Skip to content

Issues: uclaml/SPIN

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

20 Open 11 Closed
20 Open 11 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

Question about potential overfitting
#38 opened Oct 25, 2024 by kang-0909
Hello, I would like to ask, when you are training the model, do you only use the first round of dialogue from the ultrachat_200k?
#37 opened Jul 21, 2024 by jackwwy
1
Cannot reproduce the result
#36 opened Jul 12, 2024 by Joyyang158
4
numpy version warnings
#35 opened Jul 9, 2024 by jiangjun0105
1
OOM with 8 A800
#34 opened Jul 8, 2024 by 647sherry
8
Difference between training / generation input format.
#33 opened May 30, 2024 by hbin0701
Cannot reproduce generated samples in UCLA-AGI/SPIN_iter0
#32 opened May 22, 2024 by StarDewXXX
3
Generate Result
#31 opened Apr 30, 2024 by lss11005
Theoretical Analysis and Idea of SPIN are quite weird (may not make senses)??
#30 opened Apr 20, 2024 by AGTSAAA
Question about the checkpoint provided in this repo
#28 opened Mar 20, 2024 by StarDewXXX
3
SPIN == DPO in self-iteration?
#26 opened Mar 16, 2024 by onebula
6
Confused about iterations
#25 opened Mar 13, 2024 by junkangwu
4
Significant Performance Drop in GSM8k Evaluation with Updated SFT ckpt
#24 opened Mar 12, 2024 by yinyueqin
3
Have you tried combination of SPIN, SFT, DPO
#23 opened Mar 5, 2024 by penolove
Thesis discussion: Why can the end-to-end algorithm work properly?
#22 opened Mar 1, 2024 by nomadlx
5
GPU Memory question
#21 opened Feb 29, 2024 by fangyuan-ksgk
1
What changes should I make to apply the SPIN method on Llama2?
#20 opened Feb 29, 2024 by Labmem009
1
Potential reason of the significant improvement on the TruthfulQA and GSM8k
#19 opened Feb 28, 2024 by NirViaje
the logps decrease
#13 opened Feb 22, 2024 by zhanghaoie
2
Unable to reproduce performance
#12 opened Feb 22, 2024 by guozhiyao
10
ProTip! Find all open issues with in progress development work with linked:pr.