made a much better UI

pages/3_📈_Arc_Functions.py → Arc_Functions.py

@@ -13,13 +13,14 @@ def run_streamlit_app():
     st.markdown(introduction)
     input_value = st.number_input("Input Value for arctan_node", value=1.0)
     
-    if st.button("Show Computational Graph"):
+    col1, col2 = st.columns(2)
+    with col1:
         image_path = main(input_value=input_value, mode="graph")
         if image_path and os.path.exists(image_path):
             st.image(image_path, caption="Model Graph Visualization")
         else:
             st.error("Error generating the graph.")
-    if st.button("Execute"):
+    with col2:
         results = main(input_value=input_value, mode="run")
         if results:
             st.write("### Results")

pages/2_👾_Newton_Law_Of_Cooling.py → Newton_Law_Of_Cooling.py

@@ -38,12 +38,12 @@ def run_simulation(mod_graph, duration=100, dt=0.1, mode=None, mod = None):
 
 def main():
     st.title("❄️ Newton Cooling Model Simulator")
-    with open("./markdowns/css.md") as file:
-        custom_styles = file.read()
+    # with open("./markdowns/css.md") as file:
+    #     custom_styles = file.read()
     with open("./markdowns/introduction.md") as file:
         introduction = file.read()
     st.markdown(introduction)
-    st.markdown(custom_styles, unsafe_allow_html=True)
+    # st.markdown(custom_styles, unsafe_allow_html=True)
 
     
     if "mod_graph" not in st.session_state:
@@ -57,9 +57,8 @@ def main():
     with col2:  
         duration = st.number_input("Duration", value=100)
         dt = st.number_input("Time Step", value=0.1)
-
-    button_label = "Re-Run" if "run_once" in st.session_state else "Run Simulation"
-    if st.button(button_label):
+    col1, col2 = st.columns(2)
+    with col1:
         update_model_parameters(cooling_coeff, T_a)
         times, temperatures = run_simulation(st.session_state.mod_graph, duration, dt, mode="run", mod = st.session_state.mod)
         times = times[1:]  # Adjust based on your needs
@@ -73,7 +72,7 @@ def main():
         # temperatures = temperatures  # Adjust based on your needs
         st.line_chart(chart_data, use_container_width=True, height=400, x="Time (s)", y="Temperature (K)")
         st.session_state.run_once = True
-    if st.button("Show Computational Graph"):
+    with col2:
         # from IPython.display import Image
         image_path = run_simulation(st.session_state.mod_graph, duration, dt, mode="graph", mod=st.session_state.mod)
         st.image(image_path, caption="Model Graph Visualization")

pages/1_∿_Sine_Approximator.py → Sine_Approximator.py

@@ -8,7 +8,7 @@ import onnx
 import pandas as pd
 import os
 import sys
-
+import time
 graphviz_bin_dir = os.path.join(sys.prefix, "bin")
 os.environ["PATH"] += os.pathsep + graphviz_bin_dir
 
@@ -26,6 +26,9 @@ class SineApproximator(nn.Module):
     def forward(self, x):
         return self.network(x)
 
+from time import sleep
+
+# @st.cache(allow_output_mutation=True, suppress_st_warning=True)
 def train_sine_model(model, epochs=1000, learning_rate=0.01):
     X = torch.unsqueeze(torch.linspace(-np.pi, np.pi, 200), dim=1)
     Y = torch.sin(X)
@@ -33,6 +36,10 @@ def train_sine_model(model, epochs=1000, learning_rate=0.01):
     loss_func = nn.MSELoss()
     optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)
 
+    progress_bar = st.progress(0)
+    
+
+
     for epoch in range(epochs):
         prediction = model(X)
         loss = loss_func(prediction, Y)
@@ -41,10 +48,15 @@ def train_sine_model(model, epochs=1000, learning_rate=0.01):
         loss.backward()
         optimizer.step()
 
