diff --git a/gitgud/commit-msg b/gitgud/commit-msg
index 31f3dcf..1371f79 100755
--- a/gitgud/commit-msg
+++ b/gitgud/commit-msg
@@ -19,7 +19,6 @@ def postMessage(message):
 
 
 def continue_commit_prompt():
-	print "Your commit message is bad. Do you want to continue your commit? [Y/n]"
 	valid = {"": True, "yes": True, "y": True, "no": False, "n": False}
 	sys.stdin = open("/dev/tty", "r")
 	choice = raw_input().lower()
@@ -36,12 +35,22 @@ def main():
 		txt_file.close()
 
 	# check thie return score
-	if postMessage(commit_message):
+	if postMessage(commit_message > 0.8):
 		# commit if score is good
-		print "Your commit message is good."
+		print "Your commit message is quite good."
 		sys.exit(0)
+	else if postMessage(commit_message >= 0.5):
+		# commit if score is good
+		print "Your commit message could be improved. Do you want to continue your commit? [Y/n]"
+		if continue_commit_prompt():
+			sys.exit(0)
+		else:
+			# abort if user says no
+			print("Commit abort.")
+			sys.exit(1)
 	else:
 		# otherwise, ask user if they want to continue the commit
+		print "Your commit message is bad. Do you want to continue your commit? [Y/n]"
 		if continue_commit_prompt():
 			sys.exit(0)
 		else:
diff --git a/server/model/git_good_model.py b/server/model/git_good_model.py
index 6d91ff4..3a44e12 100644
--- a/server/model/git_good_model.py
+++ b/server/model/git_good_model.py
@@ -2,10 +2,10 @@
 import numpy as np
 import http.client, urllib.request, urllib.parse, urllib.error, base64
 from scipy.sparse import hstack
-from sklearn.naive_bayes import MultinomialNB
 from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer
 from sklearn.model_selection import train_test_split
 from sklearn.metrics import precision_recall_fscore_support as score
+from sklearn.neural_network import MLPClassifier
 import random
 import nltk
 from nltk import word_tokenize, pos_tag
@@ -18,7 +18,7 @@
 class GitGudModel():
 
     def __init__(self, dataframe=None):
-        self.model = MultinomialNB()
+        self.model = MLPClassifier(100, logistic)
         self.dataframe = dataframe
         pass
 
@@ -35,7 +35,8 @@ def parse_and_split_data(self, df, *, star_thresh=200, fork_thresh=50, watchers_
 
         y = df["label"]
 
-        X_train, X_test, y_train, y_test = train_test_split(mcl_transformed, y, test_size=0.1, random_state=random.randint(1, 100))
+        X_train, X_test, y_train, y_test = train_test_split(mcl_transformed, y,
+                                                            test_size=0.1, random_state=random.randint(1, 100))
 
         return (X_train, X_test, y_train, y_test)
 
@@ -220,7 +221,7 @@ def commit_body_has_present_tense(self, msg, weight):
         return self.expon((weight * present / total), 5)
 
 
-    def label_data(self, df, good_thresh=.5):
+    def label_data(self, df):
         label = []
 
         negfunctions = [(self.commit_body_has_swear, 1)]
@@ -239,7 +240,7 @@ def label_data(self, df, good_thresh=.5):
                 score += funct(row['msg'], weight)
             for funct, weight in negfunctions:
                 score *= funct(row['msg'], weight)
-            label.append(GOOD if score > good_thresh else BAD)
+            label.append(score)
 
         df['label'] = pd.Series(label)
         return df