增加多篇幅检测

flask_check_bert_test.py

@@ -116,8 +116,7 @@ def rouge_pre_m(text, df_train_nuoche):
 
     re1 = [(i[0], i[1]) for i in sorted(list(enumerate(index_rouge_list)), key=lambda x: x[1], reverse=True)]
 
-    return_list.append(re1[0][1])
-    return_list.append(re1[0][0])
+    return_list.extend(re1)
 
     return return_list
 
@@ -178,11 +177,13 @@ def accurate_check_rouge(
 
     # rouge算法
     for i in range(len(data_zong)):
-        if data_zong[i][0] > 0.47:
-            bool_check_sentense.append([i,data_zong[i][1]])
+        for j in range(len(data_zong[i])):
+            if data_zong[i][j][1] > 0.47:
+                bool_check_sentense.append([i,data_zong[i][j][0]])
     biao_red = biaohong(bool_check_sentense, data_zong, recall_data_list) # [[[0, 1, 2], [479, 480, 481]], [[3, 4, 5], [481, 482, 483]], [[6, 7, 8], [484, 485, 486]]]
 
     print("bert精确查重时间", t1-t0)
+    print(biao_red)
 
 
     sentence_0_list = []
@@ -625,7 +626,7 @@ def ulit_recall_paper(recall_data_list_dict):
 
 
     data = []
-    for i in list(recall_data_list_dict.items())[:5]:
+    for i in list(recall_data_list_dict.items()):
         data_one = processing_one_text(i[0])
         data.extend(data_one)