update status

.idea/misc.xml

@@ -1,4 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <project version="4">
   <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.8" project-jdk-type="Python SDK" />
+  <component name="PyCharmProfessionalAdvertiser">
+    <option name="shown" value="true" />
+  </component>
 </project>

.idea/modules.xml

@@ -2,7 +2,7 @@
 <project version="4">
   <component name="ProjectModuleManager">
     <modules>
-      <module fileurl="file://$PROJECT_DIR$/.idea/exam-question-classification.iml" filepath="$PROJECT_DIR$/.idea/exam-question-classification.iml" />
+      <module fileurl="file://$PROJECT_DIR$/.idea/mt-clerk-classification-algo.iml" filepath="$PROJECT_DIR$/.idea/mt-clerk-classification-algo.iml" />
     </modules>
   </component>
 </project>

README.md

@@ -1,12 +0,0 @@
-# exam-question-classification
-
-This is the experimental code for Test Case Classification via Few-shot Learning
-Test case classification by BILSTM
-## Requirements
-python==3.8
-bert4keras==0.7.7
-tensorflow==2.2.0
-torch==1.11.0
-pyltp==0.4.0
-## Run
-python classify_service/bilstm_attention.py

classify_service/bilstm_attention.py

@@ -31,7 +31,7 @@ epochs = 10
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
-app_names = ["决赛自主可控众测web自主可控运维管理系统", "航天中认自主可控众包测试练习赛", "趣享GIF众包测试201908试题"]
+app_names = ["趣享GIF众包测试201908试题"]
 bug_type = ["不正常退出", "功能不完整", "用户体验", "页面布局缺陷", "性能", "安全"]
 lexicon = {0: [], 1: [], 2: [], 3: [], 4: [], 5: []}
 n = 5  # 选择置信度最高的前n条数据
@@ -297,7 +297,7 @@ def test_origin(model, test_loader, loss_func):
         preds = torch.argmax(preds, dim=1)
         labels = torch.argmax(labels, dim=1)
 
-        recall = sklearn.metrics.recall_score(labels, preds, average="macro")
+        recall = sklearn.metrics.recall_score(labels, preds, average="macro",zero_division=0)
         f1 = sklearn.metrics.f1_score(labels, preds, average="macro", zero_division=0)
         pre = sklearn.metrics.precision_score(labels, preds, average="macro", zero_division=0)
         corrects += torch.sum(preds == labels).item()
@@ -317,6 +317,7 @@ def test_origin(model, test_loader, loss_func):
 
 def train_origin(model, train_loader, test_loader, optimizer, loss_func, epochs):
     best_val_acc = 0.0
+    # best_val_acc = test_origin(model, test_loader, loss_func)[0]
     best_model_params = copy.deepcopy(model.state_dict())
     for epoch in range(epochs):
         model.train()
@@ -325,27 +326,23 @@ def train_origin(model, train_loader, test_loader, optimizer, loss_func, epochs)
         for datas, labels in train_loader:
             datas = datas.to(device)
             labels = labels.to(device)
-
             attention_w, preds = model.forward(datas)  # 使用model预测数据
             loss = loss_func(preds, labels)
-
             optimizer.zero_grad()
             loss.backward()
             optimizer.step()
-
             loss_val += loss.item() * datas.size(0)
-
             # 获取预测的最大概率出现的位置
             preds = torch.argmax(preds, dim=1)
             labels = torch.argmax(labels, dim=1)
             corrects += torch.sum(preds == labels).item()
         train_loss = loss_val / len(train_loader.dataset)
-        train_acc = corrects / len(train_loader.dataset)
-
+        # train_acc = corrects / len(train_loader.dataset)
         if epoch % 2 == 0:
             # print("Train Loss: {}, Train Acc: {}".format(train_loss, train_acc))
             test_acc = test_origin(model, test_loader, loss_func)[0]
             if best_val_acc < test_acc:
+                print("best:",best_val_acc," new_best:",test_acc)
                 best_val_acc = test_acc
                 best_model_params = copy.deepcopy(model.state_dict())
     model.load_state_dict(best_model_params)
@@ -431,10 +428,8 @@ if __name__ == "__main__":
             i = i + 1
             # 从发展集中构建词库
             new_labeled_data = test_with_lexicon(model, develop_loader, develop_features, word2index)
-
             print("重新贴标签的数据是{}".format(new_labeled_data))
             print("现在的词库是{}".format(lexicon))
-
             if len(new_labeled_data) == 0:
                 break
             train_features, develop_features, train_labels, develop_labels = develop_to_train(new_labeled_data,
@@ -442,9 +437,7 @@ if __name__ == "__main__":
                                                                                               develop_features,
                                                                                               train_labels,
                                                                                               develop_labels)
-
             embed = nn.Embedding(vocab_size + 2, embedding_size)  # https://www.jianshu.com/p/63e7acc5e890
-
             train_features_after1 = torch.LongTensor(train_features)
             train_features_after1 = embed(train_features_after1)
             train_features_after2 = Variable(train_features_after1, requires_grad=False)

classify_service/contrast_experiment.py

@@ -1,19 +1,22 @@
 #!/usr/bin/python
-#coding=utf-8
+# coding=utf-8
 
 import os
 import numpy as np
 import logging
 import sklearn
-from sklearn.model_selection import train_test_split     #导入切分训练集、测试集模块
+import torch
+from sklearn.model_selection import train_test_split  # 导入切分训练集、测试集模块
 from sklearn.neighbors import KNeighborsClassifier
 from sklearn import svm
 from sklearn.naive_bayes import GaussianNB
 
