Skip to content

Issues: allenai/OLMo

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

52 Open 161 Closed
52 Open 161 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

High CrossEntropy and Z Loss variance after loading from checkpoint type/bug An issue about a bug
#776 opened Jan 6, 2025 by abhijangda
Generating training mix of OLMo2 from dolmino-mix type/question An issue that's a question
#775 opened Jan 5, 2025 by Cy-47
1
OLMo2 checkpoints for continued pretraining (non-HF) type/question An issue that's a question
#773 opened Dec 30, 2024 by SimonSuster
2
Sudden data error during training type/bug An issue about a bug
#766 opened Dec 16, 2024 by faresobeid
11
tokenizer.encode function`s param add_special_tokens=False not work. type/bug An issue about a bug
#765 opened Dec 12, 2024 by xiaohan2909
Difference Between DDP and FSDP Modes type/question An issue that's a question
#762 opened Dec 6, 2024 by lllabmaster
About eos_token_id in config file (20M, 1B) type/question An issue that's a question
#757 opened Nov 29, 2024 by lllabmaster
Fail to load tokenizer for checkpoints type/bug An issue about a bug
#741 opened Oct 24, 2024 by tresiwald
Error Encountered During Multi-Node Pretraining with Torchrun type/bug An issue about a bug
#737 opened Oct 21, 2024 by Zehui127
Missing OLMo checkpoints
#726 opened Oct 3, 2024 by mirandrom
1
8-bit allgather support type/question An issue that's a question
#722 opened Sep 19, 2024 by yaroslavvb
1
Expected Data Format type/question An issue that's a question
#715 opened Aug 27, 2024 by aflah02
1
Which mmlu validation setting is recommend? type/question An issue that's a question
#714 opened Aug 27, 2024 by mathfinder
1
[Quick question]: How do I turn off FSDP? type/question An issue that's a question
#703 opened Aug 15, 2024 by candygocandy
1
RuntimeError: Triton Error [CUDA]: invalid device context type/bug An issue about a bug
#700 opened Aug 13, 2024 by andymvp2018
4
slurm script for: configs/official/OLMo-7B.yaml type/question An issue that's a question
#699 opened Aug 13, 2024 by andymvp2018
3
Gflops computation is faulty for FSDP due to bug in OLMo.num_params()
#695 opened Aug 7, 2024 by AkshitaB
1
why CrossEntropyLoss is zero,i type/question An issue that's a question
#692 opened Aug 6, 2024 by aizhweiwei
2
Olmo 0724 -hf checkpoints don't load the proper config when instantiating with OLMoForCausalLM type/bug An issue about a bug
#689 opened Aug 5, 2024 by sarahwie
2
Model ladder has no documentation type/documentation An issue or pull request related to documentation
#683 opened Jul 31, 2024 by IanMagnusson
1
mlp_ratio not adjusted in config if mlp_hidden_size is set type/bug An issue about a bug
#673 opened Jul 21, 2024 by Muennighoff
Does global_train_batch_size support gradient accumulation? type/question An issue that's a question
#672 opened Jul 21, 2024 by jinzhuoran
1
Is there explicitly instruction-following data in the version of Dolma used to train v1? type/question An issue that's a question
#658 opened Jul 15, 2024 by john-hewitt
1
Can long text be splitted into short texts? type/question An issue that's a question
#655 opened Jul 12, 2024 by CoinCheung
1
Cannot convert internal OLMo checkpoint to HF type/bug An issue about a bug
#654 opened Jul 11, 2024 by viking-sudo-rm
2
ProTip! Add no:assignee to see everything that’s not assigned.