{"id":15326417,"url":"https://github.com/jieguangzhou/textclassification","last_synced_at":"2025-10-24T12:38:14.108Z","repository":{"id":108643355,"uuid":"154838696","full_name":"jieguangzhou/TextClassification","owner":"jieguangzhou","description":"基于tensorflow的文本分类 Text classification","archived":false,"fork":false,"pushed_at":"2018-10-30T15:27:10.000Z","size":309,"stargazers_count":3,"open_issues_count":0,"forks_count":2,"subscribers_count":1,"default_branch":"master","last_synced_at":"2024-10-15T22:41:15.774Z","etag":null,"topics":["cnn","deep-learning","fasttext","python","rnn","tensorflow","textcnn","textrnn"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/jieguangzhou.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2018-10-26T13:26:33.000Z","updated_at":"2022-01-22T07:31:37.000Z","dependencies_parsed_at":null,"dependency_job_id":"496162a9-aa6a-4c3f-94fa-954fd60d58a1","html_url":"https://github.com/jieguangzhou/TextClassification","commit_stats":{"total_commits":8,"total_committers":1,"mean_commits":8.0,"dds":0.0,"last_synced_commit":"8e0d3864eee44d09bba7b01183f0923492f52e12"},"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jieguangzhou%2FTextClassification","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jieguangzhou%2FTextClassification/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jieguangzhou%2FTextClassification/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jieguangzhou%2FTextClassification/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/jieguangzhou","download_url":"https://codeload.github.com/jieguangzhou/TextClassification/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":228483627,"owners_count":17927363,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cnn","deep-learning","fasttext","python","rnn","tensorflow","textcnn","textrnn"],"created_at":"2024-10-01T09:35:40.977Z","updated_at":"2025-10-24T12:38:09.045Z","avatar_url":"https://github.com/jieguangzhou.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"## Text Classification\n\n​\t基于神经网络的文本分类模型实现包括TextRNN，TextCNN和fasttext。\n\n### 环境\n\n+ python3\n+ tensorflow1.9\n+ numpy\n\n### 数据集合\n\n​\t感谢[gaussic](https://github.com/gaussic/text-classification-cnn-rnn)提供的关于THUCNews的一个子集。\n\n### 可选参数\n\n参数文件可见[text_classification/opt.py](https://github.com/jieguangzhou/TextClassification/blob/master/text_classification/opt.py)\n\n```shell\nnn:\n  -model MODEL          use model, fasttext, textrnn or textcnn\n  -embedding_size EMBEDDING_SIZE\n                        embedding size\n  -vocab_size VOCAB_SIZE\n                        vocab size\n  -embedding_path EMBEDDING_PATH\n                        embedding path, 暂不使用\n  -keep_drop_prob KEEP_DROP_PROB\n                        keep_drop_prob\n  -class_num CLASS_NUM  class_num\n\nrnn:\n  -num_units NUM_UNITS  rnn cell hidden size\n  -layer_num LAYER_NUM  rnn layer number\n  -cell_type CELL_TYPE  rnn cell type, gru or lstm\n  -bidirectional        use bidirectional\n\ncnn:\n  -filter_num FILTER_NUM\n                        cnn filter num\n  -kernel_sizes KERNEL_SIZES [KERNEL_SIZES ...]