-
 fileName = './constract.log'
-handler = [logging.FileHandler(filename=fileName,encoding="utf-8")]
-logging.basicConfig(level = logging.DEBUG, handlers = handler)
+formatter = logging.Formatter('%(asctime)s [%(levelname)s] %(module)s: %(message)s',
+                              datefmt='%m/%d/%Y %H:%M:%S')
+handler = logging.FileHandler(filename=fileName, encoding="utf-8")
+handler.setFormatter(formatter)
+logging.basicConfig(level=logging.DEBUG, handlers=[handler])
 
 parent_path = os.path.dirname(os.path.realpath(__file__))
 grander_path = os.path.dirname(parent_path)
@@ -28,6 +31,7 @@ embedding_size = 64
 batch_size = 16
 random_state = 15
 
+
 def contrast():
     logging.info("正在加载初始数据")
     txts = np.load(word_list_data_path_base + str(dataset_name) + ".npy", allow_pickle=True)
@@ -47,51 +51,91 @@ def contrast():
     for txt in txts:
         text_feature = text_to_feature(txt, word2index, max_len)
         features.append(text_feature)
-    #np.save(, features)
+    # np.save(, features)
     score_knn_lowest = 100
     score_svm_lowest = 100
     score_nb_lowest = 100
+
     score_knn_all = 0
+    recall_knn_all = 0
+    f1_knn_all = 0
+
     score_svm_all = 0
+    recall_svm_all = 0
+    f1_svm_all = 0
+
     score_nb_all = 0
+    recall_nb_all = 0
+    f1_nb_all = 0
     for i in range(random_state):
-        train_data, test_data, train_label, test_label = sklearn.model_selection.train_test_split(features, labels_new, random_state = i, train_size = 0.2,test_size = 0.8)
-        
+        train_data, test_data, train_label, test_label = sklearn.model_selection.train_test_split(features, labels_new,
+                                                                                                  random_state=i,
+                                                                                                  train_size=0.2,
+                                                                                                  test_size=0.8)
+
         logging.info("正在训练k最近邻分类器")
         knn_classifier = KNeighborsClassifier()
         knn_classifier.fit(train_data, train_label)
+        knn_predict = knn_classifier.predict(test_data)
+        recall_knn = sklearn.metrics.recall_score(test_label, knn_predict, average="macro")
+        f1_knn = sklearn.metrics.f1_score(test_label, knn_predict, average="macro")
         score_knn = knn_classifier.score(test_data, test_label)
         if score_knn < score_knn_lowest:
             score_knn_lowest = score_knn
         score_knn_all = score_knn_all + score_knn
+        recall_knn_all += recall_knn
+        f1_knn_all += f1_knn
         logging.info("k最近邻分类器准确率为{}".format(score_knn))
-        
+        logging.info("k最近邻分类器召回率为{}".format(recall_knn))
+        logging.info("k最近邻分类器f1_score为{}".format(f1_knn))
+
         logging.info("正在训练SVM分类器")
-        svm_classifier = svm.SVC(C=2,kernel='rbf',gamma=10,decision_function_shape='ovr')
+        svm_classifier = svm.SVC(C=2, kernel='rbf', gamma=10, decision_function_shape='ovr')
         svm_classifier.fit(train_data, train_label)
+        svm_predict = svm_classifier.predict(test_data)
+        recall_svm = sklearn.metrics.recall_score(test_label, svm_predict, average="macro")
+        f1_svm = sklearn.metrics.f1_score(test_label, svm_predict, average="macro")
         score_svm = svm_classifier.score(test_data, test_label)
         if score_svm < score_svm_lowest:
             score_svm_lowest = score_svm
         score_svm_all = score_svm_all + score_svm
+        recall_svm_all += recall_svm
+        f1_svm_all += f1_svm
         logging.info("SVM分类器准确率为{}".format(score_svm))
-        
+        logging.info("SVM分类器召回率为{}".format(recall_svm))
+        logging.info("SVM分类器f1_score为{}".format(f1_svm))
+
         logging.info("正在训练朴素贝叶斯分类器")
         muNB_classifier = GaussianNB()
         muNB_classifier.fit(train_data, train_label)
+        muNB_predict = muNB_classifier.predict(test_data)
+        recall_nb = sklearn.metrics.recall_score(test_label, muNB_predict, average="macro")
+        f1_nb = sklearn.metrics.f1_score(test_label, muNB_predict, average="macro")
         score_nb = muNB_classifier.score(test_data, test_label)
+
         if score_nb < score_nb_lowest:
             score_nb_lowest = score_nb
         score_nb_all = score_nb_all + score_nb
+        recall_nb_all += recall_nb
+        f1_nb_all += f1_nb
         logging.info("朴素贝叶斯分类器准确率为{}".format(score_nb))
+        logging.info("朴素贝叶斯分类器召回率为{}".format(recall_nb))
+        logging.info("朴素贝叶斯分类器f1_score为{}".format(f1_nb))
+
     logging.info("k最近邻分类器最低准确率为{}".format(score_knn_lowest))
     logging.info("SVM分类器最低准确率为{}".format(score_svm_lowest))
     logging.info("朴素贝叶斯分类器最低准确率为{}".format(score_nb_lowest))
     logging.info("k最近邻分类器平均准确率为{}".format(score_knn_all / random_state))
     logging.info("SVM分类器平均准确率为{}".format(score_svm_all / random_state))
     logging.info("朴素贝叶斯分类器平均准确率为{}".format(score_nb_all / random_state))
-    
+    logging.info("k最近邻分类器平均召回率为{}".format(recall_knn_all / random_state))
+    logging.info("SVM分类器平均召回率为{}".format(recall_svm_all / random_state))
+    logging.info("朴素贝叶斯分类器平均召回率为{}".format(recall_nb_all / random_state))
+    logging.info("k最近邻分类器平均f1_score为{}".format(f1_knn_all / random_state))
+    logging.info("SVM分类器平均f1_score为{}".format(f1_svm_all / random_state))
+    logging.info("朴素贝叶斯分类器平均f1_score为{}".format(f1_nb_all / random_state))
+
 