-        # if epoch % 100 == 0:
-            # st.write(f'Epoch {epoch}, Loss {loss.item()}')
+        # Update progress bar and status text
+        progress_bar.progress((epoch + 1) / epochs)
 
-    st.success("Training complete.")
+        
+        # Dummy sleep to simulate long-running task
+        sleep(0.01)
+
+    progress_bar.empty()
+    return model
 
 def export_to_onnx(model):
     dummy_input = torch.tensor([[0.0]])
@@ -72,6 +84,7 @@ def plot_sine_curve(model):
 
 
 def main():
+    
     # Streamlit UI setup
     st.title("Sine Function Approximator")
 
@@ -89,34 +102,34 @@ def main():
     epochs = st.slider("Epochs", min_value=0, max_value=100, value=50)
     graph_col1, graph_col2 = st.columns(2)
     with graph_col1:
-        if st.button("Train Model"):
-            model = SineApproximator()
-            train_sine_model(model, epochs, learning_rate)
-            export_to_onnx(model)
-            plot_sine_curve(model)
-            onnx_model = onnx.load("./sine_approximator.onnx")
-            mdf_model = onnx_to_mdf(onnx_model)
-            mdf_model.to_json_file("sine_approximator.json")
-            mdf_model.to_yaml_file("sine_approximator.yaml")
-            st.success("Conversion to MDF complete. Files saved as sine_approximator.json and sine_approximator.yaml")
+        status_text = st.empty()
+        model = SineApproximator()
+        train_sine_model(model, epochs, learning_rate)
+        export_to_onnx(model)
+        plot_sine_curve(model)
+        onnx_model = onnx.load("./sine_approximator.onnx")
+        mdf_model = onnx_to_mdf(onnx_model)
+        mdf_model.to_json_file("sine_approximator.json")
+        mdf_model.to_yaml_file("sine_approximator.yaml")
+        status_text.success("Training complete.")
     with graph_col2:
-        if st.button("Generate Graph"):
-            onnx_model = onnx.load("sine_approximator.onnx")
-            mdf_model = onnx_to_mdf(onnx_model)
-            mdf_model.to_graph_image(
-                engine="dot",
-                output_format="png",
-                view_on_render=False,
-                level=3,
-                filename_root="sine_approximator",
-                only_warn_on_fail=True,
-            )
-            # st.image("sine_approximator.png", caption="Model Graph Visualization")
-            from IPython.display import Image
-            Image(filename="sine_approximator.png")
-            image_path = "./sine_approximator.png"
-            st.image(image_path, caption="Model Graph Visualization")
-            st.success('Graph generated successfully.')
+        onnx_model = onnx.load("sine_approximator.onnx")
+        mdf_model = onnx_to_mdf(onnx_model)
+        mdf_model.to_graph_image(
+            engine="dot",
+            output_format="png",
+            view_on_render=False,
+            level=3,
+            filename_root="sine_approximator",
+            only_warn_on_fail=True,
+        )
+        # st.image("sine_approximator.png", caption="Model Graph Visualization")
+        from IPython.display import Image
+        Image(filename="sine_approximator.png")
+        image_path = "./sine_approximator.png"
+        st.success('Graph generated successfully.')
+        st.image(image_path, caption="Model Graph Visualization")
+        
 if __name__ == "__main__":
     main()
 

app.py

@@ -31,11 +31,16 @@ with col3:
     st.image(gmail_icon, width=30)
     st.markdown("[Email](mailto:rimjhimittal2003@gmail.com)", unsafe_allow_html=True)
 