\n                        cnn kernel_sizes, a list of int\n\ntrain:\n  -learning_rate LEARNING_RATE\n                        learning_rate\n  -batch_size BATCH_SIZE\n                        batch_size\n  -epoch_num EPOCH_NUM\n  -print_every_step PRINT_EVERY_STEP\n  -save_path SAVE_PATH\n\ndata:\n  -train_data TRAIN_DATA\n                        train data path\n  -val_data VAL_DATA    val data path\n  -test_data TEST_DATA  test data path\n  -vocab_path VOCAB_PATH\n                        vocab_pathe\n  -label_path LABEL_PATH\n                        label_path\n  -cut_length CUT_LENGTH\n                        cut_length\n  -reverse              reverse the sequence\n\n```\n\n\n\n### 训练与验证\n\n#### 1 FastText\n\n```shell\npython3.5 -m \"text_classification.main\" -model fasttext -save_path save/fasttext -epoch_num 5\n```\n\n```shell\ncreate model\n\n --------------------\nFastText : parms\nFasttext/embedding:0 (5000, 128)\nFasttext/dense/kernel:0 (128, 10)\nFasttext/dense/bias:0 (10,)\n-------------------- \n\nload data set\nEpoch: 1\nstep:   100, train loss:   2.2, train accuarcy: 42.23%, val loss :   2.2, val accuarcy: 34.26%, cost:0:00:01.963413\nstep:   200, train loss:   1.9, train accuarcy: 50.22%, val loss :   2.0, val accuarcy: 48.36%, cost:0:00:03.807075\nstep:   300, train loss:   1.6, train accuarcy: 67.36%, val loss :   1.7, val accuarcy: 59.76%, cost:0:00:05.653884\nstep:   400, train loss:   1.3, train accuarcy: 75.27%, val loss :   1.5, val accuarcy: 65.62%, cost:0:00:07.465123\nstep:   500, train loss:   1.0, train accuarcy: 79.73%, val loss :   1.3, val accuarcy: 68.18%, cost:0:00:09.304550\nstep:   600, train loss:  0.85, train accuarcy: 81.56%, val loss :   1.2, val accuarcy: 71.32%, cost:0:00:11.191120\nstep:   700, train loss:  0.74, train accuarcy: 83.41%, val loss :   1.0, val accuarcy: 72.88%, cost:0:00:13.064502\nEpoch: 2\nstep:   800, train loss:  0.63, train accuarcy: 85.42%, val loss :  0.91, val accuarcy: 74.66%, cost:0:00:00.761135\nstep:   900, train loss:  0.58, train accuarcy: 85.97%, val loss :  0.83, val accuarcy: 75.72%, cost:0:00:02.631185\nstep:  1000, train loss:  0.54, train accuarcy: 86.86%, val loss :  0.76, val accuarcy: 76.76%, cost:0:00:04.526994\nstep:  1100, train loss:  0.48, train accuarcy: 87.98%, val loss :  0.69, val accuarcy: 78.82%, cost:0:00:06.391131\nstep:  1200, train loss:  0.48, train accuarcy: 88.19%, val loss :  0.64, val accuarcy: 80.36%, cost:0:00:08.226393\nstep:  1300, train loss:  0.42, train accuarcy: 89.55%, val loss :  0.59, val accuarcy: 82.72%, cost:0:00:10.066286\nstep:  1400, train loss:  0.39, train accuarcy: 90.25%, val loss :  0.55, val accuarcy: 84.10%, cost:0:00:11.896519\nstep:  1500, train loss:  0.37, train accuarcy: 90.34%, val loss :  0.53, val accuarcy: 84.36%, cost:0:00:13.723557\nEpoch: 3\nstep:  1600, train loss:  0.36, train accuarcy: 90.36%, val loss :   0.5, val accuarcy: 85.36%, cost:0:00:00.998591\nstep:  1700, train loss:  0.34, train accuarcy: 91.38%, val loss :  0.47, val accuarcy: 86.24%, cost:0:00:02.844450\nstep:  1800, train loss:  0.32, train accuarcy: 91.50%, val loss :  0.46, val accuarcy: 86.76%, cost:0:00:04.688984\nstep:  1900, train loss:  0.33, train accuarcy: 91.97%, val loss :  0.45, val accuarcy: 87.16%, cost:0:00:06.528917\nstep:  2000, train loss:  0.29, train accuarcy: 92.62%, val loss :  0.43, val accuarcy: 87.98%, cost:0:00:08.379237\nstep:  2100, train loss:   0.3, train accuarcy: 92.39%, val loss :  0.41, val accuarcy: 