张量流中 embedding_rnn_seq2seq 模型中的 output_projection 参数是什么？
What is output_projection argument in embedding_rnn_seq2seq model in tensorflow?
正如我所读 API 它说我们可以保留上面的参数 None 或设置输出层的权重参数。这东西在做什么？
https://www.tensorflow.org/api_docs/python/tf/contrib/legacy_seq2seq/embedding_rnn_seq2seq
seq2seq tutorial. The short explanation is that it allows an efficient approximation of the loss when there are huge numbers of classes (via sampled_softmax_loss)中有部分解释。
正如我所读 API 它说我们可以保留上面的参数 None 或设置输出层的权重参数。这东西在做什么？
https://www.tensorflow.org/api_docs/python/tf/contrib/legacy_seq2seq/embedding_rnn_seq2seq
seq2seq tutorial. The short explanation is that it allows an efficient approximation of the loss when there are huge numbers of classes (via sampled_softmax_loss)中有部分解释。