终于用上了bert,踩了一些坑,和你们分享一下。函数
我主要参考了奇点机智的文章,用bert作了两个中文任务:文本分类和类似度计算。这两个任务都是直接用封装好的run_classifer,py,另外两个没有仔细看,用到了再补充。lua
Step1:写好本身的processor,照着例子写就能够,必定要shuffle!!!spa
Step2:加到main函数的processors字典里code
Step1:建一个hookget
early_stopping_hook = tf.contrib.estimator.stop_if_no_decrease_hook(
estimator=estimator,
metric_name='eval_loss',
max_steps_without_decrease=FLAGS.max_steps_without_decrease,
eval_dir=None,
min_steps=0,
run_every_secs=None,
run_every_steps=FLAGS.save_checkpoints_steps)复制代码
Step2:加到estimator.train里input
estimator.train(input_fn=train_input_fn, max_steps=num_train_steps, hooks=[early_stopping_hook])复制代码
须要用tensorboard查看训练曲线的话比较好it
Step1:建立train和eval的spec,这里须要把early stopping的hook加到trainSpecclass
train_spec = tf.estimator.TrainSpec(input_fn=train_input_fn, max_steps=num_train_steps,
hooks=[early_stopping_hook])
eval_spec = tf.estimator.EvalSpec(input_fn=eval_input_fn, throttle_secs=0)
tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)复制代码
默认Eval和Predict的batch size都很小,记得改一下sso
<-未完待续->im