Skip to content
This repository has been archived by the owner on Jul 7, 2023. It is now read-only.

Issues: tensorflow/tensor2tensor

575 Open 672 Closed
575 Open 672 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

ValueError: Cannot use 'while/transformer/problem_choice' as input to 'Equal' because 'while/transformer/problem_choice' is in a while loop. See info log for more details. bug
#613 opened Feb 23, 2018 by dixiematt8
1
Non deterministic Decoding. How can I do in Transformer. question
#607 opened Feb 21, 2018 by dixiematt8
2
Multiple beam results slowdown: "return_beams=True" increases decode time feature request
#601 opened Feb 19, 2018 by ndvbd
18
DEADLINE EXCEEDED error [query.py] feature request
#599 opened Feb 19, 2018 by ndvbd
1
Trying to replicate walk through code using Transformer library but unable to finish it. question
#591 opened Feb 16, 2018 by dixiematt8
1
Continue training with new data question
#588 opened Feb 15, 2018 by tmkhalil
6
How to use SNLI dataset with transformer. question
#583 opened Feb 13, 2018 by dixiematt8
Setting max_length low makes BLEU unexpectedly worse question
#582 opened Feb 13, 2018 by martinpopel
4
OOM Error after long training bug
#581 opened Feb 13, 2018 by martinpopel
13
Decode issue at inference bug
#579 opened Feb 13, 2018 by rafaelbou
4
Unable to decode a pre-processed input with sub-word segmentation question
#578 opened Feb 12, 2018 by surafelml
4
ERROR:tensorflow:Model diverged with loss = NaN. question
#575 opened Feb 11, 2018 by yudianer
there may have some problem about tokenizer.py bug
#567 opened Feb 8, 2018 by drockser
2
decoding bug: LanguagemodelWikiNorefV8kL1k bug
#561 opened Feb 6, 2018 by rafaelbou
1
continuous_train_and_eval makes BLEU worse (saving two checkpoints instead of one) bug
#556 opened Feb 5, 2018 by martinpopel
9
t2t & simple timeseries? Any example available? question
#544 opened Jan 27, 2018 by TanCari
3
Bug introduced in v1.3.0 causing training divergence bug
#529 opened Jan 19, 2018 by martinpopel
27
T2T 1.4.1 transformer beam search result different with 1.3.2 bug
#525 opened Jan 18, 2018 by efeiefei
6
Issue with EnFr task. Maybe a tokenization problem? bug
#518 opened Jan 15, 2018 by gsoul
11
Wrong/unclear exp and sqrt learning_rate_decay_scheme bug
#517 opened Jan 13, 2018 by martinpopel
5
eval autoregressive error bug
#511 opened Jan 11, 2018 by YichenGong
1
Using this model for reading text in images question
#509 opened Jan 10, 2018 by MBleeker
_fast_decode for multiple datashards feature request
#502 opened Jan 5, 2018 by apeterswu
2
Adding the option of Weighted Transformer Network for improvement feature request
#480 opened Dec 21, 2017 by AranKomat
2
logprob scores when decoding from file, with return_beams=True feature request
#459 opened Dec 4, 2017 by thompsonb
3
ProTip! Mix and match filters to narrow down what you’re looking for.