Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / bojone/attention issues and pull requests

#20 - Keras的batch_dot()如何用pytorch改写

Issue - State: open - Opened by linjingxu over 4 years ago - 2 comments

#19 - a = K.permute_dimensions(a, (0, 3, 2, 1))

Issue - State: open - Opened by monkeyshichi over 4 years ago - 2 comments

#18 - Is there tensorflow implementation of sparse attention?

Issue - State: open - Opened by hugddygff about 5 years ago - 3 comments

#17 - I have no idea what method "to_mask" helps.

Issue - State: open - Opened by fatLime about 5 years ago - 1 comment

#16 - 模型加载问题

Issue - State: open - Opened by xiong3134 about 5 years ago - 2 comments

#15 - K，Q，V映射

Issue - State: open - Opened by Vipning over 5 years ago

#14 - 请问Q,K,V分别是从哪里输入的呢？

Issue - State: open - Opened by Ironeie over 5 years ago - 2 comments

#13 - attention_keras.py

Issue - State: closed - Opened by myndtt over 5 years ago

#11 - tf中position

Issue - State: open - Opened by lrx1213 almost 6 years ago - 4 comments

#11 - tf中position

Issue - State: open - Opened by lrx1213 almost 6 years ago - 4 comments

#10 - attention_keras.py中，Position_Embedding的pos的计算

Issue - State: open - Opened by rivershadowy about 6 years ago - 2 comments

#9 - 行为识别　用attention

Issue - State: open - Opened by guofuzheng about 6 years ago - 1 comment

#9 - 行为识别　用attention

Issue - State: open - Opened by guofuzheng about 6 years ago - 1 comment

#8 - 感觉好像少了参数Wo

Issue - State: closed - Opened by duanyu about 6 years ago - 1 comment

#8 - 感觉好像少了参数Wo

Issue - State: closed - Opened by duanyu about 6 years ago - 1 comment

#7 - 不支持mask

Issue - State: closed - Opened by Zhao6924 over 6 years ago - 2 comments

#6 - How to understand the decoder input(target)?

Issue - State: closed - Opened by zoe218 over 6 years ago - 1 comment

#6 - How to understand the decoder input(target)?

Issue - State: closed - Opened by zoe218 over 6 years ago - 1 comment

#5 - position_ij = tf.concat([tf.cos(position_ij), tf.sin(position_ij)], 1)

Issue - State: closed - Opened by imhuay over 6 years ago - 1 comment

#5 - position_ij = tf.concat([tf.cos(position_ij), tf.sin(position_ij)], 1)

Issue - State: closed - Opened by imhuay over 6 years ago - 1 comment

#4 - position_ij = tf.matmul(position_i, position_j) 应该用 multiply 而不是 matmul

Issue - State: closed - Opened by imhuay over 6 years ago - 1 comment

#4 - position_ij = tf.matmul(position_i, position_j) 应该用 multiply 而不是 matmul

Issue - State: closed - Opened by imhuay over 6 years ago - 1 comment

#3 - K.permute_dimensions(A, (0,3,2,1))运行时报错

Issue - State: closed - Opened by njuccq almost 7 years ago - 13 comments

#2 - Position_Embedding肯能有些问题

Issue - State: closed - Opened by liangoy almost 7 years ago - 1 comment

#2 - Position_Embedding肯能有些问题

Issue - State: closed - Opened by liangoy almost 7 years ago - 1 comment

#1 - 请问 Position Embedding 可以被训练吗？

Issue - State: closed - Opened by ZJUguquan almost 7 years ago - 3 comments

#1 - 请问 Position Embedding 可以被训练吗？

Issue - State: closed - Opened by ZJUguquan almost 7 years ago - 3 comments