+def hide_pages(pages_to_hide):
+    for page in pages_to_hide:
+        st.sidebar.markdown(f"## {page}")
+
 def main():
+    hide_pages(["Arc_Functions", "Newton Cooling Model Simulator", "Arc Functions"])
     with open("./markdowns/css.md") as file:
         custom_styles = file.read()
     st.markdown(custom_styles, unsafe_allow_html=True)
-
+    
 
 if __name__ == "__main__":
     main()

arc_functions

@@ -0,0 +1,10 @@
+digraph arc_functions {
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	arctan_node [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>arctan_node</b></td></tr><tr><td><font color="#1666ff">input_value</font> = 1.0</td></tr><tr><td><font color="#1666ff">arctan_result</font> = <i>arctan</i>(<font color="#1666ff">input_value</font>, 0.4)</td></tr><tr><td><i>arctan(variable0, scale) = scale * arctan(variable0)</i></td></tr><tr><td><font color="#cc3355">output</font> = <font color="#1666ff">arctan_result</font> </td></tr></table>>]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	arccos_node [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>arccos_node</b></td></tr><tr><td><font color="#188855">input_from_arctan</font> </td></tr><tr><td><font color="#1666ff">arccos_result</font> = <i>arccos</i>(<font color="#188855">input_from_arctan</font>, 0.3)</td></tr><tr><td><i>arccos(variable0, scale) = scale * arccos(variable0)</i></td></tr><tr><td><font color="#cc3355">output</font> = <font color="#1666ff">arccos_result</font> </td></tr></table>>]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	arcsin_node [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>arcsin_node</b></td></tr><tr><td><font color="#188855">input_from_arccos</font> </td></tr><tr><td><font color="#1666ff">arcsin_result</font> = <i>arcsin</i>(<font color="#188855">input_from_arccos</font>, 0.2)</td></tr><tr><td><i>arcsin(variable0, scale) = scale * arcsin(variable0)</i></td></tr><tr><td><font color="#cc3355">output</font> = <font color="#1666ff">arcsin_result</font> </td></tr></table>>]
+	arctan_node -> arccos_node [label=<h (<font color="#cc3355">output</font> -&gt; <font color="#188855">input_from_arctan</font>)> arrowhead=empty]
+	arccos_node -> arcsin_node [label=<i (<font color="#cc3355">output</font> -&gt; <font color="#188855">input_from_arccos</font>)> arrowhead=empty]
+}

markdowns/css.md

@@ -1,5 +1,5 @@
 <style>
-
+/* 
 .stApp {
             background-image: linear-gradient(to right, rgba(13, 19, 48, 1) 0%, rgba(30, 37, 88, 1) 50%, #084349 100%), url("https://dl.dropboxusercontent.com/scl/fi/vrlu5xm5uh7nj6tor97my/glow_bg-7-1.png?rlkey=raqq4j45y0ogyuvu4rgvjcc3a&dl=1");
             background-size: cover, cover;
@@ -39,7 +39,7 @@
     background-image: linear-gradient(to right, #5456F7, #27B1FF) !important;
     border: none !important;
     color: white !important;
-}   
+}    */
 
 </style>
 

model_graph

@@ -0,0 +1,5 @@
+digraph cooling_process {
+	graph [rankdir=LR]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	node1 [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>node1</b></td></tr><tr><td><font color="#1666ff">cooling_coeff</font> = 0.1</td></tr><tr><td><font color="#1666ff">T_a</font> = 20</td></tr><tr><td><b><font color="#1666ff">T_curr</font></b> = <i>def init value:</i> 90, <i>d/dt:</i> <font color="#1666ff">dT_dt</font></td></tr><tr><td><b><font color="#1666ff">dT_dt</font></b> = -<font color="#1666ff">cooling_coeff</font>*(<font color="#1666ff">T_curr</font> - <font color="#1666ff">T_a</font>)<i>def init value:</i> 0</td></tr><tr><td><font color="#cc3355">out_port</font> = <font color="#1666ff">T_curr</font> </td></tr></table>>]
+}

models/arc_functions.py

@@ -87,7 +87,6 @@ def main(input_value=1.0, mode=None):
                 view_on_render=False,
                 level=3,
                 filename_root="arc_functions",
-                is_horizontal=True
             )
 
