diff --git a/samples/pigu_hybrid/graph_hash.txt b/samples/pigu_hybrid/graph_hash.txt
new file mode 100644
index 000000000..359b0a72d
--- /dev/null
+++ b/samples/pigu_hybrid/graph_hash.txt
@@ -0,0 +1 @@
+b0d0d50c6f76c99353442db322e1a6b1994b47187e1ce995eaee1ef8c104400a
\ No newline at end of file
diff --git a/samples/pigu_hybrid/graph_net.json b/samples/pigu_hybrid/graph_net.json
new file mode 100644
index 000000000..660acd178
--- /dev/null
+++ b/samples/pigu_hybrid/graph_net.json
@@ -0,0 +1,7 @@
+{
+    "framework": "torch",
+    "num_devices_required": 1,
+    "num_nodes_required": 1,
+    "dynamic": true,
+    "model_name": "pigu_hybrid"
+}
\ No newline at end of file
diff --git a/samples/pigu_hybrid/input_meta.py b/samples/pigu_hybrid/input_meta.py
new file mode 100644
index 000000000..e69de29bb
diff --git a/samples/pigu_hybrid/input_tensor_constraints.py b/samples/pigu_hybrid/input_tensor_constraints.py
new file mode 100644
index 000000000..e69de29bb
diff --git a/samples/pigu_hybrid/model.py b/samples/pigu_hybrid/model.py
new file mode 100644
index 000000000..392377529
--- /dev/null
+++ b/samples/pigu_hybrid/model.py
@@ -0,0 +1,992 @@
+import torch
+
+
+class GraphModule(torch.nn.Module):
+    def forward(
+        self,
+        s10: torch.SymInt,
+        L_sensor_vals_: torch.Tensor,
+        L_sensor_coords_: torch.Tensor,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_1_eps: torch.Tensor,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_attn_dropout: torch.Tensor,
+        L_self_modules_sensor_encoder_modules_norm_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_norm_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_sensor_encoder_modules_norm_eps: torch.Tensor,
+        L_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_grid_coords_: torch.Tensor,
+        L_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_transition_x: torch.Tensor,
+        L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_: torch.nn.parameter.Parameter,
+        G_np_pi: torch.Tensor,
+        L_self_modules_projector_modules_sensor_embedding_buffers_B_high_: torch.Tensor,
+        L_self_modules_projector_modules_key_proj_smooth_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_modules_key_proj_smooth_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_modules_key_proj_detail_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_modules_key_proj_detail_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_projector_scale: torch.Tensor,
+        L_base_flow_: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_1_momentum: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_1_eps: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_4_momentum: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_in_conv_modules_double_conv_modules_4_eps: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_1_eps: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down1_modules_1_modules_double_conv_modules_4_eps: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_1_eps: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_down2_modules_1_modules_double_conv_modules_4_eps: torch.Tensor,
+        L_self_modules_up1_scale_factor: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_1_momentum: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_1_eps: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_4_momentum: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up1_modules_double_conv_modules_4_eps: torch.Tensor,
+        L_self_modules_up2_scale_factor: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_1_momentum: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_1_eps: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_4_momentum: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_: torch.Tensor,
+        L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_: torch.nn.parameter.Parameter,
+        L_self_modules_conv_up2_modules_double_conv_modules_4_eps: torch.Tensor,
+        L_self_modules_out_conv_parameters_weight_: torch.nn.parameter.Parameter,
+        L_self_modules_out_conv_parameters_bias_: torch.nn.parameter.Parameter,
+    ):
+        l_sensor_vals_ = L_sensor_vals_
+        l_sensor_coords_ = L_sensor_coords_
+        l_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_ = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_
+        )
+        l_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_ = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_
+        )
+        l_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_ = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_
+        )
+        l_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_ = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_
+        )
+        l_self_modules_sensor_encoder_modules_mlp1_modules_1_eps = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_1_eps
+        )
+        l_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_ = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_
+        )
+        l_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_ = (
+            L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_
+        )
+        l_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_ = (
+            L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_
+        )
+        l_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_ = (
+            L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_
+        )
+        l_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_ = L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_
+        l_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_ = (
+            L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_
+        )
+        l_self_modules_sensor_encoder_modules_attn_dropout = (
+            L_self_modules_sensor_encoder_modules_attn_dropout
+        )
+        l_self_modules_sensor_encoder_modules_norm_parameters_weight_ = (
+            L_self_modules_sensor_encoder_modules_norm_parameters_weight_
+        )
+        l_self_modules_sensor_encoder_modules_norm_parameters_bias_ = (
+            L_self_modules_sensor_encoder_modules_norm_parameters_bias_
+        )
+        l_self_modules_sensor_encoder_modules_norm_eps = (
+            L_self_modules_sensor_encoder_modules_norm_eps
+        )
+        l_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_ = (
+            L_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_
+        )
+        l_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_ = (
+            L_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_
+        )