88.14%, cost:0:00:10.242890\nstep:  2200, train loss:  0.29, train accuarcy: 92.28%, val loss :   0.4, val accuarcy: 88.18%, cost:0:00:12.107764\nstep:  2300, train loss:  0.29, train accuarcy: 92.44%, val loss :  0.38, val accuarcy: 89.04%, cost:0:00:13.989467\nEpoch: 4\nstep:  2400, train loss:  0.27, train accuarcy: 93.43%, val loss :  0.38, val accuarcy: 88.54%, cost:0:00:01.251994\nstep:  2500, train loss:  0.26, train accuarcy: 93.34%, val loss :  0.37, val accuarcy: 88.94%, cost:0:00:03.085236\nstep:  2600, train loss:  0.24, train accuarcy: 93.69%, val loss :  0.36, val accuarcy: 89.62%, cost:0:00:04.896725\nstep:  2700, train loss:  0.25, train accuarcy: 93.77%, val loss :  0.35, val accuarcy: 89.74%, cost:0:00:06.758444\nstep:  2800, train loss:  0.25, train accuarcy: 93.47%, val loss :  0.35, val accuarcy: 89.76%, cost:0:00:08.586772\nstep:  2900, train loss:  0.25, train accuarcy: 93.39%, val loss :  0.34, val accuarcy: 90.62%, cost:0:00:10.414334\nstep:  3000, train loss:  0.25, train accuarcy: 93.77%, val loss :  0.34, val accuarcy: 90.22%, cost:0:00:12.246485\nstep:  3100, train loss:  0.23, train accuarcy: 93.84%, val loss :  0.33, val accuarcy: 90.32%, cost:0:00:14.072625\nEpoch: 5\nstep:  3200, train loss:  0.23, train accuarcy: 94.27%, val loss :  0.33, val accuarcy: 90.54%, cost:0:00:01.473536\nstep:  3300, train loss:  0.22, train accuarcy: 94.28%, val loss :  0.32, val accuarcy: 90.82%, cost:0:00:03.316607\nstep:  3400, train loss:  0.21, train accuarcy: 94.50%, val loss :  0.31, val accuarcy: 91.04%, cost:0:00:05.151259\nstep:  3500, train loss:  0.21, train accuarcy: 94.48%, val loss :  0.31, val accuarcy: 91.08%, cost:0:00:07.003222\nstep:  3600, train loss:  0.22, train accuarcy: 94.28%, val loss :  0.31, val accuarcy: 91.22%, cost:0:00:08.854277\nstep:  3700, train loss:   0.2, train accuarcy: 94.59%, val loss :  0.31, val accuarcy: 91.16%, cost:0:00:10.716808\nstep:  3800, train loss:  0.21, train accuarcy: 94.25%, val loss :  0.31, val accuarcy: 91.18%, cost:0:00:12.528761\nstep:  3900, train loss:  0.21, train accuarcy: 94.80%, val loss :  0.31, val accuarcy: 91.40%, cost:0:00:14.358694\neval test data\nloss:  0.27, accuarcy: 92.09%, cost:0:00:15.470319\n```\n\n\n\n#### 2 TextRnn\n\n```shell\npython3.5 -m \"text_classification.main\" -model textrnn -save_path save/textrnn -epoch_num 5\n```\n\n```shell\ncreate model\n\n --------------------\nTextRNN : parms\nTextRnn/embedding:0 (5000, 128)\nTextRnn/Rnn/rnn/multi_rnn_cell/cell_0/gru_cell/gates/kernel:0 (192, 128)\nTextRnn/Rnn/rnn/multi_rnn_cell/cell_0/gru_cell/gates/bias:0 (128,)\nTextRnn/Rnn/rnn/multi_rnn_cell/cell_0/gru_cell/candidate/kernel:0 (192, 64)\nTextRnn/Rnn/rnn/multi_rnn_cell/cell_0/gru_cell/candidate/bias:0 (64,)\nTextRnn/dense/kernel:0 (64, 10)\nTextRnn/dense/bias:0 (10,)\n-------------------- \n\nload data set\nEpoch: 1\nstep:   100, train loss:   2.2, train accuarcy: 18.56%, val loss :   2.1, val accuarcy: 20.84%, cost:0:00:34.051950\nstep:   200, train loss:   2.1, train accuarcy: 26.08%, val loss :   2.1, val accuarcy: 21.56%, cost:0:01:07.838350\nstep:   300, train loss:   1.9, train accuarcy: 31.70%, val loss :   1.8, val accuarcy: 31.06%, cost:0:01:41.759550\nstep:   400, train loss:   1.8, train accuarcy: 34.08%, val loss :   1.9, val accuarcy: 