feat: Add table parsing pipeline separately

python/table_parse/computation_graph_table_parse.py

@@ -0,0 +1,56 @@
+import torch
+from torch.autograd import Variable
+
+from network.dense import Dense
+
+
+class TableParse(torch.nn.Module):
+    def __init__(self, num_embeddings, num_positional, num_convolutional):
+        super(TableParse, self).__init__()
+
+        self.moduleA_embeddings = Dense(num_embeddings, config = [300, 'R', 300, 'R', 100])
+        self.moduleA_positional = Dense(num_positional, config = [20, 'S', 20, 'R', 100])
+        self.moduleA_convolutional = Dense(num_convolutional, config = [100, 'R', 100, 'R', 100])
+
+        self.moduleA_project_down = Dense(300, config=[100, 'R'])
+
+        self.moduleB = torch.nn.GRUCell(500, 100).cuda()
+        self.moduleBO_1 = Dense(100, config = [100, 'R', 100, 'R', 100, 'R'])
+        self.moduleBO_2 = Dense(100, config = [100, 'R', 100, 'R', 100, 'R'])
+        self.moduleC = Dense(100, config = [100,'R',100])
+        self.moduleD = Dense(100, config = [100, 'R', 100, 'R'])
+
+        self.iterations = 1
+
+    def set_iterations(self, iterations):
+        self.iterations = iterations
+
+    def concat(self, x, indices, indices_not_found, num_words):
+        y = Variable(torch.zeros(num_words, 100 * 5)).cuda()
+        y[:, 000:100] = x[indices[:, 0]]
+        y[:, 100:200] = x[indices[:, 1]]
+        y[:, 200:300] = x[indices[:, 2]]
+        y[:, 300:400] = x[indices[:, 3]]
+        y[:, 400:500] = x[indices[:, 4]]
+        y[indices_not_found] = 0
+
+        return y
+
+    def forward(self, indices, indices_not_found, word_embeddings, positional_features, convolutional_features, num_words):
+        uu_embeddings = self.moduleA_embeddings.forward(word_embeddings)
+        uu_positional = self.moduleA_positional.forward(positional_features)
+        uu_convolutional = self.moduleA_convolutional.forward(convolutional_features)
+
+        uu_combined = torch.cat((uu_embeddings, uu_positional, uu_convolutional), dim=1)
+        uu = self.moduleA_project_down.forward(uu_combined)
+
+        hh = Variable(torch.zeros(num_words,100)).cuda()
+
+        for i in range(self.iterations):
+            ww = self.concat(uu, indices, indices_not_found, num_words)
+            bb = self.moduleB.forward(ww, hh)
+            oo, hh = self.moduleBO_1.forward(bb), self.moduleBO_2.forward(bb)
+            ll = self.moduleC.forward(oo)
+            uu = self.moduleD.forward(hh)
+
+        return ll