Fix some operators

aidge_export_arm_cortexm/_Aidge_Arm/kernels/BatchNorm/aidge_batchnorm2d_chw_float32.h

+#include <math.h>
+
+
+void aidge_batchnorm2d_chw_float32 (float* inputs,
+                                float* outputs,
+                                float* input_mean,
+                                float* input_var,
+                                float* scale,
+                                float* bias,
+                                float epsilon,
+                                const int nb_channels,
+                                const int channel_width, const int channel_height)
+{
+    int featureMapSize = channel_width * channel_height;
+    for (int ch = 0; ch < nb_channels; ++ch) 
+    {
+        int ioIndex = ch * featureMapSize;
+        for (int i = ioIndex; i < ioIndex + featureMapSize; i++){
+            outputs[i] = bias[ch];
+        }
+        float var =sqrt(input_var[ch] + epsilon);
+
+        for (int feature = 0; feature<featureMapSize; ++feature) {
+            outputs[ioIndex + feature] += scale[ch] * (inputs[ioIndex + feature]-input_mean[ch]) / var;
+        }
+    
+    }
+}
\ No newline at end of file

aidge_export_arm_cortexm/_Aidge_Arm/kernels/Div/aidge_div.h

+template <unsigned int SIZE, typename Input_T, typename Output_T>
+__attribute__((always_inline)) inline static
+void aidge_div(Input_T* __restrict input_a, Input_T* __restrict input_b, Output_T* __restrict output) {
+  for (unsigned int i = 0; i < SIZE; ++i) {
+      // Handle division by zero case
+    if(input_b[i] != static_cast<Input_T>(0)) {
+      output[i] = input_a[i] / input_b[i];
+    } else {
+      output[i] = static_cast<Output_T>(0); // or some other error handling ? 
+    }
+  }
+}
\ No newline at end of file

aidge_export_arm_cortexm/_Aidge_Arm/kernels/Div/aidge_div_float32.h

-
-
-void aidge_div_float32(float* input_a, 
-                       float* input_b, 
-                       float* output, 
-                       unsigned int size)
-{
-    for (unsigned int i = 0; i < size; ++i) {
-        output[i] = input_a[i] / input_b[i];
-    }
-}
\ No newline at end of file

aidge_export_arm_cortexm/_Aidge_Arm/kernels/MatMul/aidge_matmul_chw_float32.h

-void aidge_matmul_chw_float32(float* input_a, 
-                       float* input_b, 
+void aidge_matmul_chw_float32(const float* input_a, 
+                       const float* input_b, 
                        float* output,
-                       int dim_a[],
-                       int dim_b[],
-                       int output_Dim[],
-                       int size_dima,
-                       int size_dimb,
-                       int size_outputDim)
+                       const int dim_a[],
+                       const int dim_b[],
+                       const int output_Dim[],
+                       const int size_dima,
+                       const int size_dimb,
+                       const int size_outputDim)
 {
         //initialize arrays storing broadcasted(or not) dims
         int ndim_a[size_outputDim];     

aidge_export_arm_cortexm/_Aidge_Arm/kernels/Reshape/aidge_reshape.h

+#include <cstring>
+template <unsigned int SIZE, typename Input_T, typename Output_T>
+__attribute__((always_inline)) inline static
+void aidge_reshape(Input_T* __restrict input, Output_T* __restrict output, unsigned int size) {
+    // Copy the input data to the output data
+    std::memcpy(output, input, size * sizeof(float));
+}
\ No newline at end of file

aidge_export_arm_cortexm/_Aidge_Arm/kernels/Sigmoid/aidge_sigmoid.h

+
+template <unsigned int SIZE, typename Input_T, typename Output_T>
+__attribute__((always_inline)) inline static
+void aidge_sigmoid(Input_T* __restrict input, Output_T* __restrict output) {
+    for (unsigned int i = 0; i < SIZE; ++i) {
+        output[i] = 1 / ( 1 + exp(-input[i]) );
+    }
+}

aidge_export_arm_cortexm/_Aidge_Arm/kernels/Softmax/aidge_softmax_chw_float32.h

 #include <math.h>
 
-void aidge_softmax_chw_float32(float* inputs, 
+void aidge_softmax_chw_float32(const float* inputs, 
                             float* outputs,
                             const int inputDims[],
                             int axis,

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/atan.jinja

@@ -4,6 +4,6 @@
 
 {% include "./_meminfo.jinja" %}
 
-#define {{ name|upper }}_OUTPUTS_SIZE {{ out_chan[0] * out_height[0] * out_width[0] }}
+#define {{ name|upper }}_OUTPUTS_SIZE {{out_size[0]}}
 
