self.loss_rewards = seq2seq.sequence_loss(

logits=decoder_logits_train,

targets=self.decoder_inputs,

weights=self.masks_rewards,

average_across_timesteps=True,

average_across_batch=True)

self.loss = seq2seq.sequence_loss(

logits=decoder_logits_train,

targets=self.decoder_inputs,

weights=self.masks,

average_across_timesteps=True,

average_across_batch=True)

这两个函数有什么区别？

为什么mask要乘reward？

计算loss_reward有什么用处？