test.py

import argparse
import os
from importlib import import_module

import gradio as gr
from tqdm import tqdm
import models.TextCNN
import torch
import pickle as pkl
from utils import  build_dataset
classes=['finance','realty','stocks','education','science','society','politics','sports','game','entertainment']

MAX_VOCAB_SIZE = 10000  # 词表长度限制
UNK, PAD = '<UNK>', '<PAD>'  # 未知字，padding符号
def build_vocab(file_path, tokenizer, max_size, min_freq):
    vocab_dic = {}
    with open(file_path, 'r', encoding='UTF-8') as f:
        for line in tqdm(f):
            lin = line.strip()
            if not lin:
                continue
            content = lin.split('\t')[0]
            for word in tokenizer(content):
                vocab_dic[word] = vocab_dic.get(word, 0) + 1
        vocab_list = sorted([_ for _ in vocab_dic.items() if _[1] >= min_freq], key=lambda x: x[1], reverse=True)[:max_size]
        vocab_dic = {word_count[0]: idx for idx, word_count in enumerate(vocab_list)}
        vocab_dic.update({UNK: len(vocab_dic), PAD: len(vocab_dic) + 1})
    return vocab_dic

parser = argparse.ArgumentParser(description='Chinese Text Classification')
parser.add_argument('--word', default=False, type=bool, help='True for word, False for char')
args = parser.parse_args()
model_name='TextCNN'
dataset = 'THUCNews'  # 数据集
embedding = 'embedding_SougouNews.npz'
x = import_module('models.' + model_name)

config = x.Config(dataset, embedding)
device='cuda:0'
model=models.TextCNN.Model(config)

# vocab, train_data, dev_data, test_data = build_dataset(config, args.word)
model.load_state_dict(torch.load('THUCNews/saved_dict/TextCNN.ckpt'))
model.to(device)
model.eval()


tokenizer = lambda x: [y for y in x]  # char-level
if os.path.exists(config.vocab_path):
    vocab = pkl.load(open(config.vocab_path, 'rb'))
else:
    vocab = build_vocab(config.train_path, tokenizer=tokenizer, max_size=MAX_VOCAB_SIZE, min_freq=1)
    pkl.dump(vocab, open(config.vocab_path, 'wb'))
print(f"Vocab size: {len(vocab)}")


# content='时评：“国学小天才”录取缘何少佳话'
content=input('输入语句:')

words_line = []
token = tokenizer(content)
seq_len = len(token)
pad_size=32
contents=[]

if pad_size:
    if len(token) < pad_size:
        token.extend([PAD] * (pad_size - len(token)))
    else:
        token = token[:pad_size]
        seq_len = pad_size
# word to id
for word in token:
    words_line.append(vocab.get(word, vocab.get(UNK)))

contents.append((words_line, seq_len))
print(words_line)
# input = torch.LongTensor(words_line).unsqueeze(1).to(device)  # convert words_line to LongTensor and add batch dimension
x = torch.LongTensor([_[0] for _ in contents]).to(device)

        # pad前的长度(超过pad_size的设为pad_size)
seq_len = torch.LongTensor([_[1] for _ in contents]).to(device)
input=(x,seq_len)
print(input)
with torch.no_grad():
    output = model(input)
    predic = torch.max(output.data, 1)[1].cpu().numpy()
print(predic)
print('类别为：{}'.format(classes[predic[0]]))





# with torch.no_grad():
#     output=model(input)
# print(output)

#
# start_time = time.time()
# test_iter = build_iterator(test_data, config)
# with torch.no_grad():
#     predict_all = np.array([], dtype=int)
#     labels_all = np.array([], dtype=int)
#     for texts, labels in test_iter:
#         # texts=texts.to(device)
#         print(texts)
#         outputs = model(texts)
#         loss = F.cross_entropy(outputs, labels)
#         labels = labels.data.cpu().numpy()
#         predic = torch.max(outputs.data, 1)[1].cpu().numpy()
#         labels_all = np.append(labels_all, labels)
#         predict_all = np.append(predict_all, predic)
#         break
#     print(labels_all)
#     print(predict_all)
#
#