-    
 def text_to_feature(text, word2index, max_len):
     feature = []
     for word in text:
@@ -99,10 +143,17 @@ def text_to_feature(text, word2index, max_len):
             feature.append(word2index[word])
         else:
             feature.append(word2index["<unk>"])
-        if(len(feature) == max_len):
+        if len(feature) == max_len:
             break
     feature = feature + [word2index["<pad>"]] * (max_len - len(feature))
     return feature
 
+
+def calculate_bi_standards(name):
+    model = torch.load(name)
+
+    pass
+
+
 if __name__ == "__main__":
     contrast()

classify_service/data_processor.py

@@ -183,13 +183,14 @@ class DataProcessor(object):
         # 注，由于nn.Embedding每次生成的词嵌入不固定，因此此处同时获取训练数据的词嵌入和测试数据的词嵌入
         # 测试数据的词表也用训练数据创建
 
-        logging.info('正在从数据库读取原始数据')
-        txt_origin, label_origin = self.read_text_from_db()
-
+        # logging.info('正在从数据库读取原始数据')
+        # txt_origin, label_origin = self.read_text_from_db()
+        # txt_origin = np.load(self.datas_path, allow_pickle=True).tolist()
+        # label_origin = np.load(self.labels_path, allow_pickle=True).tolist()
         logging.info('正在对原始数据进行数据扩增')
-        txt_origin, label_origin = self.increase_data()
-        # txt_origin = np.load(self.datas_increase_path, allow_pickle=True).tolist()
-        # label_origin = np.load(self.labels_increase_path, allow_pickle=True).tolist()
+        # txt_origin, label_origin = self.increase_data()
+        txt_origin = np.load(self.datas_increase_path, allow_pickle=True).tolist()
+        label_origin = np.load(self.labels_increase_path, allow_pickle=True).tolist()
 
         label_count = {0: 0, 1: 0, 2: 0, 3: 0, 4: 0, 5: 0}
         for i in label_origin:

