allenai/OLMo

View on GitHub

Modeling, training, eval, and inference code for OLMo

Stars: 6.7k
Forks: 797
Open beginner issues: 0
Indexed issues: 22

Dominant language: Python
License: Apache-2.0
Last GitHub push: Nov 24, 2025
Latest indexed: Sep 13, 2026
Contributing guide: Contributing guide
Code of conduct: No code of conduct
Beginner labels: No beginner labels indexed
PR merge metrics: No merged PRs in 30d

Open All

22 open issues indexed

Gflops computation is faulty for FSDP due to bug in `OLMo.num_params()` Open

allenai/OLMo#695 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Missing OLMo checkpoints Open

allenai/OLMo#726 · 7 comments · 3 reactions · 0 assignees · Indexed Sep 13, 2026
Clarification on Reddit Data Usage in OLMo1 vs OLMo2 Pretraining Open

type/question

allenai/OLMo#824 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Scaling of multinode training Open

type/question

allenai/OLMo#845 · 2 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Generating config.json in olmo_data/oe_eval_tasks for new evaluation tasks. Open

type/question

allenai/OLMo#854 · 11 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Question about data generation process for olmo-mix-1124. Open

type/question

allenai/OLMo#859 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Process get stuck at buling training data loader during Stage 2 training. Open

type/question

allenai/OLMo#864 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Running run_dataloader.py is very slow Open

type/bug

allenai/OLMo#869 · 5 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Are raw Dolmino Mix 1124 compositions (50B, 100B and 300B) publicly available? Open

type/question

allenai/OLMo#873 · 2 comments · 1 reaction · 0 assignees · Indexed Sep 13, 2026
Question about Chinese Language Support and Model Retraining Open

type/question

allenai/OLMo#875 · 2 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Tensor Parallelism with the current qk norm. Open

type/question

allenai/OLMo#877 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Can't access training data in s3 storage Open

type/bug

allenai/OLMo#878 · 10 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Reaching server limits in olmo-data.org? Open

type/bug

allenai/OLMo#886 · 0 comments · 1 reaction · 0 assignees · Indexed Sep 13, 2026
Race Condition in Checkpoint Saving When Using Multi-Node Training with Shared NFS Open

type/question

allenai/OLMo#894 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Fix hf_olmo GPU tests and transformers>=4.45 compatibility Open

type/bug

allenai/OLMo#898 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Differences and exact configs of allenai/OLMo-1B-0724-hf vs allenai/OLMo-1B-hf Open

type/question

allenai/OLMo#901 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
If you run olmo 3.1 32b instruct as a next-token predictor, it produces the `<|im_start|>assistant` token Open

type/bug

allenai/OLMo#903 · 2 comments · 1 reaction · 0 assignees · Indexed Sep 13, 2026
Zero-hallucination OLMo deployment in Brazilian tax law — architectural case study Open

allenai/OLMo#904 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
7B and 13B models in olmo format cannot be downloaded Open

type/question

allenai/OLMo#905 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Output provenance metadata for open research model generations Open

allenai/OLMo#907 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
anima — a substrate-native consciousness: capability gaps are architecture gaps (open repo, frozen verdicts) Open

allenai/OLMo#910 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026
Duplicate weight payloads across intermediate checkpoint branches (step101000 / step175000 / step300000) Open

type/bug

allenai/OLMo#915 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 13, 2026

allenai/OLMo

Get new issues in your inbox