-
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
FIM implementation question
feature request
New feature or request
#932
by liddalidd
was closed May 22, 2023
updated May 22, 2023
Support fp16 scale tolerance
feature request
New feature or request
help wanted
This issue needs assistance
#829
by Quentin-Anthony
was closed May 18, 2023
updated May 20, 2023
Fix output_layer_parallelism = "row"
bug
Something isn't working
help wanted
This issue needs assistance
#916
by Quentin-Anthony
was closed May 19, 2023
updated May 19, 2023
[Question] about summarization tasks
question
#943
by phamkhactu
was closed May 19, 2023
updated May 19, 2023
1.0 HF conversion script fails on Python 3.8
bug
Something isn't working
#875
by syskn
was closed May 18, 2023
updated May 18, 2023
deepspeed.ops.op_builder
bug
Something isn't working
#797
by zscwind
was closed May 18, 2023
updated May 18, 2023
How to convert a model parallel model to hugging face model?
feature request
New feature or request
#880
by guozhiyao
was closed May 18, 2023
updated May 18, 2023
RuntimeError: The expanded size of the tensor (1) must match the existing size (10) at non-singleton dimension 2
bug
Something isn't working
#870
by crazyofapple
was closed Apr 13, 2023
updated May 18, 2023
Import megatron.data.helpers failing
bug
Something isn't working
#934
by convexstrictly
was closed May 12, 2023
updated May 16, 2023
[Broken] Generation with Sequential Model
bug
Something isn't working
help wanted
This issue needs assistance
#854
by satpalsr
was closed May 15, 2023
updated May 15, 2023
Default of Something isn't working
output_layer_parallelism = "row"
is broken for model-parallel training
bug
#905
by cbcase
was closed May 12, 2023
updated May 12, 2023
Is Support FIM mode?
feature request
New feature or request
#930
by zxyscz
was closed May 11, 2023
updated May 11, 2023
Got an empty gpt2-tokenizer while pretraining with THE-PILE dataset
bug
Something isn't working
#876
by LostSpirit1307
was closed May 11, 2023
updated May 11, 2023
Add arguments for turning off fused kernels to work with native pytorch
feature request
New feature or request
#906
by mayank31398
was closed May 11, 2023
updated May 11, 2023
when training with multi node, raise filenotfound error
bug
Something isn't working
#919
by cateto
was closed May 9, 2023
updated May 9, 2023
Make Configs Consistent
bug
Something isn't working
good first issue
Good for newcomers
#920
by StellaAthena
was closed May 9, 2023
updated May 9, 2023
Reconstructing the FP32 weights | HuggingFace conversion
feature request
New feature or request
#922
by davidvblumenthal
was closed May 8, 2023
updated May 8, 2023
pipeline parallelism lead to slower speeds
#903
by cdj0311
was closed May 8, 2023
updated May 8, 2023
Pre-trained Model with Sparse/BlockSparse Attention for Long Sequences and Reduced GPU Memory Consumption
feature request
New feature or request
#924
by puyuanOT
was closed May 5, 2023
updated May 5, 2023
Why GPT2Dataset shuffles documents?
feature request
New feature or request
#923
by kosstbarz
was closed May 5, 2023
updated May 5, 2023
read in/have a additional column in the training data
feature request
New feature or request
#865
by davidvblumenthal
was closed Apr 5, 2023
updated May 4, 2023
Do pythia untied embedding and unembedding matrics?
bug
Something isn't working
#913
by Life-0-1
was closed May 1, 2023
updated May 4, 2023
How to deploy multi-nodes training without hostfile?
feature request
New feature or request
#811
by SefaZeng
was closed Apr 23, 2023
updated May 3, 2023
None of the urls in corpora.py are live
question
#909
by VivekBits2210
was closed May 2, 2023
updated May 2, 2023
ProTip!
Exclude everything labeled
bug
with -label:bug.