classify_service/model_train.log

@@ -274,3 +274,467 @@ INFO:root:正在从数据库读取原始数据
 09/06/2022 09:04:33 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.18221373733188662
 09/06/2022 09:04:33 [INFO] data_processor: 正在从数据库读取原始数据
 09/06/2022 09:10:11 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 13:00:44 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 13:00:44 [INFO] data_processor: 正在制作词表
+09/06/2022 13:00:44 [INFO] data_processor: 正在获取词向量
+09/06/2022 13:00:44 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 13:00:44 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 13:00:44 [INFO] bilstm_attention: 模型初始化
+09/06/2022 13:00:44 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 13:01:00 [INFO] bilstm_attention: 初始分类器accuracy为0.4664429530201342
+09/06/2022 13:01:00 [INFO] bilstm_attention: 初始分类器召回率为0.24035087719298245
+09/06/2022 13:01:00 [INFO] bilstm_attention: 初始分类器precision为0.1123355263157895
+09/06/2022 13:01:00 [INFO] bilstm_attention: 初始分类器f1_score为0.15088844742849322
+09/06/2022 13:01:02 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 13:01:33 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 13:02:07 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 13:02:49 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 13:03:32 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 13:04:16 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 13:05:00 [INFO] bilstm_attention: 开始第7次重训练
+09/06/2022 13:05:44 [INFO] bilstm_attention: 开始第8次重训练
+09/06/2022 13:06:29 [INFO] bilstm_attention: 开始第9次重训练
+09/06/2022 13:07:13 [INFO] bilstm_attention: 开始第10次重训练
+09/06/2022 13:07:58 [INFO] bilstm_attention: 开始第11次重训练
+09/06/2022 13:08:42 [INFO] bilstm_attention: 开始第12次重训练
+09/06/2022 13:09:26 [INFO] bilstm_attention: 开始第13次重训练
+09/06/2022 13:10:10 [INFO] bilstm_attention: 开始第14次重训练
+09/06/2022 13:10:59 [INFO] bilstm_attention: 开始第15次重训练
+09/06/2022 13:11:45 [INFO] bilstm_attention: 开始第16次重训练
+09/06/2022 13:12:31 [INFO] bilstm_attention: 开始第17次重训练
+09/06/2022 13:13:16 [INFO] bilstm_attention: 开始第18次重训练
+09/06/2022 13:14:47 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.35906040268456374
+09/06/2022 13:14:47 [INFO] bilstm_attention: 训练完成，测试集召回率为0.1565782511835144
+09/06/2022 13:14:47 [INFO] bilstm_attention: 训练完成，测试集Precision为0.10065357644305015
+09/06/2022 13:14:47 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.11665934562881802
+以上是 sql3读出的结果↑
+09/06/2022 13:19:20 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 13:19:22 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 13:19:22 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 13:19:22 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 13:19:39 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 13:19:40 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 13:19:40 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 13:19:40 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 13:19:40 [INFO] data_processor: 正在制作词表
+09/06/2022 13:19:40 [INFO] data_processor: 正在获取词向量
+09/06/2022 13:19:40 [INFO] bilstm_attention: 开始训练模型:决赛自主可控众测web自主可控运维管理系统
+09/06/2022 13:19:40 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 13:19:40 [INFO] bilstm_attention: 模型初始化
+09/06/2022 13:19:40 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 13:19:59 [INFO] bilstm_attention: 初始分类器accuracy为0.5275
+09/06/2022 13:19:59 [INFO] bilstm_attention: 初始分类器召回率为0.2619166666666666
+09/06/2022 13:19:59 [INFO] bilstm_attention: 初始分类器precision为0.14285416666666667
+09/06/2022 13:19:59 [INFO] bilstm_attention: 初始分类器f1_score为0.1802682066489447
+09/06/2022 13:20:03 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 13:20:27 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 13:21:04 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 13:21:48 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 13:22:40 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 13:23:33 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 13:24:28 [INFO] bilstm_attention: 开始第7次重训练
+09/06/2022 13:25:23 [INFO] bilstm_attention: 开始第8次重训练
+09/06/2022 13:26:19 [INFO] bilstm_attention: 开始第9次重训练
+09/06/2022 13:27:16 [INFO] bilstm_attention: 开始第10次重训练
+09/06/2022 13:28:16 [INFO] bilstm_attention: 开始第11次重训练
+09/06/2022 13:29:14 [INFO] bilstm_attention: 开始第12次重训练
+09/06/2022 13:30:12 [INFO] bilstm_attention: 开始第13次重训练
+09/06/2022 13:31:10 [INFO] bilstm_attention: 开始第14次重训练
+09/06/2022 13:32:09 [INFO] bilstm_attention: 开始第15次重训练
+09/06/2022 13:33:08 [INFO] bilstm_attention: 开始第16次重训练
+09/06/2022 13:34:08 [INFO] bilstm_attention: 开始第17次重训练
+09/06/2022 13:35:10 [INFO] bilstm_attention: 开始第18次重训练
+09/06/2022 13:36:13 [INFO] bilstm_attention: 开始第19次重训练
+09/06/2022 13:37:16 [INFO] bilstm_attention: 开始第20次重训练
+09/06/2022 13:38:19 [INFO] bilstm_attention: 开始第21次重训练
+09/06/2022 13:39:22 [INFO] bilstm_attention: 开始第22次重训练
+09/06/2022 13:40:23 [INFO] bilstm_attention: 开始第23次重训练
+09/06/2022 13:41:24 [INFO] bilstm_attention: 开始第24次重训练
+09/06/2022 13:43:27 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.33875
+09/06/2022 13:43:27 [INFO] bilstm_attention: 训练完成，测试集召回率为0.26684637769637765
+09/06/2022 13:43:27 [INFO] bilstm_attention: 训练完成，测试集Precision为0.2025890183890184
+09/06/2022 13:43:27 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.19970878392333496
+09/06/2022 13:43:27 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 13:43:27 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 13:43:28 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 13:43:28 [INFO] data_processor: 正在制作词表
+09/06/2022 13:43:28 [INFO] data_processor: 正在获取词向量
+09/06/2022 13:43:28 [INFO] bilstm_attention: 开始训练模型:航天中认自主可控众包测试练习赛
+09/06/2022 13:43:28 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 13:43:28 [INFO] bilstm_attention: 模型初始化
+09/06/2022 13:43:28 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 13:43:42 [INFO] bilstm_attention: 初始分类器accuracy为0.508833922261484
+09/06/2022 13:43:42 [INFO] bilstm_attention: 初始分类器召回率为0.2717592592592592
