search

lc222 / seq2seq_chatbot_new

基于seq2seq模型的简单对话系统的tf实现，具有embedding、attention、beam_search等功能，数据集是Cornell Movie Dialogs
143 stars 64 forks source link

训练时报错 #17

Open heppytt opened 5 years ago

heppytt commented 5 years ago

InvalidArgumentError (see above for traceback): ConcatOp : Dimensions of inputs should match: shape[0] = [128,1024] vs. shape[1] = [640,1024] [[Node: decoder/decoder/while/BasicDecoderStep/decoder/Attention_Wrapper/concat = ConcatV2[N=2, T=DT_FLOAT, Tidx=DT_INT32, _class=["loc:@decod...apeN/f_acc"], _device="/job:localhost/replica:0/task:0/device:CPU:0"](decoder/decoder/while/Switch_12:1, decoder/decoder/while/Switch_8:1, decoder/decoder/while/BasicDecoderStep/decoder/Attention_Wrapper/concat/axis)]]

xuyl0104 commented 5 years ago

我遇到了同样的问题，请问你的问题解决了吗 @heppytt 估计是beam_search扩大维度的时候出错了，我没好到哪里出问题。能否帮我们解答一下？ @lc222 谢谢！