         return image_path

pages/1_Model_Category_1.py

@@ -0,0 +1,24 @@
+import streamlit as st
+from importlib import import_module
+
+# Page configuration
+st.set_page_config(page_title="Dynamic Model Selection", layout="wide")
+
+# Model names and the corresponding module functions to be imported
+model_functions = {
+    "Sine_Approximator": "main",
+    "Arc_Functions": "run_streamlit_app"
+}
+
+# Define the sidebar selection dropdown
+model_name = st.sidebar.selectbox('Choose a model:', options=list(model_functions.keys()))
+
+# Function to dynamically import and run the selected model's main function
+def run_model(model_key):
+    module = import_module(model_key)
+    main_function = getattr(module, model_functions[model_key])
+    main_function()
+
+# Run the selected model when the button is pressed
+if st.sidebar.button('Run Model'):
+    run_model(model_name)

pages/2_Model_Category_2.py

@@ -0,0 +1,24 @@
+import streamlit as st
+from importlib import import_module
+
+# Page configuration
+st.set_page_config(page_title="Dynamic Model Selection", layout="wide")
+
+# Model names and the corresponding module functions to be imported
+model_functions = {
+    "Newton_Law_Of_Cooling": "main",
+    "Arc_Functions": "run_streamlit_app"
+}
+
+# Define the sidebar selection dropdown
+model_name = st.sidebar.selectbox('Choose a model:', options=list(model_functions.keys()))
+
+# Function to dynamically import and run the selected model's main function
+def run_model(model_key):
+    module = import_module(model_key)
+    main_function = getattr(module, model_functions[model_key])
+    main_function()
+
+# Run the selected model when the button is pressed
+if st.sidebar.button('Run Model'):
+    run_model(model_name)