28.04%, cost:0:02:15.653036\nstep:   500, train loss:   1.6, train accuarcy: 41.55%, val loss :   1.5, val accuarcy: 46.80%, cost:0:02:49.297939\nstep:   600, train loss:   1.3, train accuarcy: 52.55%, val loss :   1.6, val accuarcy: 48.78%, cost:0:03:23.266585\nstep:   700, train loss:   1.2, train accuarcy: 58.94%, val loss :   1.1, val accuarcy: 60.48%, cost:0:03:57.252852\nEpoch: 2\nstep:   800, train loss:  0.88, train accuarcy: 69.36%, val loss :  0.97, val accuarcy: 64.06%, cost:0:00:11.253570\nstep:   900, train loss:  0.75, train accuarcy: 76.33%, val loss :  0.89, val accuarcy: 69.96%, cost:0:00:45.180775\nstep:  1000, train loss:  0.62, train accuarcy: 82.56%, val loss :  0.82, val accuarcy: 77.70%, cost:0:01:19.110621\nstep:  1100, train loss:  0.54, train accuarcy: 85.58%, val loss :  0.74, val accuarcy: 78.06%, cost:0:01:53.019108\nstep:  1200, train loss:  0.48, train accuarcy: 88.23%, val loss :  0.49, val accuarcy: 86.38%, cost:0:02:27.000502\nstep:  1300, train loss:  0.43, train accuarcy: 89.41%, val loss :  0.45, val accuarcy: 88.58%, cost:0:03:01.183978\nstep:  1400, train loss:  0.37, train accuarcy: 91.00%, val loss :  0.42, val accuarcy: 88.12%, cost:0:03:34.768332\nstep:  1500, train loss:  0.38, train accuarcy: 90.30%, val loss :  0.41, val accuarcy: 88.46%, cost:0:04:08.398122\nEpoch: 3\nstep:  1600, train loss:  0.31, train accuarcy: 93.23%, val loss :  0.43, val accuarcy: 88.48%, cost:0:00:16.136509\nstep:  1700, train loss:  0.29, train accuarcy: 92.81%, val loss :  0.39, val accuarcy: 89.34%, cost:0:00:49.646578\nstep:  1800, train loss:   0.3, train accuarcy: 92.36%, val loss :  0.41, val accuarcy: 88.92%, cost:0:01:23.859286\nstep:  1900, train loss:  0.27, train accuarcy: 93.34%, val loss :  0.35, val accuarcy: 89.88%, cost:0:01:57.909257\nstep:  2000, train loss:  0.26, train accuarcy: 93.48%, val loss :  0.37, val accuarcy: 89.22%, cost:0:02:31.951416\nstep:  2100, train loss:   0.3, train accuarcy: 92.09%, val loss :  0.42, val accuarcy: 87.84%, cost:0:03:05.799133\nstep:  2200, train loss:  0.27, train accuarcy: 93.34%, val loss :  0.38, val accuarcy: 89.26%, cost:0:03:39.777590\nstep:  2300, train loss:  0.24, train accuarcy: 94.09%, val loss :  0.35, val accuarcy: 90.10%, cost:0:04:13.197220\nEpoch: 4\nstep:  2400, train loss:   0.2, train accuarcy: 95.52%, val loss :   0.3, val accuarcy: 91.94%, cost:0:00:21.144021\nstep:  2500, train loss:  0.18, train accuarcy: 95.62%, val loss :  0.32, val accuarcy: 91.62%, cost:0:00:54.737925\nstep:  2600, train loss:  0.19, train accuarcy: 95.33%, val loss :  0.39, val accuarcy: 89.18%, cost:0:01:28.416508\nstep:  2700, train loss:  0.19, train accuarcy: 95.27%, val loss :  0.35, val accuarcy: 89.66%, cost:0:02:01.925069\nstep:  2800, train loss:  0.16, train accuarcy: 95.95%, val loss :  0.28, val accuarcy: 92.36%, cost:0:02:35.620588\nstep:  2900, train loss:   0.2, train accuarcy: 95.06%, val loss :  0.32, val accuarcy: 91.22%, cost:0:03:09.146150\nstep:  3000, train loss:  0.19, train accuarcy: 95.31%, val loss :  0.29, val accuarcy: 92.38%, cost:0:03:42.562971\nstep:  3100, train loss:  0.22, train accuarcy: 94.97%, val loss :  0.29, val accuarcy: 92.28%, cost:0:04:16.138834\nEpoch: 5\nstep:  3200, train loss:  0.15, train accuarcy: 96.42%, val loss :  0.29, val