more debug

kermitt2 · Jan 28, 2024 · bf19c65 · bf19c65
1 parent 7061ff9
commit bf19c65
Showing 1 changed file with 6 additions and 6 deletions.
diff --git a/delft/applications/licenseClassifier.py b/delft/applications/licenseClassifier.py
@@ -141,7 +141,7 @@ def train_binary(embeddings_name, fold_count, architecture="gru", transformer=No
     print('loading multiclass copyright/license dataset...')
     xtr, y_copyrights = _read_data('data/textClassification/licenses/copyrights-licenses-data-validated.csv', data_type="copyrights")
 
-    report_training_contexts(y_copyrights)
+    report_training_copyrights(y_copyrights)
 
     for class_rank in range(len(list_classes_copyright)):
         model_name = 'copyright_' + list_classes_copyright[class_rank] + '_'+architecture
@@ -161,9 +161,9 @@ def train_binary(embeddings_name, fold_count, architecture="gru", transformer=No
             class_weights=class_weights, transformer_name=transformer)
 
         if fold_count == 1:
-            model.train(x_train, y_train_class_rank)
+            model.train(xtr, y_train_class_rank)
         else:
-            model.train_nfold(x_train, y_train_class_rank)
+            model.train_nfold(xtr, y_train_class_rank)
         # saving the model
         model.save()
 
@@ -186,9 +186,9 @@ def train_binary(embeddings_name, fold_count, architecture="gru", transformer=No
             class_weights=class_weights, transformer_name=transformer)
 
         if fold_count == 1:
-            model.train(x_train, y_train_class_rank)
+            model.train(xtr, y_train_class_rank)
         else:
-            model.train_nfold(x_train, y_train_class_rank)
+            model.train_nfold(xtr, y_train_class_rank)
         # saving the model
         model.save()
 
@@ -197,7 +197,7 @@ def train_and_eval_binary(embeddings_name, fold_count, architecture="gru", trans
     print('loading multiclass copyright/license dataset...')
     xtr, y_copyrights = _read_data('data/textClassification/licenses/copyrights-licenses-data-validated.csv', data_type="copyrights")
 
-    report_training_contexts(y_copyrights)
+    report_training_copyrights(y_copyrights)
 
     # segment train and eval sets
     x_train, y_train, x_test, y_test = split_data_and_labels(xtr, y_copyrights, 0.9)