Updated

app.py

@@ -1,7 +1,7 @@
 import gradio as gr
 import numpy as np
 from datetime import date
-from predictors import predict_bc_scores, predict_mc_scores, predict_1on1_scores
+from predictors import predict_bc_scores, predict_mc_scores
 from analysis import depth_analysis
 from predictors import predict_quillbot
 from plagiarism import plagiarism_check, build_date
@@ -27,7 +27,7 @@ def ai_generated_test(option, input, models):
     if option == "Human vs AI":
         return predict_bc_scores(input), None
     elif option == "Human vs AI Source Models":
-        return predict_bc_scores(input), predict_1on1_scores(input, models)
+        return predict_bc_scores(input), predict_mc_scores(input, models)
     return None, None
 
 
@@ -59,7 +59,7 @@ def main(
     )
     depth_analysis_plot = depth_analysis(input)
     bc_score = predict_bc_scores(input)
-    mc_score = predict_1on1_scores(input, models)
+    mc_score = predict_mc_scores(input, models)
     quilscore = predict_quillbot(input)
 
     return (
@@ -73,7 +73,7 @@ def main(
 
 # START OF GRADIO
 
-title = "Copyright Checker"
+title = "AI Detection and Source Analysis"
 months = {
     "January": "01",
     "February": "02",

predictors.py

@@ -227,43 +227,6 @@ def predict_mc(model, tokenizer, text):
         output_norm = softmax(output.logits.detach().cpu().numpy(), 1)[0]
         return output_norm
 
-
-def predict_mc_scores(input):
-    bc_scores = []
-    mc_scores = []
-
-    samples_len_bc = len(split_text_allow_complete_sentences_nltk(input, type_det="bc"))
-    segments_bc = split_text_allow_complete_sentences_nltk(input, type_det="bc")
-    for i in range(samples_len_bc):
-        cleaned_text_bc = remove_special_characters(segments_bc[i])
-        bc_score = predict_bc(text_bc_model, text_bc_tokenizer, cleaned_text_bc)
-        bc_scores.append(bc_score)
-    bc_scores_array = np.array(bc_scores)
-    average_bc_scores = np.mean(bc_scores_array, axis=0)
-    bc_score_list = average_bc_scores.tolist()
-    bc_score = {"AI": bc_score_list[1], "HUMAN": bc_score_list[0]}
-    segments_mc = split_text_allow_complete_sentences_nltk(input, type_det="mc")
-    samples_len_mc = len(split_text_allow_complete_sentences_nltk(input, type_det="mc"))
-    for i in range(samples_len_mc):
-        cleaned_text_mc = remove_special_characters(segments_mc[i])
-        mc_score = predict_mc(text_mc_model, text_mc_tokenizer, cleaned_text_mc)
-        mc_scores.append(mc_score)
-    mc_scores_array = np.array(mc_scores)
-    average_mc_scores = np.mean(mc_scores_array, axis=0)
-    mc_score_list = average_mc_scores.tolist()
-    mc_score = {}
-    for score, label in zip(mc_score_list, mc_label_map):
-        mc_score[label.upper()] = score
-
-    sum_prob = 1 - bc_score["HUMAN"]
-    for key, value in mc_score.items():
-        mc_score[key] = value * sum_prob
-    if sum_prob < 0.01:
-        mc_score = {}
-
-    return mc_score
-
-
 def predict_bc_scores(input):
     bc_scores = []
     samples_len_bc = len(split_text_allow_complete_sentences_nltk(input, type_det="bc"))
@@ -313,7 +276,7 @@ def predict_1on1_single(input, model):
     return predictions
 
 
-def predict_1on1_scores(input, models):
+def predict_mc_scores(input, models):
 
     if len(models) == 0:
         return {}