sine_approximator

@@ -0,0 +1,16 @@
+digraph main_graph {
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	"/network/network.0/Gemm" [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>/network/network.0/Gemm</b></td></tr><tr><td><font color="#188855">input</font> (shape: (1, 1), type: float)</td></tr><tr><td><font color="#188855">network.0.weight</font> (shape: (64, 1), type: float)</td></tr><tr><td><font color="#188855">network.0.bias</font> (shape: (64,), type: float)</td></tr><tr><td><font color="#1666ff">alpha</font> = 1.0</td></tr><tr><td><font color="#1666ff">beta</font> = 1.0</td></tr><tr><td><font color="#1666ff">transB</font> = 1</td></tr><tr><td><font color="#1666ff">/network/network.0/Gemm</font> = <i>onnx::Gemm</i>(<font color="#188855">input</font>, <font color="#188855">network.0.weight</font>, <font color="#188855">network.0.bias</font>)</td></tr><tr><td><i>onnx::Gemm(A, B, C) = onnx_ops.gemm(A, B, C, alpha, beta, transA, transB)</i></td></tr><tr><td><font color="#cc3355">_network_network.0_Gemm_output_0</font> = <font color="#1666ff">/network/network.0/Gemm</font> </td></tr></table>>]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	"/network/network.1/Relu" [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>/network/network.1/Relu</b></td></tr><tr><td><font color="#188855">_network_network.0_Gemm_output_0</font> (shape: (1, 64), type: float)</td></tr><tr><td><font color="#1666ff">/network/network.1/Relu</font> = <i>onnx::Relu</i>(<font color="#188855">_network_network.0_Gemm_output_0</font>)</td></tr><tr><td><i>onnx::Relu(X) = onnx_ops.relu(X)</i></td></tr><tr><td><font color="#cc3355">_network_network.1_Relu_output_0</font> = <font color="#1666ff">/network/network.1/Relu</font> </td></tr></table>>]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	"/network/network.2/Gemm" [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>/network/network.2/Gemm</b></td></tr><tr><td><font color="#188855">_network_network.1_Relu_output_0</font> (shape: (1, 64), type: float)</td></tr><tr><td><font color="#188855">network.2.weight</font> (shape: (64, 64), type: float)</td></tr><tr><td><font color="#188855">network.2.bias</font> (shape: (64,), type: float)</td></tr><tr><td><font color="#1666ff">alpha</font> = 1.0</td></tr><tr><td><font color="#1666ff">beta</font> = 1.0</td></tr><tr><td><font color="#1666ff">transB</font> = 1</td></tr><tr><td><font color="#1666ff">/network/network.2/Gemm</font> = <i>onnx::Gemm</i>(<font color="#188855">_network_network.1_Relu_output_0</font>, <font color="#188855">network.2.weight</font>, <font color="#188855">network.2.bias</font>)</td></tr><tr><td><i>onnx::Gemm(A, B, C) = onnx_ops.gemm(A, B, C, alpha, beta, transA, transB)</i></td></tr><tr><td><font color="#cc3355">_network_network.2_Gemm_output_0</font> = <font color="#1666ff">/network/network.2/Gemm</font> </td></tr></table>>]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	