+        l_grid_coords_ = L_grid_coords_
+        l_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_ = (
+            L_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_
+        )
+        l_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_ = (
+            L_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_
+        )
+        l_self_modules_projector_transition_x = L_self_modules_projector_transition_x
+        l_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_ = L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_
+        l_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_ = L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_
+        g_np_pi = G_np_pi
+        l_self_modules_projector_modules_sensor_embedding_buffers_b_high_ = (
+            L_self_modules_projector_modules_sensor_embedding_buffers_B_high_
+        )
+        l_self_modules_projector_modules_key_proj_smooth_parameters_weight_ = (
+            L_self_modules_projector_modules_key_proj_smooth_parameters_weight_
+        )
+        l_self_modules_projector_modules_key_proj_smooth_parameters_bias_ = (
+            L_self_modules_projector_modules_key_proj_smooth_parameters_bias_
+        )
+        l_self_modules_projector_modules_key_proj_detail_parameters_weight_ = (
+            L_self_modules_projector_modules_key_proj_detail_parameters_weight_
+        )
+        l_self_modules_projector_modules_key_proj_detail_parameters_bias_ = (
+            L_self_modules_projector_modules_key_proj_detail_parameters_bias_
+        )
+        l_self_modules_projector_scale = L_self_modules_projector_scale
+        l_base_flow_ = L_base_flow_
+        l_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_1_momentum = (
+            L_self_modules_in_conv_modules_double_conv_modules_1_momentum
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_1_eps = (
+            L_self_modules_in_conv_modules_double_conv_modules_1_eps
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_4_momentum = (
+            L_self_modules_in_conv_modules_double_conv_modules_4_momentum
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_ = (
+            L_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_
+        )
+        l_self_modules_in_conv_modules_double_conv_modules_4_eps = (
+            L_self_modules_in_conv_modules_double_conv_modules_4_eps
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_ = L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_ = L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum = (
+            L_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_ = L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_ = L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_ = L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_ = L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_eps = (
+            L_self_modules_down1_modules_1_modules_double_conv_modules_1_eps
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_ = L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_ = L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum = (
+            L_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_ = L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_ = L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_ = L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_ = L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_eps = (
+            L_self_modules_down1_modules_1_modules_double_conv_modules_4_eps
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_ = L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_ = L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum = (
+            L_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_ = L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_ = L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_ = L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_ = L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_eps = (
+            L_self_modules_down2_modules_1_modules_double_conv_modules_1_eps
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_ = L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_ = L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum = (
+            L_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_ = L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_ = L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_ = L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_ = L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_eps = (
+            L_self_modules_down2_modules_1_modules_double_conv_modules_4_eps
+        )
+        l_self_modules_up1_scale_factor = L_self_modules_up1_scale_factor
+        l_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_1_momentum = (
+            L_self_modules_conv_up1_modules_double_conv_modules_1_momentum
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_1_eps = (
+            L_self_modules_conv_up1_modules_double_conv_modules_1_eps
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_4_momentum = (
+            L_self_modules_conv_up1_modules_double_conv_modules_4_momentum
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_ = (
+            L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_
+        )
+        l_self_modules_conv_up1_modules_double_conv_modules_4_eps = (
+            L_self_modules_conv_up1_modules_double_conv_modules_4_eps
+        )
+        l_self_modules_up2_scale_factor = L_self_modules_up2_scale_factor
+        l_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_1_momentum = (
+            L_self_modules_conv_up2_modules_double_conv_modules_1_momentum
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_1_eps = (
+            L_self_modules_conv_up2_modules_double_conv_modules_1_eps
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_4_momentum = (
+            L_self_modules_conv_up2_modules_double_conv_modules_4_momentum
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_ = (
+            L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_
+        )
+        l_self_modules_conv_up2_modules_double_conv_modules_4_eps = (
+            L_self_modules_conv_up2_modules_double_conv_modules_4_eps
+        )
+        l_self_modules_out_conv_parameters_weight_ = (
+            L_self_modules_out_conv_parameters_weight_
+        )
+        l_self_modules_out_conv_parameters_bias_ = (
+            L_self_modules_out_conv_parameters_bias_
+        )
+        node_in = torch.cat([l_sensor_vals_, l_sensor_coords_], dim=-1)
+        l_sensor_vals_ = None
+        input_1 = torch.nn.functional.linear(
+            node_in,
