Skip to content

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.

Sign up for GitHub

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

Clear current search query, filters, and sorts
42 Open 286 Closed
42 Open 286 Closed
Author
Filter by author
Label
Filter by label
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Milestones
Filter by milestone
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

Cannot convert neox model to HF bug Something isn't working
#1231 opened May 28, 2024 by srivassid
1
How to set the ffn hidden size parameter in gpt neox feature request New feature or request
#1230 opened May 28, 2024 by IronMan-WangJinxi
The results of running eval show only 1 digit after decimal point for acc on all tested tasks bug Something isn't working
#1227 opened May 22, 2024 by lernerjenny
2
My servers used for multi-node training do not have ssh. How can I launch multi-node training using the torchrun command? feature request New feature or request
#1203 opened Apr 23, 2024 by dingning97
1
PyTorch Lightning Fused optimizer step feature request New feature or request
#1160 opened Feb 29, 2024 by jahatef
Tests fail when run with pytest --forked bug Something isn't working
#1132 opened Jan 25, 2024 by segyges
1
Integrate TransformerEngine feature request New feature or request
#1098 opened Dec 21, 2023 by Quentin-Anthony
6
Interoperability and GPT-NeoX documentation Improvements or additions to documentation question
#1058 opened Oct 12, 2023 by StellaAthena
2
Support for Mosaic Models feature request New feature or request
#1057 opened Oct 6, 2023 by rajveer43
1
[BUG] Inconsistent loss between overlap_comm=true and overlap_comm=false bug Something isn't working
#1004 opened Jul 27, 2023 by 0x6b64
4
Convert HF Llama Checkpoints to Neox Checkpoints feature request New feature or request
#994 opened Jul 10, 2023 by sxthunder
1
AssertionError: zero stage 1 requires an optimizer bug Something isn't working good first issue Good for newcomers help wanted This issue needs assistance
#987 opened Jul 4, 2023 by yonglianglan
3
How to preserve Pythia's sampling order but for different batch size. bug Something isn't working
#984 opened Jul 3, 2023 by lintangsutawika
3
Why we need to average LayerNorm values over mp ranks when converting to HFformat checkpoint?
#983 opened Jun 26, 2023 by forceshorty
Bias weights are multi-added when using gpt_j_residual in model-parallel execution bug Something isn't working good first issue Good for newcomers
#962 opened May 31, 2023 by cbcase
1
Can't finetune 20B model from slim weights with zero optimizer enabled bug Something isn't working
#926 opened May 5, 2023 by coreystatendet
2
Problems on generating with llama model
#921 opened May 4, 2023 by wiio12
21
Fine-tuning gpt-neox on 8 A100s feature request New feature or request
#892 opened Apr 20, 2023 by rajhans
6
OOM error when training on a 220G Memory machine with 8 V100. feature request New feature or request
#867 opened Apr 2, 2023 by SefaZeng
2
Add support for pytorch 2.0 ? deprioritized Issues that are not closed, but are low priority and unlikely to be solved soon feature request New feature or request
#858 opened Mar 27, 2023 by guozhiyao
1
Finetuning loss explode when not loading deepspeed zero optimal states bug Something isn't working
#843 opened Mar 19, 2023 by sxthunder
9
Implement Prefix-LM attention masking feature request New feature or request
#805 opened Mar 1, 2023 by TokyoExpress
1
Cannot load the checkpoint bug Something isn't working
#782 opened Feb 6, 2023 by jmlongriver12
5
Unable to load model checkpoint with model parallelism feature request New feature or request
#773 opened Jan 20, 2023 by RaoNikitha
4
Multi-node training without shared memory deprioritized Issues that are not closed, but are low priority and unlikely to be solved soon feature request New feature or request
#765 opened Jan 6, 2023 by VHellendoorn
ProTip! Updated in the last three days: updated:>2024-06-16.