add train-only binary classifiers for software contexts

delft/applications/softwareContextClassifier.py

@@ -40,7 +40,7 @@ def configure(architecture):
  if architecture == "bert":
  batch_size = 32
  early_stop = False
- max_epoch = 5
+ max_epoch = 6
  maxlen = 100
 
  return batch_size, maxlen, patience, early_stop, max_epoch
@@ -99,6 +99,35 @@ def train_and_eval(embeddings_name, fold_count, architecture="gru", transformer=
  model.save()
 
 
+def train_binary(embeddings_name, fold_count, architecture="gru", transformer=None):
+ print('loading multiclass software context dataset...')
+ x_train, y_train = load_software_context_corpus_json("data/textClassification/software/software-contexts.json.gz")
+
+ report_training_contexts(y_train)
+
+ for class_rank in range(len(list_classes)):
+ model_name = 'software_context_' + list_classes[class_rank] + '_'+architecture
+ class_weights = None
+
+ batch_size, maxlen, patience, early_stop, max_epoch = configure(architecture)
+
+ y_train_class_rank = [ [1, 0] if y[class_rank] == 1.0 else [0, 1] for y in y_train ]
+ y_train_class_rank = np.array(y_train_class_rank)
+
+ list_classes_rank = [list_classes[class_rank], "not_"+list_classes[class_rank]]
+
+ model = Classifier(model_name, architecture=architecture, list_classes=list_classes_rank, max_epoch=max_epoch, fold_number=fold_count, patience=patience,
+ use_roc_auc=True, embeddings_name=embeddings_name, batch_size=batch_size, maxlen=maxlen, early_stop=early_stop,
+ class_weights=class_weights, transformer_name=transformer)
+
+ if fold_count == 1:
+ model.train(x_train, y_train_class_rank)
+ else:
+ model.train_nfold(x_train, y_train_class_rank)
+ # saving the model
+ model.save()
+
+
 def train_and_eval_binary(embeddings_name, fold_count, architecture="gru", transformer=None): 
  print('loading multiclass software context dataset...')
  xtr, y = load_software_context_corpus_json("data/textClassification/software/software-contexts.json.gz")
@@ -203,8 +232,8 @@ def report_training_contexts(y):
 
  args = parser.parse_args()
 
- if args.action not in ('train', 'train_eval', 'classify'):
- print('action not specified, must be one of [train,train_eval,classify]')
+ if args.action not in ('train', 'train_eval', 'classify', 'train_binary', 'train_eval_binary'):
+ print('action not specified, must be one of [train,train_binary,train_eval,train_eval_binary,classify]')
 
  embeddings_name = args.embedding
  transformer = args.transformer
@@ -223,6 +252,12 @@ def report_training_contexts(y):
 
  train(embeddings_name, args.fold_count, architecture=architecture, transformer=transformer)
 
+ if args.action == 'train_binary':
+ if args.fold_count < 1:
+ raise ValueError("fold-count should be equal or more than 1")
+
+ train_binary(embeddings_name, args.fold_count, architecture=architecture, transformer=transformer)
+
  if args.action == 'train_eval':
  if args.fold_count < 1:
  raise ValueError("fold-count should be equal or more than 1")

setup.py

@@ -27,7 +27,8 @@
  'truecase',
  'requests',
  'pandas==1.3.5',
- 'pytest'
+ 'pytest',
+ 'tensorflow-addons==0.15.0' 
  ],
  classifiers=[
  "Programming Language :: Python :: 3.7",