+            l_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_,
+            l_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_,
+        )
+        node_in = (
+            l_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_
+        ) = l_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_ = None
+        item = l_self_modules_sensor_encoder_modules_mlp1_modules_1_eps.item()
+        l_self_modules_sensor_encoder_modules_mlp1_modules_1_eps = None
+        input_2 = torch.nn.functional.layer_norm(
+            input_1,
+            (64,),
+            l_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_,
+            l_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_,
+            item,
+        )
+        input_1 = (
+            l_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_
+        ) = item = None
+        input_3 = torch.nn.functional.silu(input_2, inplace=False)
+        input_2 = None
+        input_4 = torch.nn.functional.linear(
+            input_3,
+            l_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_,
+            l_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_,
+        )
+        input_3 = (
+            l_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_
+        ) = l_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_ = None
+        query = input_4.transpose(1, 0)
+        item_1 = l_self_modules_sensor_encoder_modules_attn_dropout.item()
+        l_self_modules_sensor_encoder_modules_attn_dropout = None
+        multi_head_attention_forward = torch.nn.functional.multi_head_attention_forward(
+            query,
+            query,
+            query,
+            64,
+            4,
+            l_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_,
+            l_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_,
+            None,
+            None,
+            False,
+            item_1,
+            l_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_,
+            l_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_,
+            training=False,
+            key_padding_mask=None,
+            need_weights=True,
+            attn_mask=None,
+            average_attn_weights=True,
+            is_causal=False,
+        )
+        query = (
+            l_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_
+        ) = (
+            l_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_
+        ) = (
+            item_1
+        ) = l_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_ = (
+            l_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_
+        ) = None
+        attn_output = multi_head_attention_forward[0]
+        multi_head_attention_forward = None
+        h_attn = attn_output.transpose(1, 0)
+        attn_output = None
+        add = input_4 + h_attn
+        input_4 = h_attn = None
+        item_2 = l_self_modules_sensor_encoder_modules_norm_eps.item()
+        l_self_modules_sensor_encoder_modules_norm_eps = None
+        h = torch.nn.functional.layer_norm(
+            add,
+            (64,),
+            l_self_modules_sensor_encoder_modules_norm_parameters_weight_,
+            l_self_modules_sensor_encoder_modules_norm_parameters_bias_,
+            item_2,
+        )
+        add = (
+            l_self_modules_sensor_encoder_modules_norm_parameters_weight_
+        ) = l_self_modules_sensor_encoder_modules_norm_parameters_bias_ = item_2 = None
+        global_context = torch.mean(h, dim=1)
+        input_5 = torch.nn.functional.linear(
+            l_grid_coords_,
+            l_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_,
+            l_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_,
+        )
+        l_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_ = (
+            l_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_
+        ) = None
+        input_6 = torch.nn.functional.silu(input_5, inplace=False)
+        input_5 = None
+        input_7 = torch.nn.functional.linear(
+            input_6,
+            l_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_,
+            l_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_,
+        )
+        input_6 = (
+            l_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_
+        ) = l_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_ = None
+        Q_smooth = input_7[(Ellipsis, slice(None, 64, None))]
+        Q_detail = input_7[(Ellipsis, slice(64, None, None))]
+        input_7 = None
+        grid_x = l_grid_coords_[(Ellipsis, slice(0, 1, None))]
+        item_3 = l_self_modules_projector_transition_x.item()
+        l_self_modules_projector_transition_x = None
+        sub = grid_x - item_3
+        grid_x = item_3 = None
+        mul = 50.0 * sub
+        sub = None
+        beta = torch.sigmoid(mul)
+        mul = None
+        Q_detail_1 = Q_detail * beta
+        Q_detail = beta = None
+        linear_4 = torch.nn.functional.linear(
+            l_sensor_coords_,
+            l_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_,
+            l_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_,
+        )
+        l_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_ = l_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_ = (None)
+        feat_smooth = torch.nn.functional.silu(linear_4)
+        linear_4 = None
+        item_4 = g_np_pi.item()
+        g_np_pi = None
+        mul_2 = 2 * item_4
+        item_4 = None
+        mul_3 = mul_2 * l_sensor_coords_
+        mul_2 = l_sensor_coords_ = None
+        proj_high = (
+            mul_3 @ l_self_modules_projector_modules_sensor_embedding_buffers_b_high_
+        )
+        mul_3 = l_self_modules_projector_modules_sensor_embedding_buffers_b_high_ = None
+        sin = torch.sin(proj_high)
+        cos = torch.cos(proj_high)
+        proj_high = None
+        feat_detail = torch.cat([sin, cos], dim=-1)
+        sin = cos = None
+        K_smooth = torch.nn.functional.linear(
+            feat_smooth,
+            l_self_modules_projector_modules_key_proj_smooth_parameters_weight_,
+            l_self_modules_projector_modules_key_proj_smooth_parameters_bias_,
+        )
+        feat_smooth = (
+            l_self_modules_projector_modules_key_proj_smooth_parameters_weight_
+        ) = l_self_modules_projector_modules_key_proj_smooth_parameters_bias_ = None
+        K_detail = torch.nn.functional.linear(
+            feat_detail,
+            l_self_modules_projector_modules_key_proj_detail_parameters_weight_,
+            l_self_modules_projector_modules_key_proj_detail_parameters_bias_,
+        )