"/network/network.3/Relu" [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>/network/network.3/Relu</b></td></tr><tr><td><font color="#188855">_network_network.2_Gemm_output_0</font> (shape: (1, 64), type: float)</td></tr><tr><td><font color="#1666ff">/network/network.3/Relu</font> = <i>onnx::Relu</i>(<font color="#188855">_network_network.2_Gemm_output_0</font>)</td></tr><tr><td><i>onnx::Relu(X) = onnx_ops.relu(X)</i></td></tr><tr><td><font color="#cc3355">_network_network.3_Relu_output_0</font> = <font color="#1666ff">/network/network.3/Relu</font> </td></tr></table>>]
+	node [color="#444444" fontcolor="#444444" penwidth=1 shape=box style=rounded]
+	"/network/network.4/Gemm" [label=<<table border="0" cellborder="0"><tr><td colspan="2"><b>/network/network.4/Gemm</b></td></tr><tr><td><font color="#188855">_network_network.3_Relu_output_0</font> (shape: (1, 64), type: float)</td></tr><tr><td><font color="#188855">network.4.weight</font> (shape: (1, 64), type: float)</td></tr><tr><td><font color="#188855">network.4.bias</font> (shape: (1,), type: float)</td></tr><tr><td><font color="#1666ff">alpha</font> = 1.0</td></tr><tr><td><font color="#1666ff">beta</font> = 1.0</td></tr><tr><td><font color="#1666ff">transB</font> = 1</td></tr><tr><td><font color="#1666ff">/network/network.4/Gemm</font> = <i>onnx::Gemm</i>(<font color="#188855">_network_network.3_Relu_output_0</font>, <font color="#188855">network.4.weight</font>, <font color="#188855">network.4.bias</font>)</td></tr><tr><td><i>onnx::Gemm(A, B, C) = onnx_ops.gemm(A, B, C, alpha, beta, transA, transB)</i></td></tr><tr><td><font color="#cc3355">output</font> = <font color="#1666ff">/network/network.4/Gemm</font> </td></tr></table>>]
+	"/network/network.0/Gemm" -> "/network/network.1/Relu" [label=</network/network.0/Gemm._network_network.0_Gemm_output_0_/network/network.1/Relu._network_network.0_Gemm_output_0 (<font color="#cc3355">_network_network.0_Gemm_output_0</font> -&gt; <font color="#188855">_network_network.0_Gemm_output_0</font>)> arrowhead=empty]
+	"/network/network.1/Relu" -> "/network/network.2/Gemm" [label=</network/network.1/Relu._network_network.1_Relu_output_0_/network/network.2/Gemm._network_network.1_Relu_output_0 (<font color="#cc3355">_network_network.1_Relu_output_0</font> -&gt; <font color="#188855">_network_network.1_Relu_output_0</font>)> arrowhead=empty]
+	"/network/network.2/Gemm" -> "/network/network.3/Relu" [label=</network/network.2/Gemm._network_network.2_Gemm_output_0_/network/network.3/Relu._network_network.2_Gemm_output_0 (<font color="#cc3355">_network_network.2_Gemm_output_0</font> -&gt; <font color="#188855">_network_network.2_Gemm_output_0</font>)> arrowhead=empty]
+	"/network/network.3/Relu" -> "/network/network.4/Gemm" [label=</network/network.3/Relu._network_network.3_Relu_output_0_/network/network.4/Gemm._network_network.3_Relu_output_0 (<font color="#cc3355">_network_network.3_Relu_output_0</font> -&gt; <font color="#188855">_network_network.3_Relu_output_0</font>)> arrowhead=empty]
+}