accuarcy: 92.02%, cost:0:00:26.234004\nstep:  3300, train loss:  0.17, train accuarcy: 95.89%, val loss :  0.35, val accuarcy: 90.48%, cost:0:01:00.304539\nstep:  3400, train loss:  0.15, train accuarcy: 96.25%, val loss :  0.31, val accuarcy: 92.16%, cost:0:01:34.078860\nstep:  3500, train loss:  0.15, train accuarcy: 96.27%, val loss :  0.28, val accuarcy: 92.38%, cost:0:02:07.846796\nstep:  3600, train loss:  0.17, train accuarcy: 95.75%, val loss :  0.36, val accuarcy: 90.72%, cost:0:02:41.827863\nstep:  3700, train loss:  0.15, train accuarcy: 96.27%, val loss :  0.26, val accuarcy: 93.08%, cost:0:03:15.554558\nstep:  3800, train loss:  0.15, train accuarcy: 96.34%, val loss :  0.25, val accuarcy: 93.24%, cost:0:03:49.024171\nstep:  3900, train loss:  0.14, train accuarcy: 96.41%, val loss :  0.29, val accuarcy: 92.30%, cost:0:04:22.491117\neval test data\nloss:  0.23, accuarcy: 93.86%, cost:0:04:37.632397\n```\n\n\n\n#### 3 TextCnn\n\n```shell\npython3.5 -m \"text_classification.main\" -model textcnn -save_path save/textcnn -epoch_num 5\n```\n\n```shell\ncreate model\n\n --------------------\nTextCNN : parms\nFasttext/embedding:0 (5000, 128)\nFasttext/CNN/conv2d/kernel:0 (5, 128, 1, 128)\nFasttext/CNN/conv2d/bias:0 (128,)\nFasttext/dense/kernel:0 (128, 10)\nFasttext/dense/bias:0 (10,)\n-------------------- \n\nload data set\nEpoch: 1\nstep:   100, train loss:   1.8, train accuarcy: 44.06%, val loss :   1.2, val accuarcy: 72.78%, cost:0:00:04.801294\nstep:   200, train loss:  0.68, train accuarcy: 81.02%, val loss :  0.69, val accuarcy: 80.74%, cost:0:00:08.525310\nstep:   300, train loss:  0.48, train accuarcy: 85.89%, val loss :   0.5, val accuarcy: 84.88%, cost:0:00:12.036122\nstep:   400, train loss:  0.36, train accuarcy: 89.66%, val loss :  0.42, val accuarcy: 87.58%, cost:0:00:15.809756\nstep:   500, train loss:  0.31, train accuarcy: 91.27%, val loss :  0.33, val accuarcy: 90.22%, cost:0:00:19.548510\nstep:   600, train loss:  0.26, train accuarcy: 92.39%, val loss :  0.29, val accuarcy: 91.58%, cost:0:00:23.657402\nstep:   700, train loss:  0.27, train accuarcy: 92.31%, val loss :  0.26, val accuarcy: 92.30%, cost:0:00:27.475161\nEpoch: 2\nstep:   800, train loss:  0.18, train accuarcy: 94.97%, val loss :  0.27, val accuarcy: 92.46%, cost:0:00:01.377078\nstep:   900, train loss:  0.18, train accuarcy: 94.73%, val loss :  0.25, val accuarcy: 92.76%, cost:0:00:05.129438\nstep:  1000, train loss:  0.17, train accuarcy: 94.75%, val loss :  0.25, val accuarcy: 92.70%, cost:0:00:08.793668\nstep:  1100, train loss:  0.17, train accuarcy: 95.19%, val loss :  0.25, val accuarcy: 92.12%, cost:0:00:12.762015\nstep:  1200, train loss:  0.16, train accuarcy: 94.98%, val loss :  0.22, val accuarcy: 93.70%, cost:0:00:16.546919\nstep:  1300, train loss:  0.15, train accuarcy: 95.69%, val loss :  0.21, val accuarcy: 93.98%, cost:0:00:20.877066\nstep:  1400, train loss:  0.16, train accuarcy: 95.31%, val loss :   0.2, val accuarcy: 93.88%, cost:0:00:25.291857\nstep:  1500, train loss:  0.15, train accuarcy: 95.41%, val loss :  0.24, val accuarcy: 92.82%, cost:0:00:29.547380\nEpoch: 3\nstep:  1600, train loss:   0.1, train accuarcy: 97.09%, val loss :  0.21, val accuarcy: 93.76%, cost:0:00:01.791284\nstep:  1700, train loss:  0.11, train