+        feat_detail = (
+            l_self_modules_projector_modules_key_proj_detail_parameters_weight_
+        ) = l_self_modules_projector_modules_key_proj_detail_parameters_bias_ = None
+        score_smooth = torch.functional.einsum("bhwc,bsc->bhws", Q_smooth, K_smooth)
+        Q_smooth = K_smooth = None
+        score_detail = torch.functional.einsum("bhwc,bsc->bhws", Q_detail_1, K_detail)
+        Q_detail_1 = K_detail = None
+        add_1 = score_smooth + score_detail
+        score_smooth = score_detail = None
+        item_5 = l_self_modules_projector_scale.item()
+        l_self_modules_projector_scale = None
+        attn_logits = add_1 * item_5
+        add_1 = item_5 = None
+        attn = torch.nn.functional.softmax(attn_logits, dim=-1)
+        attn_logits = None
+        out = torch.functional.einsum("bhws,bsc->bhwc", attn, h)
+        attn = h = None
+        grid_feats = out.permute(0, 3, 1, 2)
+        out = None
+        view = global_context.view(1, 64, 1, 1)
+        global_context = None
+        global_map = view.expand(1, 64, 128, 256)
+        view = None
+        coords_map = l_grid_coords_.permute(0, 3, 1, 2)
+        l_grid_coords_ = None
+        coords_map_1 = torch.nn.functional.interpolate(
+            coords_map, size=(128, 256), mode="bilinear", align_corners=True
+        )
+        coords_map = None
+        base_flow = torch.nn.functional.interpolate(
+            l_base_flow_, size=(128, 256), mode="bilinear", align_corners=True
+        )
+        l_base_flow_ = None
+        x = torch.cat([grid_feats, coords_map_1, global_map, base_flow], dim=1)
+        grid_feats = coords_map_1 = global_map = None
+        input_8 = torch.conv2d(
+            x,
+            l_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_,
+            l_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        x = (
+            l_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_
+        ) = l_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_ = None
+        item_6 = l_self_modules_in_conv_modules_double_conv_modules_1_momentum.item()
+        l_self_modules_in_conv_modules_double_conv_modules_1_momentum = None
+        item_7 = l_self_modules_in_conv_modules_double_conv_modules_1_eps.item()
+        l_self_modules_in_conv_modules_double_conv_modules_1_eps = None
+        input_9 = torch.nn.functional.batch_norm(
+            input_8,
+            l_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_,
+            l_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_,
+            l_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_,
+            l_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_,
+            False,
+            item_6,
+            item_7,
+        )
+        input_8 = (
+            l_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_
+        ) = (
+            l_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_
+        ) = (
+            l_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_
+        ) = item_6 = item_7 = None
+        input_10 = torch.nn.functional.silu(input_9, inplace=True)
+        input_9 = None
+        input_11 = torch.conv2d(
+            input_10,
+            l_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_,
+            l_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_10 = (
+            l_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_
+        ) = l_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_ = None
+        item_8 = l_self_modules_in_conv_modules_double_conv_modules_4_momentum.item()
+        l_self_modules_in_conv_modules_double_conv_modules_4_momentum = None
+        item_9 = l_self_modules_in_conv_modules_double_conv_modules_4_eps.item()
+        l_self_modules_in_conv_modules_double_conv_modules_4_eps = None
+        input_12 = torch.nn.functional.batch_norm(
+            input_11,
+            l_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_,
+            l_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_,
+            l_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_,
+            l_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_,
+            False,
+            item_8,
+            item_9,
+        )
+        input_11 = (
+            l_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_
+        ) = (
+            l_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_
+        ) = (
+            l_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_
+        ) = item_8 = item_9 = None
+        input_13 = torch.nn.functional.silu(input_12, inplace=True)
+        input_12 = None
+        input_14 = torch.nn.functional.max_pool2d(
+            input_13, 2, 2, 0, 1, ceil_mode=False, return_indices=False
+        )
+        input_15 = torch.conv2d(
+            input_14,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_14 = l_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_ = l_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_ = (None)
+        item_10 = (
+            l_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum.item()
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum = None
+        item_11 = (
+            l_self_modules_down1_modules_1_modules_double_conv_modules_1_eps.item()
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_1_eps = None
+        input_16 = torch.nn.functional.batch_norm(
+            input_15,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_,
+            False,
+            item_10,
+            item_11,
+        )
+        input_15 = l_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_ = l_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_ = l_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_ = l_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_ = (item_10) = (
+            item_11
+        ) = None
+        input_17 = torch.nn.functional.silu(input_16, inplace=True)
+        input_16 = None
+        input_18 = torch.conv2d(
+            input_17,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_17 = l_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_ = l_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_ = (None)
+        item_12 = (
+            l_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum.item()
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum = None
+        item_13 = (
+            l_self_modules_down1_modules_1_modules_double_conv_modules_4_eps.item()
+        )
+        l_self_modules_down1_modules_1_modules_double_conv_modules_4_eps = None
+        input_19 = torch.nn.functional.batch_norm(
+            input_18,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_,
