update

main.lua

@@ -90,12 +90,14 @@ function run(tr, n_epoches, dset_train, dset_dev, dset_test)
     header('Evaluating on test set')
     printf('-- using model with dev score = %.4f\n', best_score)
     local test_preds = best_trainer:eval(dset_test)
+    local flag = false
     if tr.task == 'SICK' then
         local pearson_score = stats.pearson(test_preds, dset_test.labels)
         local spearman_score = stats.spearmanr(test_preds, dset_test.labels)
         local mse_score = stats.mse(test_preds, dset_test.labels)
         printf('-- Test pearson = %.4f, spearmanr = %.4f, mse = %.4f \n',
             pearson_score, spearman_score, mse_score)
+        if pearson_score > 0.87 then flag = true end
     elseif tr.task == 'MSRP' then
         local accuracy = stats.accuracy(test_preds, dset_test.labels)
         local f1 = stats.f1(test_preds, dset_test.labels)
@@ -116,9 +118,11 @@ function run(tr, n_epoches, dset_train, dset_dev, dset_test)
         local accuracy = stats.accuracy(test_preds, dset_test.labels)
         printf('-- Test accuracy = %.4f \n', accuracy)
     end
+    if flag then
     print('save parameters')
     local path = 'data/params/params-' .. tr.task .. '-' .. tr.structure .. '.t7'
     best_trainer:save(path)
+    end
 end
 
 function test(ts, dset_test)

models/AttTreeGRU.lua

@@ -21,8 +21,6 @@ function AttTreeGRU:__init(config)
     self.output_module = self:new_output_module()
     self.output_modules = {}
 
-    -- task
-    self.task = config.task or true
 end
 
 function AttTreeGRU:new_composer()
@@ -34,9 +32,7 @@ function AttTreeGRU:new_composer()
         nn.Linear(self.mem_dim, self.mem_dim, false)(prev_res)
     })
     local temp = nn.Linear(self.mem_dim, 1)(M)
-    local attention_weights = (task == true)
-            and nn.Transpose({1,2})(nn.SoftMax()(nn.Transpose({1,2})(temp)))
-            or nn.Transpose({1,2})(nn.Sigmoid()(nn.Transpose({1,2})(temp)))
+    local attention_weights = nn.Transpose({1,2})(nn.SoftMax()(nn.Transpose({1,2})(temp)))
     local child_h_att = nn.MM(true, false)({ child_h, attention_weights })
     local child_h_sum = nn.Reshape(self.mem_dim)(child_h_att)
 

models/AttTreeLSTM.lua

@@ -23,8 +23,6 @@ function AttTreeLSTM:__init(config)
     self.output_module = self:new_output_module()
     self.output_modules = {}
 
-    -- task
-    self.task = config.task or true
 end
 
 function AttTreeLSTM:new_composer()
@@ -37,9 +35,7 @@ function AttTreeLSTM:new_composer()
         nn.Linear(self.mem_dim, self.mem_dim, false)(atte_h)
     })
     local temp = nn.Linear(self.mem_dim, 1)(M)
-    local attention_weights = (task == true)
-            and nn.Transpose({1,2})(nn.SoftMax()(nn.Transpose({1,2})(temp)))
-            or nn.Transpose({1,2})(nn.Sigmoid()(nn.Transpose({1,2})(temp)))
+    local attention_weights = nn.Transpose({1,2})(nn.SoftMax()(nn.Transpose({1,2})(temp)))
     local child_h_att = nn.MM(true, false)({ child_h, attention_weights })
     local child_h_sum = nn.Reshape(self.mem_dim)(child_h_att)