Skip to content

Issues: eric-mitchell/direct-preference-optimization

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

40 Open 43 Closed
40 Open 43 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

When trying to reproduce the complete example, "NotImplementedError: offload_to_cpu=True and NO_SHARD is not supported yet" is thrown
#91 opened Nov 12, 2024 by ZSvedic
1
ValueError when using peft on FSDPTrainer
#90 opened Nov 5, 2024 by AragornHorse
In DPO training, I got this ‘train stats after 160768 examples: {'rewards_train/chosen': 'nan', 'rewards_train/rejected': 'nan', 'rewards_train/accuracies': '0', 'rewards_train/margins': 'nan', 'l -ogps_train/rejected': 'nan', 'logps_train/chosen': 'nan', 'loss/train': 'nan', 'examples_per_second': '5.4876', 'grad_norm': 'nan', 'counters/examples': 160768, 'counters/up -dates': 5024}’
#89 opened Sep 23, 2024 by Alan-D-Chen
2
GPT4 prompt when evaluating DPO
#88 opened Sep 5, 2024 by kygguo
How to gurantee the output.logits.shape[:-1] == labels.shape
#87 opened Aug 13, 2024 by foreverhell
How are evals done on trained models?
#83 opened May 22, 2024 by lesnikow
where is config document of ipo?
#81 opened May 7, 2024 by 3244we
1
Using Mistral 7B with transformers v4.38.1 on MATH dataset, and facing memory leaks
#80 opened May 4, 2024 by Jayant1234
Weird logits and model starts degeneration while training DPO
#77 opened Apr 9, 2024 by DungNasSa10
2
Was it your intention to recreate wandb tables in iterator?
#76 opened Apr 4, 2024 by huskydoge
Can DPO work on BERT-style Model?
#75 opened Mar 24, 2024 by Leo-T-Zang
The number of training steps in the SHP dataset
#73 opened Mar 16, 2024 by bonin147
Computing faster lopgs
#72 opened Mar 9, 2024 by alexvishnevskiy
3
Implementation for Plackett-Luce rank model
#71 opened Mar 4, 2024 by rohan598
1
What's the reference policy of Preferred-FT in Figure 2?
#70 opened Mar 4, 2024 by zetian1025
My Code to Reproduce IMDB
#69 opened Feb 26, 2024 by QiyaoWei
Why does SFT sum the cross-entropy loss within each sequence?
#68 opened Feb 17, 2024 by YJWon99
1
Using cross entropy loss to calculate DPO?
#67 opened Feb 14, 2024 by zachares
2
Unable to Run SFT
#66 opened Feb 13, 2024 by Rui-Yuan91
3
Question bout IPO loss vs DPO loss
#64 opened Jan 30, 2024 by MoonBlvd
1
Reproducing Win Rate inference for TL;DR
#62 opened Jan 9, 2024 by jdchang1
1
DPO did not achieve the expected experimental effect
#56 opened Dec 7, 2023 by Vance0124
2
How to re-implement the result of IMDB sentiment generation.
#54 opened Nov 14, 2023 by junkangwu
Llama-2-13b-chat Valid reward accuracy remains ~50%
#53 opened Nov 6, 2023 by nxphi47
Qwen model issues & embedding and loss has nan
#52 opened Nov 3, 2023 by lylcst
5
ProTip! Follow long discussions with comments:>50.