sine_approximator.json

@@ -0,0 +1,230 @@
+{
+    "ONNX Model": {
+        "format": "ModECI MDF v0.4",
+        "generating_application": "Python modeci-mdf v0.4.9",
+        "graphs": {
+            "main_graph": {
+                "nodes": {
+                    "/network/network.0/Gemm": {
+                        "input_ports": {
+                            "input": {
+                                "shape": [
+                                    1,
+                                    1
+                                ],
+                                "type": "float"
+                            },
+                            "network.0.weight": {
+                                "shape": [
+                                    64,
+                                    1
+                                ],
+                                "type": "float"
+                            },
+                            "network.0.bias": {
+                                "shape": [
+                                    64
+                                ],
+                                "type": "float"
+                            }
+                        },
+                        "parameters": {
+                            "alpha": {
+                                "value": 1.0
+                            },
+                            "beta": {
+                                "value": 1.0
+                            },
+                            "transB": {
+                                "value": 1
+                            },
+                            "/network/network.0/Gemm": {
+                                "function": "onnx::Gemm",
+                                "args": {
+                                    "A": "input",
+                                    "B": "network.0.weight",
+                                    "C": "network.0.bias"
+                                }
+                            }
+                        },
+                        "output_ports": {
+                            "_network_network.0_Gemm_output_0": {
+                                "value": "/network/network.0/Gemm"
+                            }
+                        }
+                    },
+                    "/network/network.1/Relu": {
+                        "input_ports": {
+                            "_network_network.0_Gemm_output_0": {
+                                "shape": [
+                                    1,
+                                    64
+                                ],
+                                "type": "float"
+                            }
+                        },
+                        "parameters": {
+                            "/network/network.1/Relu": {
+                                "function": "onnx::Relu",
+                                "args": {
+                                    "X": "_network_network.0_Gemm_output_0"
+                                }
+                            }
+                        },
+                        "output_ports": {
+                            "_network_network.1_Relu_output_0": {
+                                "value": "/network/network.1/Relu"
+                            }
+                        }
+                    },
+                    "/network/network.2/Gemm": {
+                        "input_ports": {
+                            "_network_network.1_Relu_output_0": {
+                                "shape": [
+                                    1,
+                                    64
+                                ],
+                                "type": "float"
+                            },
+                            "network.2.weight": {
+                                "shape": [
+                                    64,
+                                    64
+                                ],
+                                "type": "float"
+                            },
+                            "network.2.bias": {
+                                "shape": [
+                                    64
+                                ],
+                                "type": "float"
+                            }
+                        },
+                        "parameters": {
+                            "alpha": {
+                                "value": 1.0
+                            },
+                            "beta": {
+                                "value": 1.0
+                            },
+                            "transB": {
+                                "value": 1
+                            },
+                            "/network/network.2/Gemm": {
+                                "function": "onnx::Gemm",
+                                "args": {
+                                    "A": "_network_network.1_Relu_output_0",
+                                    "B": "network.2.weight",
+                                    "C": "network.2.bias"
+                                }
+                            }
+                        },
+                        "output_ports": {
+                            "_network_network.2_Gemm_output_0": {
+                                "value": "/network/network.2/Gemm"
+                            }
+                        }
+                    },
+                    "/network/network.3/Relu": {
+                        "input_ports": {
+                            "_network_network.2_Gemm_output_0": {
+                                "shape": [
+                                    1,
+                                    64
+                                ],
+                                "type": "float"
+                            }
+                        },
+                        "parameters": {
+                            "/network/network.3/Relu": {
+                                "function": "onnx::Relu",
+                                "args": {
+                                    "X": "_network_network.2_Gemm_output_0"
+                                }
+                            }
+                        },
+                        "output_ports": {
+                            "_network_network.3_Relu_output_0": {
+                                "value": "/network/network.3/Relu"
+                            }
+                        }
+                    },
+                    "/network/network.4/Gemm": {
+                        "input_ports": {
+                            "_network_network.3_Relu_output_0": {
+                                "shape": [
+                                    1,
+                                    64
+                                ],
+                                "type": "float"
+                            },
+                            "network.4.weight": {
+                                "shape": [
+                                    1,
+                                    64
+                                ],
+                                "type": "float"
+                            },
+                            "network.4.bias": {
+                                "shape": [
+                                    1
+                                ],
+                                "type": "float"
+                            }
+                        },
+                        "parameters": {
+                            "alpha": {
+                                "value": 1.0
+                            },
+                            "beta": {
+                                "value": 1.0
+                            },
+                            "transB": {
+                                "value": 1
+                            },
+                            "/network/network.4/Gemm": {
+                                "function": "onnx::Gemm",
+                                "args": {
+                                    "A": "_network_network.3_Relu_output_0",
+                                    "B": "network.4.weight",
+                                    "C": "network.4.bias"
+                                }
+                            }
+                        },
+                        "output_ports": {
+                            "output": {
+                                "value": "/network/network.4/Gemm"
+                            }
+                        }
+                    }
+                },
+                "edges": {
+                    "/network/network.0/Gemm._network_network.0_Gemm_output_0_/network/network.1/Relu._network_network.0_Gemm_output_0": {
+                        "sender": "/network/network.0/Gemm",
+                        "receiver": "/network/network.1/Relu",
+                        "sender_port": "_network_network.0_Gemm_output_0",
+                        "receiver_port": "_network_network.0_Gemm_output_0"
+                    },
+                    "/network/network.1/Relu._network_network.1_Relu_output_0_/network/network.2/Gemm._network_network.1_Relu_output_0": {
+                        "sender": "/network/network.1/Relu",
+                        "receiver": "/network/network.2/Gemm",
+                        "sender_port": "_network_network.1_Relu_output_0",
+                        "receiver_port": "_network_network.1_Relu_output_0"
+                    },
+                    "/network/network.2/Gemm._network_network.2_Gemm_output_0_/network/network.3/Relu._network_network.2_Gemm_output_0": {
+                        "sender": "/network/network.2/Gemm",
+                        "receiver": "/network/network.3/Relu",
+                        "sender_port": "_network_network.2_Gemm_output_0",
+                        "receiver_port": "_network_network.2_Gemm_output_0"
+                    },
+                    "/network/network.3/Relu._network_network.3_Relu_output_0_/network/network.4/Gemm._network_network.3_Relu_output_0": {
+                        "sender": "/network/network.3/Relu",
+                        "receiver": "/network/network.4/Gemm",
+                        "sender_port": "_network_network.3_Relu_output_0",
+                        "receiver_port": "_network_network.3_Relu_output_0"
+                    }
+                }
+            }
+        }
+    }
+}

