fix for issue 52: cpu mode epochs

dataset.lua

@@ -38,32 +38,18 @@ function DataSet:__init(loader, options)
 end
 
 function DataSet:load(loader)
-  local filename = "data/vocab.t7"
-
-  if path.exists(filename) then
-    print("Loading vocabulary from " .. filename .. " ...")
-    local data = torch.load(filename)
-    self.word2id = data.word2id
-    self.id2word = data.id2word
-    self.wordsCount = data.wordsCount
-    self.goToken = data.goToken
-    self.eosToken = data.eosToken
-    self.unknownToken = data.unknownToken
-    self.examplesCount = data.examplesCount
-  else
-    print("" .. filename .. " not found")
-    self:visit(loader:load())
-    print("Writing " .. filename .. " ...")
-    torch.save(filename, {
-      word2id = self.word2id,
-      id2word = self.id2word,
-      wordsCount = self.wordsCount,
-      goToken = self.goToken,
-      eosToken = self.eosToken,
-      unknownToken = self.unknownToken,
-      examplesCount = self.examplesCount
-    })
-  end
+  local filenam = 'data/vocab.t7'
+  self:visit(loader:load())
+  print("Writing " .. filename .. " ...")
+  torch.save(filename, {
+    word2id = self.word2id,
+    id2word = self.id2word,
+    wordsCount = self.wordsCount,
+    goToken = self.goToken,
+    eosToken = self.eosToken,
+    unknownToken = self.unknownToken,
+    examplesCount = self.examplesCount
+  })
 end
 
 function DataSet:visit(conversations)

seq2seq.lua

@@ -43,6 +43,17 @@ function Seq2Seq:float()
   end
 end
 
+-- function Seq2Seq:double()
+-- created by zhaopku to fix the problem of CPU mode
+function Seq2Seq:double()
+  self.encoder:double()
+  self.decoder:double()
+
+  if self.criterion then
+    self.criterion:double()
+  end
+end
+
 function Seq2Seq:cl()
   self.encoder:cl()
   self.decoder:cl()

train.lua

@@ -158,7 +158,7 @@ for epoch = 1, options.maxEpoch do
     params, gradParams = nil,nil
     collectgarbage()
     -- Model is saved as CPU
-    model:float()
+    model:double()
     torch.save("data/model.t7", model)
     collectgarbage()
     if options.cuda then