i want to train the encoder-decoder model on two gpu , and i have tried to understand https://github.com/harvardnlp/seq2seq-attn, but i do not very know it clearly, i always make mistakes, so if who knows the detailed logic relation, i want to discuss with you, thank you, i need your help
i want to train the encoder-decoder model on two gpu , and i have tried to understand https://github.com/harvardnlp/seq2seq-attn, but i do not very know it clearly, i always make mistakes, so if who knows the detailed logic relation, i want to discuss with you, thank you, i need your help