-
Notifications
You must be signed in to change notification settings - Fork 976
Issues: EleutherAI/gpt-neox
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
Issues installing - Bare with me please
bug
Something isn't working
#1006
by MistakingManx
was closed Jul 28, 2023
updated Jul 28, 2023
BF16 Trainning's bug
bug
Something isn't working
#991
by jiezhangGt
was closed Jul 6, 2023
updated Jul 26, 2023
train shows an unexpected exception in best_download no explanation given
bug
Something isn't working
#976
by nevakrien
was closed Jul 24, 2023
updated Jul 24, 2023
The xformers result can not match with norm attention result
feature request
New feature or request
#998
by guozhiyao
was closed Jul 20, 2023
updated Jul 20, 2023
Substantial decrease in FLOPs per GPU when training multinode
#965
by davidvblumenthal
was closed Jul 11, 2023
updated Jul 11, 2023
Distributed training with model parallelism hangs with the recent PR
bug
Something isn't working
#985
by absol13
was closed Jul 10, 2023
updated Jul 10, 2023
A clearer explanation of data-weights with more details
feature request
New feature or request
#992
by leocnj
was closed Jul 8, 2023
updated Jul 8, 2023
NCCL backend in DeepSpeed not yet implemented
bug
Something isn't working
#990
by jiezhangGt
was closed Jul 5, 2023
updated Jul 5, 2023
About gpt-neox-20B model hyperparameter
#989
by peiyingxin
was closed Jul 5, 2023
updated Jul 5, 2023
input decoding error on the args string of train.py
bug
Something isn't working
#977
by nevakrien
was closed Jun 27, 2023
updated Jun 27, 2023
Any plans to implement multi-query attention
feature request
New feature or request
#982
by crazyofapple
was closed Jun 27, 2023
updated Jun 27, 2023
Questions: does norm support sequence parallel
#980
by ftgreat
was closed Jun 25, 2023
updated Jun 25, 2023
WARNING: shuffle index length is not equal to sample index length
bug
Something isn't working
#972
by 1ittlesnow
was closed Jun 22, 2023
updated Jun 22, 2023
[Training followed tutorial] error: exits with return code = -7
bug
Something isn't working
#940
by phamkhactu
was closed May 18, 2023
updated Jun 20, 2023
Adding data to continue training failed.
feature request
New feature or request
#860
by SefaZeng
was closed May 18, 2023
updated Jun 10, 2023
Any plans on supporting modeling_tf_gpt_neox to hugging face transformers models
feature request
New feature or request
#970
by praneethgb
was closed Jun 7, 2023
updated Jun 7, 2023
RuntimeError: stack expects each tensor to be equal size
bug
Something isn't working
#929
by cateto
was closed Jun 5, 2023
updated Jun 5, 2023
How to run gpt-neox with two gtx 1080's?
#964
by therealjr
was closed Jun 4, 2023
updated Jun 4, 2023
change max-position-embeddings after pretrain.
feature request
New feature or request
#942
by guozhiyao
was closed Jun 3, 2023
updated Jun 3, 2023
Create script to allow conversion from HF ckpt to neox
feature request
New feature or request
#846
by Quentin-Anthony
was closed Jun 2, 2023
updated Jun 2, 2023
3 tasks
Running with bf16 error
bug
Something isn't working
good first issue
Good for newcomers
#939
by Life-0-1
was closed May 18, 2023
updated May 25, 2023
Preconfigured Datasets are not available
bug
Something isn't working
#949
by vogt31337
was closed May 24, 2023
updated May 24, 2023
[Question] about clearly params
feature request
New feature or request
#945
by phamkhactu
was closed May 24, 2023
updated May 24, 2023
Configs are not broadcasted properly in multi-node training
bug
Something isn't working
#925
by silverriver
was closed May 23, 2023
updated May 23, 2023
ProTip!
Follow long discussions with comments:>50.