# HG changeset patch
# User SylvainPL <sylvain.pannetier.lebeuf@umontreal.ca>
# Date 1269456302 14400
# Node ID 698313f8f6e6a42b263de9213c324996e8b22b69
# Parent  a8b92a4a708d595d36ad9c6a9e15276662d74a53
rajout de methode reliant toutes les couches cachees a la logistic et changeant seulement les parametres de la logistic durant finetune

diff -r a8b92a4a708d -r 698313f8f6e6 deep/stacked_dae/v_sylvain/nist_sda.py
--- a/deep/stacked_dae/v_sylvain/nist_sda.py	Wed Mar 24 14:44:41 2010 -0400
+++ b/deep/stacked_dae/v_sylvain/nist_sda.py	Wed Mar 24 14:45:02 2010 -0400
@@ -68,7 +68,21 @@
                                     max_minibatches=rtt)
 
     parameters=[]
-    optimizer.pretrain(datasets.nist_P07())
+    #Number of files of P07 used for pretraining
+    nb_file=0
+    if state['pretrain_choice'] == 0:
+        print('\n\tpretraining with NIST\n')
+        optimizer.pretrain(datasets.nist_all()) 
+    elif state['pretrain_choice'] == 1:
+        #To know how many file will be used during pretraining
+        nb_file = state['pretraining_epochs_per_layer'] 
+        state['pretraining_epochs_per_layer'] = 1 #Only 1 time over the dataset
+        if nb_file >=100:
+            sys.exit("The code does not support this much pretraining epoch (99 max with P07).\n"+
+            "You have to correct the code (and be patient, P07 is huge !!)\n"+
+             "or reduce the number of pretraining epoch to run the code (better idea).\n")
+        print('\n\tpretraining with P07')
+        optimizer.pretrain(datasets.nist_P07(min_file=0,max_file=nb_file)) 
     channel.save()
     
     #Set some of the parameters used for the finetuning
@@ -89,34 +103,47 @@
     
     #Decide how the finetune is done
     
-    if finetune_choice==0:
-        print('\n\n\tfinetune avec nist\n\n')
-        optimizer.reload_parameters()
-        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1)
-    if finetune_choice==1:
-        print('\n\n\tfinetune avec P07\n\n')
-        optimizer.reload_parameters()
-        optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0)
-    if finetune_choice==2:
-        print('\n\n\tfinetune avec nist suivi de P07\n\n')
-        optimizer.reload_parameters()
-        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1)
-        optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0)
-
+    if finetune_choice == 0:
+        print('\n\n\tfinetune with NIST\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(min_file=nb_file),max_finetune_epoch_NIST,ind_test=1)
+        channel.save()
+    if finetune_choice == 1:
+        print('\n\n\tfinetune with P07\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_P07(min_file=nb_file),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0)
+        channel.save()
+    if finetune_choice == 2:
+        print('\n\n\tfinetune with NIST followed by P07\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(min_file=nb_file),max_finetune_epoch_NIST,ind_test=21)
+        optimizer.finetune(datasets.nist_P07(min_file=nb_file),datasets.nist_all(),max_finetune_epoch_P07,ind_test=20)
+        channel.save()
+    if finetune_choice == 3:
+        print('\n\n\tfinetune with NIST only on the logistic regression on top (but validation on P07).\n\
+        All hidden units output are input of the logistic regression\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(min_file=nb_file),max_finetune_epoch_NIST,ind_test=1,special=1)
+        
+        
     if finetune_choice==-1:
-        print('\nSerie de 3 essais de fine-tuning')
-        print('\n\n\tfinetune avec nist\n\n')
-        optimizer.reload_parameters()
-        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1)
+        print('\nSERIE OF 3 DIFFERENT FINETUNINGS')
+        print('\n\n\tfinetune with NIST\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(min_file=nb_file),max_finetune_epoch_NIST,ind_test=1)
+        channel.save()
+        print('\n\n\tfinetune with P07\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_P07(min_file=nb_file),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0)
         channel.save()
-        print('\n\n\tfinetune avec P07\n\n')
-        optimizer.reload_parameters()
-        optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0)
+        print('\n\n\tfinetune with NIST (done earlier) followed by P07 (written here)\n\n')
+        optimizer.reload_parameters('params_finetune_NIST.txt')
+        optimizer.finetune(datasets.nist_P07(min_file=nb_file),datasets.nist_all(),max_finetune_epoch_P07,ind_test=20)
         channel.save()
-        print('\n\n\tfinetune avec nist suivi de P07\n\n')
-        optimizer.reload_parameters()
-        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(),max_finetune_epoch_NIST,ind_test=1)
-        optimizer.finetune(datasets.nist_P07(),datasets.nist_all(),max_finetune_epoch_P07,ind_test=0)
+        print('\n\n\tfinetune with NIST only on the logistic regression on top.\n\
+        All hidden units output are input of the logistic regression\n\n')
+        optimizer.reload_parameters('params_pretrain.txt')
+        optimizer.finetune(datasets.nist_all(),datasets.nist_P07(min_file=nb_file),max_finetune_epoch_NIST,ind_test=1,special=1)
         channel.save()
     
     channel.save()