+09/06/2022 13:43:42 [INFO] bilstm_attention: 初始分类器precision为0.14042245370370368
+09/06/2022 13:43:42 [INFO] bilstm_attention: 初始分类器f1_score为0.18221373733188662
+09/06/2022 13:43:44 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 13:44:16 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 13:45:17 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.508833922261484
+09/06/2022 13:45:17 [INFO] bilstm_attention: 训练完成，测试集召回率为0.2717592592592592
+09/06/2022 13:45:17 [INFO] bilstm_attention: 训练完成，测试集Precision为0.14042245370370368
+09/06/2022 13:45:17 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.18221373733188662
+09/06/2022 13:45:17 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 13:45:17 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 13:45:17 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 13:45:17 [INFO] data_processor: 正在制作词表
+09/06/2022 13:45:17 [INFO] data_processor: 正在获取词向量
+09/06/2022 13:45:17 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 13:45:17 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 13:45:17 [INFO] bilstm_attention: 模型初始化
+09/06/2022 13:45:17 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 13:45:32 [INFO] bilstm_attention: 初始分类器accuracy为0.46476510067114096
+09/06/2022 13:45:32 [INFO] bilstm_attention: 初始分类器召回率为0.2387426900584795
+09/06/2022 13:45:32 [INFO] bilstm_attention: 初始分类器precision为0.11182383040935673
+09/06/2022 13:45:32 [INFO] bilstm_attention: 初始分类器f1_score为0.1501061887514977
+09/06/2022 13:45:35 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 13:46:05 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 13:46:37 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 13:47:09 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 13:47:42 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 13:48:15 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 13:48:48 [INFO] bilstm_attention: 开始第7次重训练
+09/06/2022 13:49:22 [INFO] bilstm_attention: 开始第8次重训练
+09/06/2022 13:49:56 [INFO] bilstm_attention: 开始第9次重训练
+09/06/2022 13:50:30 [INFO] bilstm_attention: 开始第10次重训练
+09/06/2022 13:51:05 [INFO] bilstm_attention: 开始第11次重训练
+09/06/2022 13:51:41 [INFO] bilstm_attention: 开始第12次重训练
+09/06/2022 13:52:15 [INFO] bilstm_attention: 开始第13次重训练
+09/06/2022 13:52:50 [INFO] bilstm_attention: 开始第14次重训练
+09/06/2022 13:53:25 [INFO] bilstm_attention: 开始第15次重训练
+09/06/2022 13:54:00 [INFO] bilstm_attention: 开始第16次重训练
+09/06/2022 13:54:35 [INFO] bilstm_attention: 开始第17次重训练
+09/06/2022 13:55:10 [INFO] bilstm_attention: 开始第18次重训练
+09/06/2022 13:55:46 [INFO] bilstm_attention: 开始第19次重训练
+09/06/2022 13:56:21 [INFO] bilstm_attention: 开始第20次重训练
+09/06/2022 13:56:58 [INFO] bilstm_attention: 开始第21次重训练
+09/06/2022 13:58:11 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.2348993288590604
+09/06/2022 13:58:11 [INFO] bilstm_attention: 训练完成，测试集召回率为0.19923402255639094
+09/06/2022 13:58:11 [INFO] bilstm_attention: 训练完成，测试集Precision为0.1218473522091943
+09/06/2022 13:58:11 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.12357362645094842
+以上为 学长increase.npy跑的内容↑
+09/06/2022 14:36:53 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 14:36:54 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 14:36:54 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 14:36:54 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 15:03:59 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 15:04:00 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 15:11:55 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 15:11:55 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 15:11:55 [INFO] data_processor: 正在制作词表
+09/06/2022 15:11:55 [INFO] data_processor: 正在获取词向量
+09/06/2022 15:11:55 [INFO] bilstm_attention: 开始训练模型:决赛自主可控众测web自主可控运维管理系统
+09/06/2022 15:11:55 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 15:11:55 [INFO] bilstm_attention: 模型初始化
+09/06/2022 15:11:55 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 15:11:58 [INFO] bilstm_attention: 初始分类器accuracy为0.5112781954887218
+09/06/2022 15:11:58 [INFO] bilstm_attention: 初始分类器召回率为0.26296296296296295
+09/06/2022 15:11:58 [INFO] bilstm_attention: 初始分类器precision为0.1265046296296296
+09/06/2022 15:11:58 [INFO] bilstm_attention: 初始分类器f1_score为0.1678516866922664
+09/06/2022 15:11:59 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 15:12:06 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 15:12:13 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 15:12:22 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 15:12:32 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 15:12:42 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 15:12:53 [INFO] bilstm_attention: 开始第7次重训练
+09/06/2022 15:13:03 [INFO] bilstm_attention: 开始第8次重训练
+09/06/2022 15:13:14 [INFO] bilstm_attention: 开始第9次重训练
+09/06/2022 15:13:25 [INFO] bilstm_attention: 开始第10次重训练
+09/06/2022 15:13:36 [INFO] bilstm_attention: 开始第11次重训练
+09/06/2022 15:14:01 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.5112781954887218
+09/06/2022 15:14:01 [INFO] bilstm_attention: 训练完成，测试集召回率为0.26296296296296295
+09/06/2022 15:14:01 [INFO] bilstm_attention: 训练完成，测试集Precision为0.1265046296296296
+09/06/2022 15:14:01 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.1678516866922664
+09/06/2022 15:14:01 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 15:19:29 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 15:19:29 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 15:19:29 [INFO] data_processor: 正在制作词表
+09/06/2022 15:19:29 [INFO] data_processor: 正在获取词向量
+09/06/2022 15:19:29 [INFO] bilstm_attention: 开始训练模型:航天中认自主可控众包测试练习赛
+09/06/2022 15:19:29 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 15:19:29 [INFO] bilstm_attention: 模型初始化
+09/06/2022 15:19:29 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 15:19:31 [INFO] bilstm_attention: 初始分类器accuracy为0.43617021276595747
+09/06/2022 15:19:31 [INFO] bilstm_attention: 初始分类器召回率为0.2833333333333333
+09/06/2022 15:19:31 [INFO] bilstm_attention: 初始分类器precision为0.12135416666666667
+09/06/2022 15:19:31 [INFO] bilstm_attention: 初始分类器f1_score为0.16691483503077706
