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

GitHub / lucidrains/memorizing-transformers-pytorch issues and pull requests

#16 - stable_softmax

Issue - State: closed - Opened by ontocord about 2 years ago - 3 comments

#15 - is there any environment detial for use this package ?

Issue - State: open - Opened by sertg5k4ll about 2 years ago

#14 - is it a t5 arch or decoder only gpt style arch?

Issue - State: open - Opened by brando90 over 2 years ago - 1 comment

#13 - hugging face training code with demo

Issue - State: open - Opened by brando90 over 2 years ago

#12 - official repo?

Issue - State: open - Opened by brando90 over 2 years ago - 1 comment

#11 - Update KNNMemory.add to correctly index and update KNN list, after ad…

Pull Request - State: closed - Opened by LakshyAAAgrawal over 2 years ago - 1 comment

#10 - KNNMemory add() does not appear to update self.knns

Issue - State: closed - Opened by vyaivo over 2 years ago

#9 - FAISS hard reset

Issue - State: open - Opened by itsdaniele almost 3 years ago

#8 - Update train.py to correct implementation of val loss calculation

Pull Request - State: closed - Opened by LakshyAAAgrawal almost 3 years ago

#7 - index out of

Issue - State: open - Opened by chxiag almost 3 years ago

#6 - Support for Multi-GPU training?

Issue - State: open - Opened by Victorwz almost 3 years ago

#5 - Dimensionality of key and values for Attention

Issue - State: open - Opened by manestay about 3 years ago - 8 comments

#4 - Arguments to reproduce the models from the original paper?

Issue - State: closed - Opened by manestay about 3 years ago - 1 comment

#3 - Maybe scale is wrong

Issue - State: open - Opened by denadai2 about 3 years ago - 3 comments

#2 - try using joblib to parallelize adding new memories to faiss indices

Pull Request - State: closed - Opened by lucidrains over 3 years ago

#1 - Any interesting results?

Issue - State: open - Opened by rom1504 over 3 years ago - 79 comments