modules_1_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.505
+    std = 1.100
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [88, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.004
+    std = 0.224
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.048
+    std = 0.701
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.575
+    std = 0.380
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.368
+    std = 0.622
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage2_modules_2_modules_branch2_modules_4_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.002
+    std = 0.128
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 13.613
+    std = 5.222
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.685
+    std = 0.769
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage2_modules_2_modules_branch2_modules_6_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.166
+    std = 1.671
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.124
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.523
+    std = 2.254
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 3.983
+    std = 1.724
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.735
+    std = 0.401
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage2_modules_3_modules_branch2_modules_1_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.995
+    std = 1.038
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [88, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.008
+    std = 0.242
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.010
+    std = 0.722
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.693
+    std = 0.390
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.496
+    std = 0.507
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage2_modules_3_modules_branch2_modules_4_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [88, 88, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.131
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 12.921
+    std = 4.303
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.807
+    std = 0.511
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage2_modules_3_modules_branch2_modules_6_parameters_bias_"
+    shape = [88]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.367
+    std = 1.794
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch1_modules_0_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.011
+    std = 0.155
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.021
+    std = 2.232
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch1_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.024
+    std = 0.769
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.253
+    std = 0.501
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_0_modules_branch1_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch1_modules_2_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.098
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch1_modules_3_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 15.313
+    std = 9.817
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.686
+    std = 0.526
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_:
+    name = "L_self_modules_stage3_modules_0_modules_branch1_modules_3_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.478
+    std = 1.687
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.101
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.675
+    std = 3.873
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 7.140
+    std = 2.766
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.010
+    std = 0.388
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_0_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -1.323
+    std = 0.943
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.011
+    std = 0.153
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.083
+    std = 0.437
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.224
+    std = 0.224
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.334
+    std = 0.322
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_0_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.001
+    std = 0.095
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 10.838
+    std = 5.996
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.623
+    std = 0.751
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_0_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.421
+    std = 2.459
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.004
+    std = 0.080
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.430
+    std = 1.495
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.390
+    std = 0.677
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.327
+    std = 0.300
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_1_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.454
+    std = 0.946
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.184
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.020
+    std = 0.436
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.284
+    std = 0.246
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.971
+    std = 0.494
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_1_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.085
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 7.313
+    std = 3.483
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.972
+    std = 0.800
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_1_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.243
+    std = 1.983
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.003
+    std = 0.079
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.058
+    std = 1.401
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.536
+    std = 0.701
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.410
+    std = 0.306
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_2_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.018
+    std = 0.868
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.001
+    std = 0.152
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.004
+    std = 0.351
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.189
+    std = 0.268
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.007
+    std = 0.371
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_2_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.079
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 6.718
+    std = 3.474
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.945
+    std = 0.733
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_2_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.396
+    std = 1.837
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.086
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.183
+    std = 2.209
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.603
+    std = 1.286
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.426
+    std = 0.271
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_3_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.160
+    std = 0.854
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.008
+    std = 0.174
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.025
+    std = 0.356
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.222
+    std = 0.200
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.969
+    std = 0.504
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_3_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.085
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 7.023
+    std = 3.949
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.767
+    std = 0.633
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_3_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.027
+    std = 1.774
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.083
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.066
+    std = 1.985
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.811
+    std = 1.490
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.458
+    std = 0.258
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_4_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.091
+    std = 0.684
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.010
+    std = 0.162
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.053
+    std = 0.320
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.214
+    std = 0.403
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.946
+    std = 0.497
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_4_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.081
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 6.861
+    std = 3.971
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.741
+    std = 0.707
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_4_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.162
+    std = 1.646
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.078
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.218
+    std = 1.851
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.632
+    std = 0.651
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.405
+    std = 0.273
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_5_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.006
+    std = 0.902
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.001
+    std = 0.163
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.003
+    std = 0.341
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.176
+    std = 0.137
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.988
+    std = 0.462
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_5_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.001
+    std = 0.079
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 6.317
+    std = 3.185
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.874
+    std = 0.607
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_5_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.705
+    std = 1.494
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.001
+    std = 0.077
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.020
+    std = 1.892
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.527
+    std = 0.437
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.422
+    std = 0.244
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_6_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.133
+    std = 0.851
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.154
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.017
+    std = 0.326
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.174
+    std = 0.158
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.018
+    std = 0.431