+09/06/2022 15:19:32 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 15:19:39 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 15:19:47 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 15:19:55 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 15:20:04 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 15:20:22 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.43617021276595747
+09/06/2022 15:20:22 [INFO] bilstm_attention: 训练完成，测试集召回率为0.2833333333333333
+09/06/2022 15:20:22 [INFO] bilstm_attention: 训练完成，测试集Precision为0.12135416666666667
+09/06/2022 15:20:22 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.16691483503077706
+09/06/2022 15:20:22 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 15:26:13 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 15:26:13 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 15:26:13 [INFO] data_processor: 正在制作词表
+09/06/2022 15:26:13 [INFO] data_processor: 正在获取词向量
+09/06/2022 15:26:13 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 15:26:13 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 15:26:13 [INFO] bilstm_attention: 模型初始化
+09/06/2022 15:26:13 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 15:26:16 [INFO] bilstm_attention: 初始分类器accuracy为0.46464646464646464
+09/06/2022 15:26:16 [INFO] bilstm_attention: 初始分类器召回率为0.24863945578231292
+09/06/2022 15:26:16 [INFO] bilstm_attention: 初始分类器precision为0.1888214959643531
+09/06/2022 15:26:16 [INFO] bilstm_attention: 初始分类器f1_score为0.19446749268269178
+09/06/2022 15:26:17 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 15:26:25 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 15:26:34 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 15:26:44 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 15:26:53 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 15:27:11 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.494949494949495
+09/06/2022 15:27:11 [INFO] bilstm_attention: 训练完成，测试集召回率为0.22380952380952382
+09/06/2022 15:27:11 [INFO] bilstm_attention: 训练完成，测试集Precision为0.11919642857142856
+09/06/2022 15:27:11 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.15395511627395683
+09/06/2022 16:35:59 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 16:36:38 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 16:36:39 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 16:41:58 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 17:12:11 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 17:12:13 [INFO] data_processor: 正在从数据库读取原始数据
+09/06/2022 17:17:59 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 21:01:48 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 21:01:48 [INFO] data_processor: 正在制作词表
+09/06/2022 21:01:48 [INFO] data_processor: 正在获取词向量
+09/06/2022 21:01:48 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 21:01:48 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 21:01:48 [INFO] bilstm_attention: 模型初始化
+09/06/2022 21:01:48 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 21:02:04 [INFO] bilstm_attention: 初始分类器accuracy为0.4664429530201342
+09/06/2022 21:02:04 [INFO] bilstm_attention: 初始分类器召回率为0.24035087719298245
+09/06/2022 21:02:04 [INFO] bilstm_attention: 初始分类器precision为0.1123355263157895
+09/06/2022 21:02:04 [INFO] bilstm_attention: 初始分类器f1_score为0.15088844742849322
+09/06/2022 21:02:07 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 21:02:26 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 21:02:46 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 21:03:05 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 21:03:25 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 21:03:45 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 21:04:29 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.4664429530201342
+09/06/2022 21:04:29 [INFO] bilstm_attention: 训练完成，测试集召回率为0.24035087719298245
+09/06/2022 21:04:29 [INFO] bilstm_attention: 训练完成，测试集Precision为0.1123355263157895
+09/06/2022 21:04:29 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.15088844742849322
+09/06/2022 22:26:08 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 22:26:09 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 22:26:09 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 22:27:25 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 22:27:27 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 22:27:27 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 22:27:27 [INFO] data_processor: 正在制作词表
+09/06/2022 22:27:27 [INFO] data_processor: 正在获取词向量
+09/06/2022 22:27:27 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 22:27:27 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 22:27:27 [INFO] bilstm_attention: 模型初始化
+09/06/2022 22:27:27 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 22:29:56 [INFO] bilstm_attention: 初始分类器accuracy为0.46140939597315433
+09/06/2022 22:29:56 [INFO] bilstm_attention: 初始分类器召回率为0.2612202380952381
+09/06/2022 22:29:56 [INFO] bilstm_attention: 初始分类器precision为0.18331566094723992
+09/06/2022 22:29:56 [INFO] bilstm_attention: 初始分类器f1_score为0.1919964770355044
+09/06/2022 22:29:59 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 22:33:36 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 22:33:53 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 22:33:54 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 22:33:54 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 22:33:54 [INFO] data_processor: 正在制作词表
+09/06/2022 22:33:54 [INFO] data_processor: 正在获取词向量
+09/06/2022 22:33:54 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 22:33:54 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 22:33:54 [INFO] bilstm_attention: 模型初始化
+09/06/2022 22:33:54 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 22:34:23 [INFO] bilstm_attention: 初始分类器accuracy为0.46140939597315433
+09/06/2022 22:34:23 [INFO] bilstm_attention: 初始分类器召回率为0.2622906223893066
+09/06/2022 22:34:23 [INFO] bilstm_attention: 初始分类器precision为0.1856951674056937