accuarcy: 96.73%, val loss :  0.23, val accuarcy: 92.98%, cost:0:00:05.452701\nstep:  1800, train loss:  0.11, train accuarcy: 96.42%, val loss :  0.22, val accuarcy: 93.36%, cost:0:00:09.460258\nstep:  1900, train loss:  0.11, train accuarcy: 96.69%, val loss :  0.25, val accuarcy: 92.36%, cost:0:00:13.329682\nstep:  2000, train loss:   0.1, train accuarcy: 96.77%, val loss :  0.21, val accuarcy: 93.84%, cost:0:00:16.989272\nstep:  2100, train loss:  0.11, train accuarcy: 96.55%, val loss :  0.22, val accuarcy: 93.34%, cost:0:00:20.480673\nstep:  2200, train loss:  0.12, train accuarcy: 96.36%, val loss :  0.23, val accuarcy: 92.76%, cost:0:00:24.517457\nstep:  2300, train loss:  0.12, train accuarcy: 96.52%, val loss :  0.21, val accuarcy: 93.44%, cost:0:00:28.338475\nEpoch: 4\nstep:  2400, train loss: 0.076, train accuarcy: 97.86%, val loss :  0.18, val accuarcy: 94.52%, cost:0:00:02.856666\nstep:  2500, train loss: 0.077, train accuarcy: 97.58%, val loss :  0.19, val accuarcy: 94.72%, cost:0:00:07.318107\nstep:  2600, train loss: 0.083, train accuarcy: 97.50%, val loss :   0.2, val accuarcy: 94.26%, cost:0:00:11.614801\nstep:  2700, train loss: 0.087, train accuarcy: 97.28%, val loss :   0.2, val accuarcy: 94.22%, cost:0:00:16.110494\nstep:  2800, train loss: 0.077, train accuarcy: 97.70%, val loss :   0.2, val accuarcy: 94.48%, cost:0:00:20.575370\nstep:  2900, train loss: 0.078, train accuarcy: 97.53%, val loss :  0.26, val accuarcy: 93.08%, cost:0:00:25.102530\nstep:  3000, train loss: 0.083, train accuarcy: 97.31%, val loss :  0.24, val accuarcy: 93.82%, cost:0:00:29.608956\nstep:  3100, train loss: 0.089, train accuarcy: 97.45%, val loss :  0.22, val accuarcy: 94.00%, cost:0:00:34.041156\nEpoch: 5\nstep:  3200, train loss: 0.058, train accuarcy: 98.24%, val loss :  0.22, val accuarcy: 93.78%, cost:0:00:03.429175\nstep:  3300, train loss: 0.057, train accuarcy: 98.20%, val loss :   0.2, val accuarcy: 94.34%, cost:0:00:08.040378\nstep:  3400, train loss: 0.062, train accuarcy: 98.12%, val loss :  0.22, val accuarcy: 93.86%, cost:0:00:12.500552\nstep:  3500, train loss: 0.058, train accuarcy: 98.14%, val loss :   0.2, val accuarcy: 94.74%, cost:0:00:17.083240\nstep:  3600, train loss: 0.062, train accuarcy: 97.98%, val loss :  0.21, val accuarcy: 94.30%, cost:0:00:21.373843\nstep:  3700, train loss: 0.072, train accuarcy: 97.84%, val loss :   0.2, val accuarcy: 94.46%, cost:0:00:25.535575\nstep:  3800, train loss:  0.07, train accuarcy: 97.77%, val loss :  0.21, val accuarcy: 93.92%, cost:0:00:29.945531\nstep:  3900, train loss: 0.058, train accuarcy: 98.23%, val loss :  0.21, val accuarcy: 93.82%, cost:0:00:34.236629\neval test data\nloss:  0.18, accuarcy: 95.07%, cost:0:00:36.260877\n```\n\n#### 4 三个模型的训练结果图如下:\n\n​\t下面为训练结果，其中红色为TextCNN，蓝色为TextRNN，橙色为Fasttext。\n\n![train_result](https://github.com/jieguangzhou/TextClassification/blob/master/images/val_accuarcy.png)\n\n![train_result](https://github.com/jieguangzhou/TextClassification/blob/master/images/val_loss.png)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjieguangzhou%2Ftextclassification","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjieguangzhou%2Ftextclassification","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjieguangzhou%2Ftextclassification/lists"}