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_6_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.083
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 6.917
+    std = 2.009
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.758
+    std = 0.475
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_6_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.741
+    std = 1.575
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.078
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.163
+    std = 1.555
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.275
+    std = 0.390
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.411
+    std = 0.264
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage3_modules_7_modules_branch2_modules_1_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.174
+    std = 0.880
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [176, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.156
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.019
+    std = 0.327
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.183
+    std = 0.223
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.967
+    std = 0.578
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage3_modules_7_modules_branch2_modules_4_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [176, 176, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.082
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 7.152
+    std = 6.449
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.091
+    std = 0.498
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage3_modules_7_modules_branch2_modules_6_parameters_bias_"
+    shape = [176]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -1.198
+    std = 1.363
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch1_modules_0_parameters_weight_"
+    )
+    shape = [352, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.010
+    std = 0.098
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.051
+    std = 0.885
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch1_modules_1_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.197
+    std = 0.178
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.867
+    std = 0.304
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_:
+    name = "L_self_modules_stage4_modules_0_modules_branch1_modules_1_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch1_modules_2_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.063
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch1_modules_3_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 7.436
+    std = 3.067
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.399
+    std = 0.467
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_:
+    name = "L_self_modules_stage4_modules_0_modules_branch1_modules_3_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.080
+    std = 1.417
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.002
+    std = 0.064
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.474
+    std = 1.603
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 3.176
+    std = 0.962
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.658
+    std = 0.288
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage4_modules_0_modules_branch2_modules_1_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -1.625
+    std = 0.517
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [352, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.011
+    std = 0.114
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.010
+    std = 0.156
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.071
+    std = 0.184
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.107
+    std = 0.491
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage4_modules_0_modules_branch2_modules_4_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.068
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 9.406
+    std = 5.312
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.588
+    std = 0.711
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage4_modules_0_modules_branch2_modules_6_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.120
+    std = 1.808
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.006
+    std = 0.058
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.992
+    std = 0.959
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.511
+    std = 1.042
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.353
+    std = 0.191
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage4_modules_1_modules_branch2_modules_1_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.241
+    std = 0.643
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [352, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.005
+    std = 0.123
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.015
+    std = 0.186
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.084
+    std = 0.105
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.787
+    std = 0.508
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage4_modules_1_modules_branch2_modules_4_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.001
+    std = 0.061
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 7.019
+    std = 3.998
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.804
+    std = 0.764
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage4_modules_1_modules_branch2_modules_6_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.720
+    std = 1.612
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.005
+    std = 0.061
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.303
+    std = 0.829
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.992
+    std = 0.334
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.398
+    std = 0.165
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage4_modules_2_modules_branch2_modules_1_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.637
+    std = 0.606
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [352, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.008
+    std = 0.119
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.006
+    std = 0.168
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.080
+    std = 0.090
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.870
+    std = 0.540
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage4_modules_2_modules_branch2_modules_4_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.001
+    std = 0.065
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 8.798
+    std = 5.092
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.952
+    std = 0.825
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage4_modules_2_modules_branch2_modules_6_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.912
+    std = 1.514
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_0_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.004
+    std = 0.064
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.092
+    std = 1.056
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_1_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.030
+    std = 0.431
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.485
+    std = 0.196
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_:
+    name = "L_self_modules_stage4_modules_3_modules_branch2_modules_1_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.712
+    std = 0.620
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_3_parameters_weight_"
+    )
+    shape = [352, 1, 3, 3]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.005
+    std = 0.140
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.018
+    std = 0.155
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_4_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.086
+    std = 0.090
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.951
+    std = 0.483
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_:
+    name = "L_self_modules_stage4_modules_3_modules_branch2_modules_4_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_5_parameters_weight_"
+    )
+    shape = [352, 352, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.001
+    std = 0.067
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_:
+    name = "L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_mean_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.001
+    std = 0.001
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_6_buffers_running_var_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 8.151
+    std = 3.417
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_:
+    name = (
+        "L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_weight_"
+    )
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 2.269
+    std = 0.445
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_:
+    name = "L_self_modules_stage4_modules_3_modules_branch2_modules_6_parameters_bias_"
+    shape = [352]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -1.255
+    std = 1.142
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv5_modules_0_parameters_weight_:
+    name = "L_self_modules_conv5_modules_0_parameters_weight_"
+    shape = [1024, 704, 1, 1]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.004
+    std = 0.052
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv5_modules_1_buffers_running_mean_:
+    name = "L_self_modules_conv5_modules_1_buffers_running_mean_"
+    shape = [1024]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.793
+    std = 1.326
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv5_modules_1_buffers_running_var_:
+    name = "L_self_modules_conv5_modules_1_buffers_running_var_"
+    shape = [1024]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 1.959
+    std = 0.550
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv5_modules_1_parameters_weight_:
+    name = "L_self_modules_conv5_modules_1_parameters_weight_"
+    shape = [1024]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 5.407
+    std = 0.456
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv5_modules_1_parameters_bias_:
+    name = "L_self_modules_conv5_modules_1_parameters_bias_"
+    shape = [1024]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -8.129
+    std = 1.398
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_fc_parameters_weight_:
+    name = "L_self_modules_fc_parameters_weight_"
+    shape = [1000, 1024]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = 0.000
+    std = 0.060
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_fc_parameters_bias_:
+    name = "L_self_modules_fc_parameters_bias_"
+    shape = [1000]
+    dtype = "torch.float32"
+    device = "cuda:0"
+    mean = -0.000
+    std = 0.052
+    data = None