 #endif /* {{ name|upper }}_LAYER_H */

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/batchnorm2d.jinja

@@ -3,10 +3,10 @@
 #define {{ name|upper }}_LAYER_H
 
 {# For layer configuration -#}
-#define {{ name|upper }}_NB_BATCH {{ input_dims[0] }}
-#define {{ name|upper }}_NB_CHANNELS {{ input_dims[1] }}
-#define {{ name|upper }}_CHANNELS_HEIGHT {{ input_dims[2] }}
-#define {{ name|upper }}_CHANNELS_WIDTH {{ input_dims[3] }}
+#define {{ name|upper }}_NB_BATCH {{ in_dims[0] }}
+#define {{ name|upper }}_NB_CHANNELS {{ in_dims[1] }}
+#define {{ name|upper }}_CHANNELS_HEIGHT {{ in_dims[2] }}
+#define {{ name|upper }}_CHANNELS_WIDTH {{ in_dims[3] }}
 
 #define {{ name|upper }}_EPSILON {{ epsilon }}
 

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/div.jinja

+{#- For name header -#}
+#ifndef {{ name|upper }}_LAYER_H
+#define {{ name|upper }}_LAYER_H
+
+/* ElemWise - sub layer */
+{% include "./_def_io.jinja" %}
+{% include "./_meminfo.jinja" %}
+
+{# For layer configuration -#}
+#define {{ name|upper }}_INPUTS_SIZE {{ in_size[0] }}
+#define {{ name|upper }}_OUTPUTS_SIZE {{ out_size[0] }}
+
+
+#define {{in_name[0]|upper}}_NB_DIM {{ in_dims[0]|length}}
+#define {{in_name[1]|upper}}_NB_DIM {{ in_dims[1]|length}}
+#define {{out_name[0]|upper}}_NB_DIM {{ out_dims[0]|length}}
+
+static const int {{ in_name[0]|upper }}_DIMS[] = { {{ in_dims[0]|join(", ") }} };
+static const int {{ in_name[1]|upper }}_DIMS[] = { {{ in_dims[1]|join(", ") }} };
+static const int {{ out_name[0]|upper }}_DIMS[] = { {{ out_dims[0]|join(", ") }} };
+
+#endif /* {{ name|upper }}_LAYER_H */

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/matmul.jinja

@@ -2,12 +2,14 @@
 #ifndef {{ name|upper }}_LAYER_H
 #define {{ name|upper }}_LAYER_H
 
-{# For layer configuration -#}
+{% include "./_def_io.jinja" %}
+{% include "./_meminfo.jinja" %}
 
 
-#define {{ in_name[0]|upper }}_DIMS {{ in_dims[0] }}
-#define {{ in_name[1]|upper }}_DIMS {{ in_dims[1] }}
-#define {{ out_name[0]|upper }}_DIMS {{ out_dims[0] }}
+{# For layer configuration -#}
+static const int {{ in_name[0]|upper }}_DIMS[] = { {{ in_dims[0]|join(", ") }} };
+static const int {{ in_name[1]|upper}}_DIMS[] = { {{ in_dims[0]|join(", ") }} };
+static const int {{ out_name[0]|upper }}_DIMS[] = { {{ out_dims[0]|join(", ") }} };
 
 #define {{name|upper}}_INPUT_A_DIMS_SIZE {{ in_dims[0]|length}}
 #define {{name|upper}}_INPUT_B_DIMS_SIZE {{ in_dims[1]|length}}

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/reshape.jinja

 {#- For name header -#}
 #ifndef {{ name|upper }}_LAYER_H
 #define {{ name|upper }}_LAYER_H
-
+{% include "./_def_io.jinja" %}
+{% include "./_meminfo.jinja" %}
 {# For layer configuration -#}
 #define {{ name|upper }}_INPUTS_SIZE {{ nb_inputs }}
 #define {{ name|upper }}_OUTPUTS_SIZE {{ nb_outputs }}

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/sigmoid.jinja

+{#- For name header -#}
+#ifndef {{ name|upper }}_LAYER_H
+#define {{ name|upper }}_LAYER_H
+
+/* ElemWise - sub layer */
+{% include "./_def_io.jinja" %}
+{% include "./_meminfo.jinja" %}
+
+{# For layer configuration -#}
+#define {{ name|upper }}_INPUTS_SIZE {{ in_size[0] }}
+#define {{ name|upper }}_OUTPUTS_SIZE {{ out_size[0] }}
+
+
+#define {{in_name[0]|upper}}_NB_DIM {{ in_dims[0]|length}}
+#define {{out_name[0]|upper}}_NB_DIM {{ out_dims[0]|length}}
+
+static const int {{ in_name[0]|upper }}_DIMS[] = { {{ in_dims[0]|join(", ") }} };
+static const int {{ out_name[0]|upper }}_DIMS[] = { {{ out_dims[0]|join(", ") }} };
+
+#endif /* {{ name|upper }}_LAYER_H */