sine_approximator.yaml

@@ -0,0 +1,153 @@
+ONNX Model:
+    format: ModECI MDF v0.4
+    generating_application: Python modeci-mdf v0.4.9
+    graphs:
+        main_graph:
+            nodes:
+                /network/network.0/Gemm:
+                    input_ports:
+                        input:
+                            shape:
+                            - 1
+                            - 1
+                            type: float
+                        network.0.weight:
+                            shape:
+                            - 64
+                            - 1
+                            type: float
+                        network.0.bias:
+                            shape:
+                            - 64
+                            type: float
+                    parameters:
+                        alpha:
+                            value: 1.0
+                        beta:
+                            value: 1.0
+                        transB:
+                            value: 1
+                        /network/network.0/Gemm:
+                            function: onnx::Gemm
+                            args:
+                                A: input
+                                B: network.0.weight
+                                C: network.0.bias
+                    output_ports:
+                        _network_network.0_Gemm_output_0:
+                            value: /network/network.0/Gemm
+                /network/network.1/Relu:
+                    input_ports:
+                        _network_network.0_Gemm_output_0:
+                            shape:
+                            - 1
+                            - 64
+                            type: float
+                    parameters:
+                        /network/network.1/Relu:
+                            function: onnx::Relu
+                            args:
+                                X: _network_network.0_Gemm_output_0
+                    output_ports:
+                        _network_network.1_Relu_output_0:
+                            value: /network/network.1/Relu
+                /network/network.2/Gemm:
+                    input_ports:
+                        _network_network.1_Relu_output_0:
+                            shape:
+                            - 1
+                            - 64
+                            type: float
+                        network.2.weight:
+                            shape:
+                            - 64
+                            - 64
+                            type: float
+                        network.2.bias:
+                            shape:
+                            - 64
+                            type: float
+                    parameters:
+                        alpha:
+                            value: 1.0
+                        beta:
+                            value: 1.0
+                        transB:
+                            value: 1
+                        /network/network.2/Gemm:
+                            function: onnx::Gemm
+                            args:
+                                A: _network_network.1_Relu_output_0
+                                B: network.2.weight
+                                C: network.2.bias
+                    output_ports:
+                        _network_network.2_Gemm_output_0:
+                            value: /network/network.2/Gemm
+                /network/network.3/Relu:
+                    input_ports:
+                        _network_network.2_Gemm_output_0:
+                            shape:
+                            - 1
+                            - 64
+                            type: float
+                    parameters:
+                        /network/network.3/Relu:
+                            function: onnx::Relu
+                            args:
+                                X: _network_network.2_Gemm_output_0
+                    output_ports:
+                        _network_network.3_Relu_output_0:
+                            value: /network/network.3/Relu
+                /network/network.4/Gemm:
+                    input_ports:
+                        _network_network.3_Relu_output_0:
+                            shape:
+                            - 1
+                            - 64
+                            type: float
+                        network.4.weight:
+                            shape:
+                            - 1
+                            - 64
+                            type: float
+                        network.4.bias:
+                            shape:
+                            - 1
+                            type: float
+                    parameters:
+                        alpha:
+                            value: 1.0
+                        beta:
+                            value: 1.0
+                        transB:
+                            value: 1
+                        /network/network.4/Gemm:
+                            function: onnx::Gemm
+                            args:
+                                A: _network_network.3_Relu_output_0
+                                B: network.4.weight
+                                C: network.4.bias
+                    output_ports:
+                        output:
+                            value: /network/network.4/Gemm
+            edges:
+                /network/network.0/Gemm._network_network.0_Gemm_output_0_/network/network.1/Relu._network_network.0_Gemm_output_0:
+                    sender: /network/network.0/Gemm
+                    receiver: /network/network.1/Relu
+                    sender_port: _network_network.0_Gemm_output_0
+                    receiver_port: _network_network.0_Gemm_output_0
+                /network/network.1/Relu._network_network.1_Relu_output_0_/network/network.2/Gemm._network_network.1_Relu_output_0:
+                    sender: /network/network.1/Relu
+                    receiver: /network/network.2/Gemm
+                    sender_port: _network_network.1_Relu_output_0
+                    receiver_port: _network_network.1_Relu_output_0
+                /network/network.2/Gemm._network_network.2_Gemm_output_0_/network/network.3/Relu._network_network.2_Gemm_output_0:
+                    sender: /network/network.2/Gemm
+                    receiver: /network/network.3/Relu
+                    sender_port: _network_network.2_Gemm_output_0
+                    receiver_port: _network_network.2_Gemm_output_0
+                /network/network.3/Relu._network_network.3_Relu_output_0_/network/network.4/Gemm._network_network.3_Relu_output_0:
+                    sender: /network/network.3/Relu
+                    receiver: /network/network.4/Gemm
+                    sender_port: _network_network.3_Relu_output_0
+                    receiver_port: _network_network.3_Relu_output_0