introduce small-footprint backward computation (by using "-smallfootprint")

source/sample/transformer/T2TTrainer.cpp

@@ -116,6 +116,7 @@ void T2TTrainer::Init(int argc, char ** argv)
     LoadParamBool(argc, argv, "doubledend", &isDoubledEnd, false);
     LoadParamBool(argc, argv, "smallbatch", &isSmallBatch, true);
     LoadParamBool(argc, argv, "bigbatch", &isBigBatch, false);
+    LoadParamBool(argc, argv, "smallfootprint", &isSmallFootprint, false);
 
     buf  = new int[bufSize];
     buf2 = new int[bufSize];
@@ -163,6 +164,9 @@ void T2TTrainer::Train(const char * fn, const char * validFN, const char * model
     int devID = model->devID;
     XMem * mem = model->mem;
     XNet net;
+
+    if(isSmallFootprint)
+        net.SetGradEfficientFlag();
     
     PrepareModel(model);
 
@@ -213,8 +217,6 @@ void T2TTrainer::Train(const char * fn, const char * validFN, const char * model
                 ShowNTErrors("Illegal model type!");
             }
 
-            net.ShowNetwork(stderr, &output);
-
             /* back-propagation for obtaining gradients */
             if (labelSmoothingP > 0)
                 LabelSmooth(&gold, &goldSmoothed, labelSmoothingP);

source/sample/transformer/T2TTrainer.h

@@ -142,6 +142,9 @@ public:
     /* counterpart of "isSmallBatch" */
     bool isBigBatch;
 
+    /* indicates whether we use small memory footprint for backward process */
+    bool isSmallFootprint;
+
 public:
     /* constructor */
     T2TTrainer();