[天池比賽] Multimodalities Recall

比賽鏈接及其數據下載：

https://tianchi.aliyun.com/competition/entrance/231786/information

 

base model：（該base是Do大哥一小時搞定的，感謝隊友大哥）

不finetune bert是0.55左右，finetunebert應該是0.60左右。

 

import base64
import argparse

from keras_bert.layers import MaskedConv1D
from sklearn.metrics import roc_auc_score
import pandas as pd
import codecs, gc
from keras_bert import load_trained_model_from_checkpoint, Tokenizer, load_vocabulary
from keras.layers import *
from keras.callbacks import *
from keras.models import Model
from keras.optimizers import Adam
import tensorflow as tf
import keras
import warnings
import numpy as np
import json
import os

warnings.filterwarnings("ignore")

config = tf.ConfigProto()
config.gpu_options.allow_growth = True
keras.backend.set_session(tf.Session(config=config))

np.random.seed(1024)

parser = argparse.Argu