+09/06/2022 22:34:23 [INFO] bilstm_attention: 初始分类器f1_score为0.1917470831199303
+09/06/2022 22:34:25 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 22:35:17 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 22:36:12 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 22:37:14 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 22:38:18 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 22:39:27 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 22:40:39 [INFO] bilstm_attention: 开始第7次重训练
+09/06/2022 22:41:54 [INFO] bilstm_attention: 开始第8次重训练
+09/06/2022 22:43:11 [INFO] bilstm_attention: 开始第9次重训练
+09/06/2022 22:44:29 [INFO] bilstm_attention: 开始第10次重训练
+09/06/2022 22:45:46 [INFO] bilstm_attention: 开始第11次重训练
+09/06/2022 22:47:02 [INFO] bilstm_attention: 开始第12次重训练
+09/06/2022 22:48:20 [INFO] bilstm_attention: 开始第13次重训练
+09/06/2022 22:49:40 [INFO] bilstm_attention: 开始第14次重训练
+09/06/2022 22:50:59 [INFO] bilstm_attention: 开始第15次重训练
+09/06/2022 22:53:36 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.4077181208053691
+09/06/2022 22:53:36 [INFO] bilstm_attention: 训练完成，测试集召回率为0.2516920774157616
+09/06/2022 22:53:36 [INFO] bilstm_attention: 训练完成，测试集Precision为0.20253697326065748
+09/06/2022 22:53:36 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.19972863037090352
+09/06/2022 23:43:27 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 23:43:29 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 23:43:29 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 23:43:29 [INFO] data_processor: 正在制作词表
+09/06/2022 23:43:29 [INFO] data_processor: 正在获取词向量
+09/06/2022 23:43:29 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 23:43:29 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 23:43:29 [INFO] bilstm_attention: 模型初始化
+09/06/2022 23:43:29 [INFO] bilstm_attention: 开始训练基础分类器
+09/06/2022 23:43:44 [INFO] bilstm_attention: 初始分类器accuracy为0.4664429530201342
+09/06/2022 23:43:44 [INFO] bilstm_attention: 初始分类器召回率为0.24035087719298245
+09/06/2022 23:43:44 [INFO] bilstm_attention: 初始分类器precision为0.1123355263157895
+09/06/2022 23:43:44 [INFO] bilstm_attention: 初始分类器f1_score为0.15088844742849322
+09/06/2022 23:43:46 [INFO] bilstm_attention: 开始第1次重训练
+09/06/2022 23:44:05 [INFO] bilstm_attention: 开始第2次重训练
+09/06/2022 23:44:23 [INFO] bilstm_attention: 开始第3次重训练
+09/06/2022 23:44:42 [INFO] bilstm_attention: 开始第4次重训练
+09/06/2022 23:45:01 [INFO] bilstm_attention: 开始第5次重训练
+09/06/2022 23:45:20 [INFO] bilstm_attention: 开始第6次重训练
+09/06/2022 23:46:04 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.4664429530201342
+09/06/2022 23:46:04 [INFO] bilstm_attention: 训练完成，测试集召回率为0.24035087719298245
+09/06/2022 23:46:04 [INFO] bilstm_attention: 训练完成，测试集Precision为0.1123355263157895
+09/06/2022 23:46:04 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.15088844742849322
+09/06/2022 23:47:34 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/06/2022 23:47:36 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/06/2022 23:47:36 [INFO] data_processor: 正在统计原始数据的标签类型
+09/06/2022 23:47:36 [INFO] data_processor: 正在制作词表
+09/06/2022 23:47:36 [INFO] data_processor: 正在获取词向量
+09/06/2022 23:47:36 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/06/2022 23:47:36 [INFO] bilstm_attention: pytorch 初始化
+09/06/2022 23:47:36 [INFO] bilstm_attention: 模型初始化
+09/06/2022 23:47:36 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:06:09 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 00:06:10 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 00:06:10 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 00:06:10 [INFO] data_processor: 正在制作词表
+09/07/2022 00:06:10 [INFO] data_processor: 正在获取词向量
+09/07/2022 00:06:10 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 00:06:10 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 00:06:10 [INFO] bilstm_attention: 模型初始化
+09/07/2022 00:06:10 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:06:22 [INFO] bilstm_attention: 初始分类器accuracy为0.4664429530201342
+09/07/2022 00:06:22 [INFO] bilstm_attention: 初始分类器召回率为0.26258040935672505
+09/07/2022 00:06:22 [INFO] bilstm_attention: 初始分类器precision为0.16590133185527922
+09/07/2022 00:06:22 [INFO] bilstm_attention: 初始分类器f1_score为0.1871790422476922
+09/07/2022 00:06:24 [INFO] bilstm_attention: 开始第1次重训练
+09/07/2022 00:06:43 [INFO] bilstm_attention: 开始第2次重训练
+09/07/2022 00:07:12 [INFO] bilstm_attention: 开始第3次重训练
+09/07/2022 00:07:44 [INFO] bilstm_attention: 开始第4次重训练
+09/07/2022 00:08:20 [INFO] bilstm_attention: 开始第5次重训练
+09/07/2022 00:08:54 [INFO] bilstm_attention: 开始第6次重训练
+09/07/2022 00:09:28 [INFO] bilstm_attention: 开始第7次重训练
+09/07/2022 00:10:04 [INFO] bilstm_attention: 开始第8次重训练
+09/07/2022 00:10:39 [INFO] bilstm_attention: 开始第9次重训练
+09/07/2022 00:11:16 [INFO] bilstm_attention: 开始第10次重训练
+09/07/2022 00:11:55 [INFO] bilstm_attention: 开始第11次重训练
+09/07/2022 00:12:30 [INFO] bilstm_attention: 开始第12次重训练
+09/07/2022 00:13:06 [INFO] bilstm_attention: 开始第13次重训练
+09/07/2022 00:20:28 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 00:20:29 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 00:20:29 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 00:20:29 [INFO] data_processor: 正在制作词表
+09/07/2022 00:20:29 [INFO] data_processor: 正在获取词向量
+09/07/2022 00:20:29 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 00:20:29 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 00:20:29 [INFO] bilstm_attention: 模型初始化
+09/07/2022 00:20:29 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:20:41 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 00:20:42 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 00:20:42 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 00:20:42 [INFO] data_processor: 正在制作词表
+09/07/2022 00:20:42 [INFO] data_processor: 正在获取词向量
+09/07/2022 00:20:43 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 00:20:43 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 00:20:43 [INFO] bilstm_attention: 模型初始化
+09/07/2022 00:20:43 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:20:45 [INFO] bilstm_attention: 初始分类器accuracy为0.46464646464646464
