1 files changed, 49 insertions, 53 deletions

diff --git a/gd.cc b/gd.cc
index 2304a561..c245ff04 100644
--- a/gd.cc
+++ b/gd.cc
@@ -21,7 +21,6 @@ embodied in the content of this file are licensed under the BSD
 #include "gd.h"
 #include "cache.h"
 #include "simple_label.h"
-#include "delay_ring.h"
 #include "allreduce.h"
 #include "accumulate.h"
 
@@ -40,43 +39,42 @@ void* gd_thread(void *in)
   
   while ( true )
     {//this is a poor man's select operation.
-      if ((ec = get_delay_example()) != NULL)//nonblocking
+      if ((ec = get_example()) != NULL)//semiblocking operation.
 	{
+	  assert(ec->in_use);
 	  if (ec->pass != current_pass)
 	    {
+	      if(global.span_server != "") {
+		if(global.adaptive)
+		  accumulate_weighted_avg(global.span_server, params->reg);
+		else 
+		  accumulate_avg(global.span_server, params->reg, 0);	      
+	      }
+
 	      if (global.save_per_pass)
 		sync_weights(&reg);
 	      global.eta *= global.eta_decay_rate;
 	      save_predictor(*(params->final_regressor_name), current_pass);
 	      current_pass = ec->pass;
 	    }
-	  if (global.adaptive)
-	    adaptive_inline_train(reg,ec,ec->eta_round);
-	  else
-	    inline_train(reg, ec, ec->eta_round);
-	  finish_example(ec);
-	  if (global.sd->contraction < 1e-10)  // updating weights now to avoid numerical instability
-	    sync_weights(&reg);
-	}
-      else if ((ec = get_example()) != NULL)//semiblocking operation.
-	{
-	  assert(ec->in_use);
-	  if (ec->pass != current_pass && global.span_server != "")
+	  
+	  if (!command_example(ec, params))
 	    {
-	      if(global.span_server != "") {
-		if(global.adaptive)
-		  accumulate_weighted_avg(global.span_server, params->reg);
-		else 
-		  accumulate_avg(global.span_server, params->reg, 0);	      
-	      }
+	      predict(reg,ec,*(params->vars));
+	      if (ec->eta_round != 0.)
+		{
+		  if (global.adaptive)
+		    adaptive_inline_train(reg,ec,ec->eta_round);
+		  else
+		    inline_train(reg, ec, ec->eta_round);
+		  if (global.sd->contraction < 1e-10)  // updating weights now to avoid numerical instability
+		    sync_weights(&reg);
+		  
+		}
 	    }
-
-	  if (command_example(ec, params))
-	    delay_example(ec,0);
-	  else
-	    predict(reg,ec,*(params->vars));
+	  finish_example(ec);
 	}
-      else if (thread_done())
+      else if (parser_done())
 	{
 	  sync_weights(&reg);
 	  if(global.span_server != "") {
@@ -615,31 +613,35 @@ void local_predict(example* ec, gd_vars& vars, regressor& reg)
     t = global.sd->weighted_unlabeled_examples;
   else
     t = ec->example_t;
-
+  
+  ec->eta_round = 0;
   if (ld->label != FLT_MAX)
     {
       ec->loss = reg.loss->getLoss(ec->final_prediction, ld->label) * ld->weight;
       
-      double eta_t;
-      float norm;
-      if (global.adaptive && global.exact_adaptive_norm) {
-	float magx = 0.;
-	norm = compute_xGx(reg, ec, magx);
-	eta_t = global.eta * norm / magx;
-      } else {
-	eta_t = global.eta / pow(t,vars.power_t) * ld->weight;
-	norm = global.nonormalize ? 1. : ec->total_sum_feat_sq;
-      }
-
-      ec->eta_round = reg.loss->getUpdate(ec->final_prediction, ld->label, eta_t, norm) / global.sd->contraction;
-
-      if (global.training && global.reg_mode && fabs(ec->eta_round) > 1e-8) {
-	double dev1 = reg.loss->first_derivative(ec->final_prediction, ld->label);
-	double eta_bar = (fabs(dev1) > 1e-8) ? (-ec->eta_round / dev1) : 0.0;
-	if (fabs(dev1) > 1e-8)
-	  global.sd->contraction /= (1. + global.l2_lambda * eta_bar * norm);
-	global.sd->gravity += eta_bar * sqrt(norm) * global.l1_lambda;
-      }
+      if (global.training)
+	{
+	  double eta_t;
+	  float norm;
+	  if (global.adaptive && global.exact_adaptive_norm) {
+	    float magx = 0.;
+	    norm = compute_xGx(reg, ec, magx);
+	    eta_t = global.eta * norm / magx;
+	  } else {
+	    eta_t = global.eta / pow(t,vars.power_t) * ld->weight;
+	    norm = global.nonormalize ? 1. : ec->total_sum_feat_sq;
+	  }
+	  
+	  ec->eta_round = reg.loss->getUpdate(ec->final_prediction, ld->label, eta_t, norm) / global.sd->contraction;
+	  
+	  if (global.reg_mode && fabs(ec->eta_round) > 1e-8) {
+	    double dev1 = reg.loss->first_derivative(ec->final_prediction, ld->label);
+	    double eta_bar = (fabs(dev1) > 1e-8) ? (-ec->eta_round / dev1) : 0.0;
+	    if (fabs(dev1) > 1e-8)
+	      global.sd->contraction /= (1. + global.l2_lambda * eta_bar * norm);
+	    global.sd->gravity += eta_bar * sqrt(norm) * global.l1_lambda;
+	  }
+	}
     }
   else if(global.active)
     ec->revert_weight = reg.loss->getRevertingWeight(ec->final_prediction, global.eta/pow(t,vars.power_t));
@@ -676,12 +678,6 @@ void predict(regressor& r, example* ex, gd_vars& vars)
 
   local_predict(ex, vars,r);
   ex->done = true;
-  
-  if (global.training && ((label_data*)(ex->ld))->label != FLT_MAX)
-    delay_example(ex,1);
-  else
-    delay_example(ex,0);
-
 }
 
 // trains regressor r on one example ex.