mlitxr / test-classification-cnn-pytorch Goto Github PK

torch                        1.13.1
gensim                       4.2.0

python main.py -h

usage: main.py [-h] [-lr LR] [-epoch EPOCH] [-kernel_nums KERNEL_NUMS]
               [-divident DIVIDENT] [-padding_size PADDING_SIZE]
               [-test_data_path TEST_DATA_PATH] [-word2vec_size WORD2VEC_SIZE]
               [-dropout DROPOUT]

CNN model training

optional arguments:
  -h, --help            show this help message and exit
  -lr LR                learning_rate
  -epoch EPOCH          training eochs
  -kernel_nums KERNEL_NUMS
                        number of each kind of kernel
  -divident DIVIDENT    ratio of the training set to the test set
  -padding_size PADDING_SIZE
                        number of words picked in each sentence
  -test_data_path TEST_DATA_PATH
                        address of data used to train
  -word2vec_size WORD2VEC_SIZE
                        word2vec dim
  -dropout DROPOUT      dropout

python main.py
tensorboard --logdir='train_log'

python pridict.py
#需要输入上面选择的序号
#输出按照submit_example.txt的格式存储在主目录下的submit.txt中

id, pred
0, 0
1, 1
2, 4
3, 1
4, 6
5, 7
6, 5
7, 4
8, 9
9, 7
10, 6