+            l_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_,
+            False,
+            item_12,
+            item_13,
+        )
+        input_18 = l_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_ = l_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_ = l_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_ = l_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_ = (item_12) = (
+            item_13
+        ) = None
+        input_20 = torch.nn.functional.silu(input_19, inplace=True)
+        input_19 = None
+        input_21 = torch.nn.functional.max_pool2d(
+            input_20, 2, 2, 0, 1, ceil_mode=False, return_indices=False
+        )
+        input_22 = torch.conv2d(
+            input_21,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_21 = l_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_ = l_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_ = (None)
+        item_14 = (
+            l_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum.item()
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum = None
+        item_15 = (
+            l_self_modules_down2_modules_1_modules_double_conv_modules_1_eps.item()
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_1_eps = None
+        input_23 = torch.nn.functional.batch_norm(
+            input_22,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_,
+            False,
+            item_14,
+            item_15,
+        )
+        input_22 = l_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_ = l_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_ = l_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_ = l_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_ = (item_14) = (
+            item_15
+        ) = None
+        input_24 = torch.nn.functional.silu(input_23, inplace=True)
+        input_23 = None
+        input_25 = torch.conv2d(
+            input_24,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_24 = l_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_ = l_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_ = (None)
+        item_16 = (
+            l_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum.item()
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum = None
+        item_17 = (
+            l_self_modules_down2_modules_1_modules_double_conv_modules_4_eps.item()
+        )
+        l_self_modules_down2_modules_1_modules_double_conv_modules_4_eps = None
+        input_26 = torch.nn.functional.batch_norm(
+            input_25,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_,
+            l_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_,
+            False,
+            item_16,
+            item_17,
+        )
+        input_25 = l_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_ = l_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_ = l_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_ = l_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_ = (item_16) = (
+            item_17
+        ) = None
+        input_27 = torch.nn.functional.silu(input_26, inplace=True)
+        input_26 = None
+        item_18 = l_self_modules_up1_scale_factor.item()
+        l_self_modules_up1_scale_factor = None
+        x_1 = torch.nn.functional.interpolate(
+            input_27, None, item_18, "bilinear", True, recompute_scale_factor=None
+        )
+        input_27 = item_18 = None
+        x_2 = torch.cat([x_1, input_20], dim=1)
+        x_1 = input_20 = None
+        input_28 = torch.conv2d(
+            x_2,
+            l_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_,
+            l_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        x_2 = (
+            l_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_
+        ) = None
+        item_19 = l_self_modules_conv_up1_modules_double_conv_modules_1_momentum.item()
+        l_self_modules_conv_up1_modules_double_conv_modules_1_momentum = None
+        item_20 = l_self_modules_conv_up1_modules_double_conv_modules_1_eps.item()
+        l_self_modules_conv_up1_modules_double_conv_modules_1_eps = None
+        input_29 = torch.nn.functional.batch_norm(
+            input_28,
+            l_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_,
+            l_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_,
+            l_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_,
+            l_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_,
+            False,
+            item_19,
+            item_20,
+        )
+        input_28 = (
+            l_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_
+        ) = item_19 = item_20 = None
+        input_30 = torch.nn.functional.silu(input_29, inplace=True)
+        input_29 = None
+        input_31 = torch.conv2d(
+            input_30,
+            l_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_,
+            l_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_30 = (
+            l_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_
+        ) = None
+        item_21 = l_self_modules_conv_up1_modules_double_conv_modules_4_momentum.item()
+        l_self_modules_conv_up1_modules_double_conv_modules_4_momentum = None
+        item_22 = l_self_modules_conv_up1_modules_double_conv_modules_4_eps.item()
+        l_self_modules_conv_up1_modules_double_conv_modules_4_eps = None
+        input_32 = torch.nn.functional.batch_norm(
+            input_31,
+            l_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_,
+            l_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_,
+            l_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_,
+            l_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_,
+            False,
+            item_21,
+            item_22,
+        )
+        input_31 = (
+            l_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_
+        ) = item_21 = item_22 = None
+        input_33 = torch.nn.functional.silu(input_32, inplace=True)
+        input_32 = None
+        item_23 = l_self_modules_up2_scale_factor.item()
+        l_self_modules_up2_scale_factor = None
+        x_3 = torch.nn.functional.interpolate(
+            input_33, None, item_23, "bilinear", True, recompute_scale_factor=None
+        )
+        input_33 = item_23 = None
+        x_4 = torch.cat([x_3, input_13], dim=1)
+        x_3 = input_13 = None
+        input_34 = torch.conv2d(
+            x_4,
+            l_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_,
+            l_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        x_4 = (
