Re-factoring

.vscode/launch.json

@@ -1,15 +1,6 @@
 {
-    // Usare IntelliSense per informazioni sui possibili attributi.
-    // Al passaggio del mouse vengono visualizzate le descrizioni degli attributi esistenti.
-    // Per altre informazioni, visitare: https://go.microsoft.com/fwlink/?linkid=830387
-    "version": "0.2.0",
     "configurations": [
-        {
-            "name": "Python: File corrente",
-            "type": "python",
-            "request": "launch",
-            "program": "/home/christian/Documenti/GitHub/Image-Captioning/v1/NeuralNet.py",
-            "console": "integratedTerminal"
-        }
+    {"name":"Main","type":"python","request":"launch","program":"${workspaceFolder}/main.py","console":"integratedTerminal", 
+        "args": ["RNetvI", "train", "1024", "1024", "--attention", "True", "--attention_dim", "1024", "--dataset_folder", "./dataset/flickr30k_images", "--device", "cuda:0", "--splits", "1", "1", "1", "--epochs", "2"]}
     ]
 }

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2022 christianSistemiPos
+Copyright (c) 2022 christiandimaio
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

NeuralModels/Attention/IAttention.py

@@ -0,0 +1,37 @@
+import torch
+import torch.nn as nn
+
+class IAttention(nn.Module):
+    """
+        Class interface for Attention unit 
+        Args are intended as suggested.
+    """
+    def __init__(self, *args):
+        """Constructor for an Attention model 
+
+        Args:
+            encoder_dim (int): 
+                The number of features extracted from the image.
+            hidden_dim (int): 
+                The capacity of the LSTM.
+            attention_dim (int): 
+                The capacity of the Attention Model.
+        """
+        super(IAttention, self).__init__()
+
+    def forward(self, *args):
+        """Compute z_t given images and hidden state at t-1 for all the element in the batch.
+
+        Args:
+            images (torch.Tensor): `(batch_dim, image_portions, encoder_dim)`
+                The tensor of the images in the batch. 
+            lstm_hidden_states (torch.Tensor): `(batch_dim, hidden_dim)`
+                The hidden states at t-1 of the elements in the batch. 
+
+        Returns:
+            (Tuple[torch.Tensor,torch.Tensor]): `[(batch_dim, encoder_dim), (batch_dim, image_portions)]`
+                Z_t and the alphas evaluated for each portion of the image, for each image in the batch.
+        """
+        pass
+
+