pageedward

@Tracin @loralyc 请问你用text_file训练是从已有checkpoint继续训练吗，会不会出现下面问题：  需要排除部分节点？

> > @AaronYALai @interxuxing I think it should the next parents should be gather_helper(input_t.parents, parents) as the parent_idx is the traceback to the last timestep's parent at each timestep. >...

> > @AaronYALai @interxuxing I think it should the next parents should be gather_helper(input_t.parents, parents) as the parent_idx is the traceback to the last timestep's parent at each timestep. >...

@ritheshkumar95

@phhang

@wangguanhua 这里loss 只计算了1个 batch的loss 吧，

@xinxueying what is the val loss when u ended training?

@zhongwenkun886 @ypwhs Keras版本里的已经训练好的权重文件的百度网盘链接已经失效，能否共享下呢？谢谢

@mhy1998 @ypwhs 有没有预训练权重文件，共享下行不