LWISLVISLIE/keras-transformer issues and pull requests

#44 - Code example of a transformer model based on character level input sequence and word level output sequence

Issue - State: closed - Opened by AyberkYavuz over 1 year ago - 1 comment
Labels: question

#43 - what the required version of tensorflow and keras for this project?

Issue - State: closed - Opened by JulseJiang almost 3 years ago
Labels: enhancement

#42 - Update test_get_model.py

Pull Request - State: closed - Opened by shenganzhang about 3 years ago

#41 - Installation problem on MAC m1

Issue - State: closed - Opened by dauny90 about 3 years ago - 1 comment
Labels: question, wontfix

#40 - Update gelu.py

Pull Request - State: closed - Opened by BenjaminWegener about 3 years ago - 3 comments
Labels: wontfix

#39 - ValueError: An operation has `None` for gradient. Please make sure that all of your ops have a gradient defined (i.e. are differentiable). Common ops without gradient: K.argmax, K.round, K.eval.

Issue - State: closed - Opened by littlecay over 3 years ago - 1 comment
Labels: bug, wontfix

#38 - Easiest way to output attention scores

Issue - State: closed - Opened by rajicon over 3 years ago - 1 comment
Labels: question, wontfix

#37 - Question concerning default settings

Issue - State: closed - Opened by nosnibor27 almost 4 years ago - 1 comment
Labels: question, wontfix

#36 - How can I adjust model to disable teacher forcing?

Issue - State: closed - Opened by lismz almost 4 years ago - 4 comments
Labels: question, wontfix

#35 - Slow prediction

Issue - State: closed - Opened by piacenti over 4 years ago - 1 comment
Labels: question, wontfix

#34 - why get the same output with different input?

Issue - State: closed - Opened by jayeew over 4 years ago - 1 comment
Labels: question

#33 - The input of function "_get_max_suffix_repeat_times" is not correct.

Issue - State: closed - Opened by sdq11111 over 4 years ago
Labels: bug

#32 - Hierarchy of Encoders Not Working

Issue - State: closed - Opened by rajicon over 4 years ago - 1 comment
Labels: question, wontfix

#31 - How do you use masking with the transformer?

Issue - State: closed - Opened by rajicon over 4 years ago - 2 comments
Labels: question, wontfix

#30 - doesnt work when test set contains more than two sentences

Issue - State: closed - Opened by sainikmahata over 4 years ago - 2 comments
Labels: question, wontfix

#29 - How to handle multiple inputs

Issue - State: closed - Opened by rajicon almost 5 years ago - 1 comment
Labels: question, wontfix

#28 - Some question about the decode method

Issue - State: closed - Opened by qiuhuiGithub almost 5 years ago - 1 comment
Labels: question, wontfix

#27 - Fix compatibility

Pull Request - State: closed - Opened by LWISLVISLIE almost 5 years ago

#26 - module 'keras.layers' has no attribute 'Wrapper'

Issue - State: closed - Opened by Nadni almost 5 years ago - 1 comment
Labels: bug

#25 - Questions about data input and output

Issue - State: closed - Opened by zhouhao-learning almost 5 years ago - 1 comment
Labels: question, wontfix

#24 - Do I need to specify <start> and <end> of sequences

Issue - State: closed - Opened by LincLabUCCS about 5 years ago - 1 comment
Labels: wontfix

#23 - why get same output with different input?

Issue - State: closed - Opened by maozezhong about 5 years ago - 1 comment
Labels: question, wontfix

#22 - Using get_encoders in my model, output is all zeros

Issue - State: closed - Opened by UriLawgeex over 5 years ago - 3 comments
Labels: question, wontfix

#21 - How to train a transformer model with fit_generator?

Issue - State: closed - Opened by CrazyGuitarAndCode over 5 years ago - 2 comments
Labels: wontfix

#20 - star-transformer

Issue - State: closed - Opened by yongzhuo over 5 years ago - 2 comments
Labels: enhancement, wontfix

#19 - Add adapter

Pull Request - State: closed - Opened by LWISLVISLIE over 5 years ago

#18 - get metrics accuracy in compile

Issue - State: closed - Opened by kewin1807 over 5 years ago - 1 comment
Labels: bug, wontfix

#17 - experiments on Real world examples

Issue - State: closed - Opened by kaeflint over 5 years ago - 1 comment
Labels: wontfix

#16 - #15 Add beam search

Pull Request - State: closed - Opened by LWISLVISLIE over 5 years ago

#15 - beam search

Issue - State: closed - Opened by PKQ1688 over 5 years ago - 2 comments
Labels: enhancement, wontfix

#14 - Describe the Bug

Issue - State: closed - Opened by PKQ1688 over 5 years ago - 1 comment

#13 - Error in get_model() : ValueError: too many values to unpack (expected 2)

Issue - State: closed - Opened by nshaud over 5 years ago - 8 comments
Labels: bug, wontfix

#12 - Possible bug: Encoder fed every step

Issue - State: closed - Opened by veqtor over 5 years ago - 1 comment
Labels: bug, wontfix

#11 - Is there only Multi-Head Attention?

Issue - State: closed - Opened by kaen2891 over 5 years ago - 5 comments
Labels: enhancement, wontfix

#10 - Support tf.keras

Pull Request - State: closed - Opened by LWISLVISLIE over 5 years ago

#9 - Use it as a layer

Issue - State: closed - Opened by guyko81 over 5 years ago - 4 comments
Labels: enhancement, wontfix

#8 - Add training argument to return input layers and output layer

Pull Request - State: closed - Opened by psirenny over 5 years ago - 3 comments

#7 - ValueError: Unknown layer: EmbeddingRet

Issue - State: closed - Opened by Nadni over 5 years ago - 1 comment
Labels: bug

#6 - Tensorflow dependency remove

Issue - State: closed - Opened by flydsc over 5 years ago - 2 comments
Labels: enhancement, wontfix

#5 - after stacked decoder layer，I think it should be concatenate with a linear and softmax layer.

Issue - State: closed - Opened by flyboyer almost 6 years ago

#4 - Added note about "incompatible shapes" error when accuracy metric is …

Pull Request - State: closed - Opened by bedapudi6788 almost 6 years ago - 1 comment

GitHub / LWISLVISLIE/keras-transformer issues and pull requests