+            l_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_
+        ) = None
+        item_24 = l_self_modules_conv_up2_modules_double_conv_modules_1_momentum.item()
+        l_self_modules_conv_up2_modules_double_conv_modules_1_momentum = None
+        item_25 = l_self_modules_conv_up2_modules_double_conv_modules_1_eps.item()
+        l_self_modules_conv_up2_modules_double_conv_modules_1_eps = None
+        input_35 = torch.nn.functional.batch_norm(
+            input_34,
+            l_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_,
+            l_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_,
+            l_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_,
+            l_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_,
+            False,
+            item_24,
+            item_25,
+        )
+        input_34 = (
+            l_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_
+        ) = item_24 = item_25 = None
+        input_36 = torch.nn.functional.silu(input_35, inplace=True)
+        input_35 = None
+        input_37 = torch.conv2d(
+            input_36,
+            l_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_,
+            l_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_,
+            (1, 1),
+            (1, 1),
+            (1, 1),
+            1,
+        )
+        input_36 = (
+            l_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_
+        ) = None
+        item_26 = l_self_modules_conv_up2_modules_double_conv_modules_4_momentum.item()
+        l_self_modules_conv_up2_modules_double_conv_modules_4_momentum = None
+        item_27 = l_self_modules_conv_up2_modules_double_conv_modules_4_eps.item()
+        l_self_modules_conv_up2_modules_double_conv_modules_4_eps = None
+        input_38 = torch.nn.functional.batch_norm(
+            input_37,
+            l_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_,
+            l_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_,
+            l_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_,
+            l_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_,
+            False,
+            item_26,
+            item_27,
+        )
+        input_37 = (
+            l_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_
+        ) = (
+            l_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_
+        ) = item_26 = item_27 = None
+        input_39 = torch.nn.functional.silu(input_38, inplace=True)
+        input_38 = None
+        fluc_out = torch.conv2d(
+            input_39,
+            l_self_modules_out_conv_parameters_weight_,
+            l_self_modules_out_conv_parameters_bias_,
+            (1, 1),
+            (0, 0),
+            (1, 1),
+            1,
+        )
+        input_39 = (
+            l_self_modules_out_conv_parameters_weight_
+        ) = l_self_modules_out_conv_parameters_bias_ = None
+        add_2 = fluc_out + base_flow
+        fluc_out = base_flow = None
+        return (add_2,)
diff --git a/samples/pigu_hybrid/weight_meta.py b/samples/pigu_hybrid/weight_meta.py
new file mode 100644
index 000000000..08fcc32e3
--- /dev/null
+++ b/samples/pigu_hybrid/weight_meta.py
@@ -0,0 +1,1231 @@
+class Program_weight_tensor_meta_s10:
+    name = "s10"
+    shape = []
+    dtype = "torch.int64"
+    device = "cpu"
+    mean = None
+    std = None
+    data = [4]
+
+
+class Program_weight_tensor_meta_L_sensor_vals_:
+    name = "L_sensor_vals_"
+    shape = [1, 65, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.023
+    std = 0.933
+    data = None
+
+
+class Program_weight_tensor_meta_L_sensor_coords_:
+    name = "L_sensor_coords_"
+    shape = [1, 65, 2]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.012
+    std = 0.964
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_weight_"
+    shape = [64, 5]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.013
+    std = 0.249
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_0_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.003
+    std = 0.258
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_weight_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_1_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_1_eps:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_1_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_weight_"
+    shape = [64, 64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.002
+    std = 0.072
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_:
+    name = "L_self_modules_sensor_encoder_modules_mlp1_modules_3_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.002
+    std = 0.067
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_:
+    name = "L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_bias_"
+    shape = [192]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_:
+    name = "L_self_modules_sensor_encoder_modules_attn_parameters_in_proj_weight_"
+    shape = [192, 64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.001
+    std = 0.088
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_:
+    name = (
+        "L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_weight_"
+    )
+    shape = [64, 64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.072
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_:
+    name = (
+        "L_self_modules_sensor_encoder_modules_attn_modules_out_proj_parameters_bias_"
+    )
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_attn_dropout:
+    name = "L_self_modules_sensor_encoder_modules_attn_dropout"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000000]
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_norm_parameters_weight_:
+    name = "L_self_modules_sensor_encoder_modules_norm_parameters_weight_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_norm_parameters_bias_:
+    name = "L_self_modules_sensor_encoder_modules_norm_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_sensor_encoder_modules_norm_eps:
+    name = "L_self_modules_sensor_encoder_modules_norm_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_:
+    name = "L_self_modules_projector_modules_query_mlp_modules_0_parameters_weight_"
+    shape = [64, 2]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.072
+    std = 0.392
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_:
+    name = "L_self_modules_projector_modules_query_mlp_modules_0_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.033
+    std = 0.412
+    data = None
+
+
+class Program_weight_tensor_meta_L_grid_coords_:
+    name = "L_grid_coords_"