+09/07/2022 00:20:45 [INFO] bilstm_attention: 初始分类器召回率为0.2265808596165739
+09/07/2022 00:20:45 [INFO] bilstm_attention: 初始分类器precision为0.16216422466422467
+09/07/2022 00:20:45 [INFO] bilstm_attention: 初始分类器f1_score为0.18099461832707991
+09/07/2022 00:20:46 [INFO] bilstm_attention: 开始第1次重训练
+09/07/2022 00:20:50 [INFO] bilstm_attention: 开始第2次重训练
+09/07/2022 00:20:56 [INFO] bilstm_attention: 开始第3次重训练
+09/07/2022 00:21:03 [INFO] bilstm_attention: 开始第4次重训练
+09/07/2022 00:21:10 [INFO] bilstm_attention: 开始第5次重训练
+09/07/2022 00:21:17 [INFO] bilstm_attention: 开始第6次重训练
+09/07/2022 00:21:25 [INFO] bilstm_attention: 开始第7次重训练
+09/07/2022 00:21:33 [INFO] bilstm_attention: 开始第8次重训练
+09/07/2022 00:21:41 [INFO] bilstm_attention: 开始第9次重训练
+09/07/2022 00:21:58 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.29292929292929293
+09/07/2022 00:21:58 [INFO] bilstm_attention: 训练完成，测试集召回率为0.2952380952380952
+09/07/2022 00:21:58 [INFO] bilstm_attention: 训练完成，测试集Precision为0.08824404761904761
+09/07/2022 00:21:58 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.13490381798652476
+09/07/2022 00:28:18 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 00:28:20 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 00:28:20 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 00:28:20 [INFO] data_processor: 正在制作词表
+09/07/2022 00:28:20 [INFO] data_processor: 正在获取词向量
+09/07/2022 00:28:20 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 00:28:20 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 00:28:20 [INFO] bilstm_attention: 模型初始化
+09/07/2022 00:28:20 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:28:29 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 00:28:31 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 00:28:31 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 00:28:31 [INFO] data_processor: 正在制作词表
+09/07/2022 00:28:31 [INFO] data_processor: 正在获取词向量
+09/07/2022 00:28:31 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 00:28:31 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 00:28:31 [INFO] bilstm_attention: 模型初始化
+09/07/2022 00:28:31 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:28:33 [INFO] bilstm_attention: 初始分类器accuracy为0.494949494949495
+09/07/2022 00:28:33 [INFO] bilstm_attention: 初始分类器召回率为0.24995748299319726
+09/07/2022 00:28:33 [INFO] bilstm_attention: 初始分类器precision为0.22051282051282053
+09/07/2022 00:28:33 [INFO] bilstm_attention: 初始分类器f1_score为0.21254877845266865
+09/07/2022 00:28:34 [INFO] bilstm_attention: 开始第1次重训练
+09/07/2022 00:28:38 [INFO] bilstm_attention: 开始第2次重训练
+09/07/2022 00:28:45 [INFO] bilstm_attention: 开始第3次重训练
+09/07/2022 00:28:52 [INFO] bilstm_attention: 开始第4次重训练
+09/07/2022 00:29:00 [INFO] bilstm_attention: 开始第5次重训练
+09/07/2022 00:29:09 [INFO] bilstm_attention: 开始第6次重训练
+09/07/2022 00:29:17 [INFO] bilstm_attention: 开始第7次重训练
+09/07/2022 00:29:25 [INFO] bilstm_attention: 开始第8次重训练
+09/07/2022 00:29:34 [INFO] bilstm_attention: 开始第9次重训练
+09/07/2022 00:29:42 [INFO] bilstm_attention: 开始第10次重训练
+09/07/2022 00:29:50 [INFO] bilstm_attention: 开始第11次重训练
+09/07/2022 00:29:59 [INFO] bilstm_attention: 开始第12次重训练
+09/07/2022 00:30:07 [INFO] bilstm_attention: 开始第13次重训练
+09/07/2022 00:30:25 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.48484848484848486
+09/07/2022 00:30:25 [INFO] bilstm_attention: 训练完成，测试集召回率为0.2901089981447124
+09/07/2022 00:30:25 [INFO] bilstm_attention: 训练完成，测试集Precision为0.23915515701229986
+09/07/2022 00:30:25 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.253612168705048
+09/07/2022 00:31:02 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 00:31:03 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 00:31:03 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 00:31:03 [INFO] data_processor: 正在制作词表
+09/07/2022 00:31:03 [INFO] data_processor: 正在获取词向量
+09/07/2022 00:31:03 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 00:31:03 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 00:31:03 [INFO] bilstm_attention: 模型初始化
+09/07/2022 00:31:03 [INFO] bilstm_attention: 开始训练基础分类器
+09/07/2022 00:31:18 [INFO] bilstm_attention: 初始分类器accuracy为0.4664429530201342
+09/07/2022 00:31:18 [INFO] bilstm_attention: 初始分类器召回率为0.24035087719298245
+09/07/2022 00:31:18 [INFO] bilstm_attention: 初始分类器precision为0.1123355263157895
+09/07/2022 00:31:18 [INFO] bilstm_attention: 初始分类器f1_score为0.15088844742849322
+09/07/2022 00:31:21 [INFO] bilstm_attention: 开始第1次重训练
+09/07/2022 00:31:42 [INFO] bilstm_attention: 开始第2次重训练
+09/07/2022 00:32:12 [INFO] bilstm_attention: 开始第3次重训练
+09/07/2022 00:32:44 [INFO] bilstm_attention: 开始第4次重训练
+09/07/2022 00:33:16 [INFO] bilstm_attention: 开始第5次重训练
+09/07/2022 00:33:50 [INFO] bilstm_attention: 开始第6次重训练
+09/07/2022 00:34:26 [INFO] bilstm_attention: 开始第7次重训练
+09/07/2022 00:35:01 [INFO] bilstm_attention: 开始第8次重训练
+09/07/2022 00:35:39 [INFO] bilstm_attention: 开始第9次重训练
+09/07/2022 00:36:15 [INFO] bilstm_attention: 开始第10次重训练
+09/07/2022 00:36:55 [INFO] bilstm_attention: 开始第11次重训练
+09/07/2022 00:37:32 [INFO] bilstm_attention: 开始第12次重训练
+09/07/2022 00:38:09 [INFO] bilstm_attention: 开始第13次重训练
+09/07/2022 00:39:23 [INFO] bilstm_attention: 训练完成，测试集Accuracy为0.4664429530201342
+09/07/2022 00:39:23 [INFO] bilstm_attention: 训练完成，测试集召回率为0.24035087719298245
+09/07/2022 00:39:23 [INFO] bilstm_attention: 训练完成，测试集Precision为0.1123355263157895
+09/07/2022 00:39:23 [INFO] bilstm_attention: 训练完成，测试集f1_score为0.15088844742849322
+09/07/2022 15:05:27 [DEBUG] tpu_cluster_resolver: Falling back to TensorFlow client; we recommended you install the Cloud TPU client directly with pip install cloud-tpu-client.
+09/07/2022 15:05:28 [INFO] data_processor: 正在对原始数据进行数据扩增
+09/07/2022 15:05:28 [INFO] data_processor: 正在统计原始数据的标签类型
+09/07/2022 15:05:28 [INFO] data_processor: 正在制作词表
+09/07/2022 15:05:28 [INFO] data_processor: 正在获取词向量
+09/07/2022 15:05:29 [INFO] bilstm_attention: 开始训练模型:趣享GIF众包测试201908试题
+09/07/2022 15:05:29 [INFO] bilstm_attention: pytorch 初始化
+09/07/2022 15:05:29 [INFO] bilstm_attention: 模型初始化
+09/07/2022 15:05:29 [INFO] bilstm_attention: 开始训练基础分类器