aidge_export_arm_cortexm/_Aidge_Arm/templates/configuration/softmax.jinja

 
 {% include "./_def_io.jinja" %}
-
+{% include "./_meminfo.jinja" %}
 {#- For name header -#}
 #ifndef {{ name|upper }}_LAYER_H
 #define {{ name|upper }}_LAYER_H

aidge_export_arm_cortexm/_Aidge_Arm/templates/forward_call/batchnorm2d.jinja

-aidge_batchnorm2d_chw_{{dataformat}} ({{input_name}}, {{output_name}}, {{running_mean_name}}, {{running_var_name}}, {{weight_name}}, {{bias_name}}, {{ name|upper }}_EPSILON, {{ name|upper }}_NB_CHANNELS, {{ name|upper }}_CHANNELS_WIDTH, {{ name|upper }}_CHANNELS_HEIGHT);
+aidge_batchnorm2d_chw_{{dataformat|default("float32") }} 
+({{in_name}},
+ {{out_name}}, 
+ {{running_mean_name}},
+ {{running_var_name}},
+ {{weight_name}},
+ {{bias_name}}, 
+ {{ name|upper }}_EPSILON, 
+ {{ name|upper }}_NB_CHANNELS, 
+ {{ name|upper }}_CHANNELS_WIDTH, 
+ {{ name|upper }}_CHANNELS_HEIGHT);

aidge_export_arm_cortexm/_Aidge_Arm/templates/forward_call/div.jinja

+{% filter indent(width=4, first=False) %}
+{% include "./_mem_offset.jinja" %}
+aidge_div <{{name|upper}}_OUTPUTS_SIZE>
+            ({{in_name[0]}},
+            {{in_name[1]}},
+            {{out_name[0]}});
+{% endfilter %}

aidge_export_arm_cortexm/_Aidge_Arm/templates/forward_call/matmul.jinja

-aidge_matmul_chw_{{dataformat}} ({{in_name[0]}}, {{in_name[1]}}, {{out_name[0]}}, {{in_name[0]}}_DIMS, {{in_name[1]}}_DIMS, {{out_name[0]}}_DIMS, {{name|upper}}_INPUT_A_DIMS_SIZE, {{name|upper}}_INPUT_B_DIMS_SIZE, {{name|upper}}_OUTPUT_DIMS_SIZE);
+{% filter indent(width=4, first=False) %}
+{% include "./_mem_offset.jinja" %}
+aidge_matmul_chw_{{dataformat| default("float32") }} 
+({{in_name[0]}}, 
+{{in_name[1]}}, 
+{{out_name[0]}}, 
+{{in_name[0]|upper}}_DIMS, 
+{{in_name[1]|upper}}_DIMS, 
+{{out_name[0]|upper}}_DIMS, 
+{{name|upper}}_INPUT_A_DIMS_SIZE, 
+{{name|upper}}_INPUT_B_DIMS_SIZE, 
+{{name|upper}}_OUTPUT_DIMS_SIZE);
+{% endfilter %}
\ No newline at end of file

aidge_export_arm_cortexm/_Aidge_Arm/templates/forward_call/reshape.jinja

-aidge_reshape_chw_{{dataformat}}({{input_name}}, {{output_name}}, {{name|upper}}_OUTPUTS_SIZE);
+{% filter indent(width=4, first=False) %}
\ No newline at end of file
+{% include "./_mem_offset.jinja" %}
+aidge_reshape<{{name|upper}}_OUTPUTS_SIZE>
+({{input_name}}, {{output_name}}, {{name|upper}}_OUTPUTS_SIZE);
+{% endfilter %}
\ No newline at end of file

aidge_export_arm_cortexm/_Aidge_Arm/templates/forward_call/sigmoid.jinja

+{% filter indent(width=4, first=False) %}
+{% include "./_mem_offset.jinja" %}
+aidge_sigmoid<{{name|upper}}_OUTPUTS_SIZE>
+            ({{in_name[0]}},
+            {{out_name[0]}});
+{% endfilter %}

aidge_export_arm_cortexm/_Aidge_Arm/templates/forward_call/sub.jinja

 {% filter indent(width=4, first=False) %}
 {% include "./_mem_offset.jinja" %}
-aidge_sub_float32({{in_name[0]}}, {{in_name[1]}}, {{out_name[0]}}, {{name|upper}}_OUTPUTS_SIZE);
+aidge_sub<{{name|upper}}_OUTPUTS_SIZE>
+            ({{in_name[0]}},
+            {{in_name[1]}},
+            {{out_name[0]}});
 {% endfilter %}