+    shape = [1, 128, 256, 2]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.008
+    std = 1.002
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_:
+    name = "L_self_modules_projector_modules_query_mlp_modules_2_parameters_weight_"
+    shape = [128, 64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.001
+    std = 0.072
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_:
+    name = "L_self_modules_projector_modules_query_mlp_modules_2_parameters_bias_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.006
+    std = 0.075
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_transition_x:
+    name = "L_self_modules_projector_transition_x"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.250
+    std = 0.000
+    data = [0.250000]
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_:
+    name = "L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_weight_"
+    shape = [32, 2]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.071
+    std = 0.406
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_:
+    name = "L_self_modules_projector_modules_sensor_embedding_modules_raw_proj_parameters_bias_"
+    shape = [32]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.044
+    std = 0.406
+    data = [
+        -0.109570,
+        -0.500612,
+        0.499090,
+        -0.595455,
+        -0.314464,
+        0.311158,
+        0.427396,
+        -0.394249,
+        0.155079,
+        0.525865,
+        0.152357,
+        0.517181,
+        0.421748,
+        -0.172479,
+        0.703390,
+        -0.336485,
+        -0.533521,
+        0.561451,
+        -0.191133,
+        -0.205178,
+        -0.142233,
+        -0.455833,
+        -0.541973,
+        -0.376394,
+        -0.008022,
+        0.528279,
+        0.360097,
+        0.309068,
+        0.316133,
+        -0.304521,
+        0.450108,
+        0.365386,
+    ]
+
+
+class Program_weight_tensor_meta_G_np_pi:
+    name = "G_np_pi"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 3.142
+    std = 0.000
+    data = [3.141593]
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_sensor_embedding_buffers_B_high_:
+    name = "L_self_modules_projector_modules_sensor_embedding_buffers_B_high_"
+    shape = [2, 32]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.810
+    std = 11.562
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_key_proj_smooth_parameters_weight_:
+    name = "L_self_modules_projector_modules_key_proj_smooth_parameters_weight_"
+    shape = [64, 32]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.004
+    std = 0.101
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_key_proj_smooth_parameters_bias_:
+    name = "L_self_modules_projector_modules_key_proj_smooth_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.008
+    std = 0.097
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_key_proj_detail_parameters_weight_:
+    name = "L_self_modules_projector_modules_key_proj_detail_parameters_weight_"
+    shape = [64, 64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.001
+    std = 0.072
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_modules_key_proj_detail_parameters_bias_:
+    name = "L_self_modules_projector_modules_key_proj_detail_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.075
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_projector_scale:
+    name = "L_self_modules_projector_scale"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.125
+    std = 0.000
+    data = [0.125000]
+
+
+class Program_weight_tensor_meta_L_base_flow_:
+    name = "L_base_flow_"
+    shape = [1, 3, 128, 256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.002
+    std = 0.998
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_0_parameters_weight_"
+    shape = [64, 133, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_0_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.001
+    std = 0.016
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_1_momentum:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_1_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_mean_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_1_buffers_running_var_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_1_parameters_weight_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_1_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_1_eps:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_1_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_3_parameters_weight_"
+    shape = [64, 64, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.024
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_3_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.001
+    std = 0.022
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_4_momentum:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_4_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_mean_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_4_buffers_running_var_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_4_parameters_weight_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_4_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_in_conv_modules_double_conv_modules_4_eps:
+    name = "L_self_modules_in_conv_modules_double_conv_modules_4_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_weight_"
+    shape = [128, 64, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.024
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_:
+    name = (
+        "L_self_modules_down1_modules_1_modules_double_conv_modules_0_parameters_bias_"
+    )
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.001
+    std = 0.024
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_1_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_mean_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_1_buffers_running_var_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_weight_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_:
+    name = (
+        "L_self_modules_down1_modules_1_modules_double_conv_modules_1_parameters_bias_"
+    )
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_1_eps:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_1_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_weight_"
+    shape = [128, 128, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_:
+    name = (
+        "L_self_modules_down1_modules_1_modules_double_conv_modules_3_parameters_bias_"
+    )
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_4_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_mean_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_4_buffers_running_var_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_weight_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_:
+    name = (
+        "L_self_modules_down1_modules_1_modules_double_conv_modules_4_parameters_bias_"
+    )
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down1_modules_1_modules_double_conv_modules_4_eps:
+    name = "L_self_modules_down1_modules_1_modules_double_conv_modules_4_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_weight_"
+    shape = [256, 128, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_:
+    name = (
+        "L_self_modules_down2_modules_1_modules_double_conv_modules_0_parameters_bias_"
+    )
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.001
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_1_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_mean_"
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_1_buffers_running_var_"
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_weight_"
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_:
+    name = (
+        "L_self_modules_down2_modules_1_modules_double_conv_modules_1_parameters_bias_"
+    )
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_1_eps:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_1_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_weight_"
+    shape = [256, 256, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.012
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_:
+    name = (
+        "L_self_modules_down2_modules_1_modules_double_conv_modules_3_parameters_bias_"
+    )
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.002
+    std = 0.012
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_4_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_mean_"
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_4_buffers_running_var_"
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_weight_"
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_:
+    name = (
+        "L_self_modules_down2_modules_1_modules_double_conv_modules_4_parameters_bias_"
+    )
+    shape = [256]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_down2_modules_1_modules_double_conv_modules_4_eps:
+    name = "L_self_modules_down2_modules_1_modules_double_conv_modules_4_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_up1_scale_factor:
+    name = "L_self_modules_up1_scale_factor"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 2.000
+    std = 0.000
+    data = [2.000000]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_weight_"
+    shape = [128, 384, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.010
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_0_parameters_bias_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.001
+    std = 0.011
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_1_momentum:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_1_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_mean_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_1_buffers_running_var_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_weight_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_1_parameters_bias_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_1_eps:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_1_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_weight_"
+    shape = [128, 128, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_3_parameters_bias_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.001
+    std = 0.017
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_4_momentum:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_4_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_mean_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_4_buffers_running_var_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_weight_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_4_parameters_bias_"
+    shape = [128]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up1_modules_double_conv_modules_4_eps:
+    name = "L_self_modules_conv_up1_modules_double_conv_modules_4_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_up2_scale_factor:
+    name = "L_self_modules_up2_scale_factor"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 2.000
+    std = 0.000
+    data = [2.000000]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_weight_"
+    shape = [64, 192, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.000
+    std = 0.014
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_0_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = -0.001
+    std = 0.014
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_1_momentum:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_1_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_mean_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_1_buffers_running_var_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_weight_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_1_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_1_eps:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_1_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_weight_"
+    shape = [64, 64, 3, 3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.024
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_3_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.022
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_4_momentum:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_4_momentum"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.100
+    std = 0.000
+    data = [0.100000]
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_mean_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_4_buffers_running_var_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_weight_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 1.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_4_parameters_bias_"
+    shape = [64]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_conv_up2_modules_double_conv_modules_4_eps:
+    name = "L_self_modules_conv_up2_modules_double_conv_modules_4_eps"
+    shape = []
+    dtype = "torch.float64"
+    device = "cpu"
+    mean = 0.000
+    std = 0.000
+    data = [0.000010]
+
+
+class Program_weight_tensor_meta_L_self_modules_out_conv_parameters_weight_:
+    name = "L_self_modules_out_conv_parameters_weight_"
+    shape = [3, 64, 1, 1]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.001
+    std = 0.073
+    data = None
+
+
+class Program_weight_tensor_meta_L_self_modules_out_conv_parameters_bias_:
+    name = "L_self_modules_out_conv_parameters_bias_"
+    shape = [3]
+    dtype = "torch.float32"
+    device = "cpu"
+    mean = 0.019
+    std = 0.047
+    data = [0.064352, -0.028943, 0.022260]