Skip to content

@OpenEuroLLM OpenEuroLLM

All

    Repositories list

    55 repositories

    • non_web_data

      Public
      In this repository we will add the scripts and other indications used for the file extraction of non web data for the OELLM project
      Jupyter Notebook
      •1•0•0•0•Updated Oct 6, 2026Oct 6, 2026

    • JudgeArena

      Public
      Evaluating LLM with swappable judges: local, remote, openrouter on multiple benchmarks.
      Python
      •
      Apache License 2.0
      •6•18•6•2•Updated Oct 6, 2026Oct 6, 2026

    • annealing-experiments

      Public
      Calculations, data mixtures, scripts, etc. around annealing the various production runs
      Python
      •0•0•0•0•Updated Oct 6, 2026Oct 6, 2026

    • training-data-packer

      Public
      Packaging annotated datasets into final training data. Part of the OpenEuroLLM project.
      trainingaidatasets
      trainingaidatasetsllmopeneurollm
      Python
      •
      Apache License 2.0
      •1•2•11•1•Updated Oct 6, 2026Oct 6, 2026

    • quickdash

      Public
      quick eval dashboard
      JavaScript
      •
      Apache License 2.0
      •1•0•2•0•Updated Oct 6, 2026Oct 6, 2026

    • eval-framework

      Public
      Mirror of https://github.com/Aleph-Alpha-Research/eval-framework/
      Python
      •
      Apache License 2.0
      •0•0•0•0•Updated Oct 6, 2026Oct 6, 2026

    • oellm-eval

      Public
      A CLI to run reproducible evaluation across EuroHPC clusters
      Python
      •24•21•5•4•Updated Oct 5, 2026Oct 5, 2026

    • oellm-autoexp

      Public
      Merge of megatron-train, autoexperiment and oellm_pretrain.
      Python
      •
      Apache License 2.0
      •6•16•1•2•Updated Oct 5, 2026Oct 5, 2026

    • warc2text

      Public
      Extracts plain text, language identification and more metadata from WARC records
      C++
      •
      MIT License
      •7•0•1•0•Updated Oct 3, 2026Oct 3, 2026

    • lm-evaluation-harness

      Public
      A framework for few-shot evaluation of language models.
      Python
      •
      MIT License
      •3.6k•0•0•0•Updated Oct 3, 2026Oct 3, 2026

    • NVIDIA-Megatron-LM

      Public
      Ongoing research training transformer models at scale
      Python
      •
      Other
      •4.6k•0•0•4•Updated Oct 2, 2026Oct 2, 2026

    • notebooks

      Public
      Jupyter Notebook
      •
      Apache License 2.0
      •0•1•0•0•Updated Sep 30, 2026Sep 30, 2026

    • synthgen

      Public
      Python
      •
      Other
      •0•0•0•0•Updated Sep 30, 2026Sep 30, 2026

    • training-data-collection

      Public
      Internal Collection Training Data Used in Production
      Python
      •2•6•0•1•Updated Sep 29, 2026Sep 29, 2026

    • ComplexKDA

      Public
      machine-learningrnnstate-space-models
      machine-learningrnnstate-space-modelsdeltanetlinear-rnn
      Python
      •
      MIT License
      •0•8•0•0•Updated Sep 28, 2026Sep 28, 2026

    • openeurollm-model-identity

      Public
      Multilingual synthetic conversations for accurate, bounded OpenEuroLLM model self-knowledge
      Python
      •0•0•0•0•Updated Sep 24, 2026Sep 24, 2026

    • training-data-catalogue

      Public
      Curated Public Repository of LLM (Pre-)Training Data
      Shell
      •6•15•44•0•Updated Sep 22, 2026Sep 22, 2026

    • post-training

      Public
      Repo for post-training LLMs
      Python
      •5•9•2•3•Updated Sep 18, 2026Sep 18, 2026

    • WSM-merge

      Public
      small study of merge
      Python
      •0•0•0•0•Updated Sep 17, 2026Sep 17, 2026

    • tmax-reproduction

      Public archive
      Archived code-only record of a partial TMAX infrastructure reproduction on Leonardo; no RL-performance result.
      reinforcement-learningtmaxopeneurollm
      reinforcement-learningtmaxopeneurollmopen-instruct
      Python
      •
      Apache License 2.0
      •0•0•0•0•Updated Aug 31, 2026Aug 31, 2026

    • function-calling

      Public
      Public index for current and historical OpenEuroLLM function-calling data, post-training, evaluation, and research artifacts.
      0•0•0•0•Updated Aug 31, 2026Aug 31, 2026

    • function-calling-post-training

      Public
      Historical OpenEuroLLM source snapshots for the OLMo 3 pilot and concluded Qwen3.5 controlled campaign.
      Python
      •
      Apache License 2.0
      •0•0•0•0•Updated Aug 31, 2026Aug 31, 2026

    • dense_english_scaling_laws

      Public
      Jupyter Notebook
      •0•4•1•0•Updated Aug 28, 2026Aug 28, 2026

    • pii-detection-solo

      Public
      LUMI-ready deployment of PII detection inference with openai privacy-filter model
      Python
      •1•0•0•0•Updated Aug 26, 2026Aug 26, 2026

    • web-data-processing

      Public
      Pipeline for text and metadata extraction from different web data archives
      0•0•1•0•Updated Aug 3, 2026Aug 3, 2026

    • tau2-bench

      Public
      OpenEuroLLM tau2-bench evaluation harness with OLMo and Qwen serving, user-simulator, and aggregation scripts.
      evaluationtool-usefunction-calling
      evaluationtool-usefunction-callingolmotau2
      Python
      •
      MIT License
      •0•0•0•0•Updated Aug 3, 2026Aug 3, 2026

    • bfcl

      Public
      OpenEuroLLM snapshot of the Berkeley Function Calling Leaderboard evaluation harness and OLMo evaluation orchestration.
      evaluationtool-usefunction-calling
      evaluationtool-usefunction-callingolmobfcl
      Python
      •
      Apache License 2.0
      •0•0•0•0•Updated Aug 3, 2026Aug 3, 2026

    • oellm-tasks

      Public
      Evaluation tasks created within the OpenEuroLLM project to be used in lm-eval harness.
      Python
      •
      MIT License
      •1•0•1•1•Updated Jul 30, 2026Jul 30, 2026

    • pretraining-decontamination

      Public
      Pipeline for contamination identification in pretraining datasets (2nd cycle)
      Python
      •0•0•0•0•Updated Jul 21, 2026Jul 21, 2026

    • titan-oellm

      Public
      OpenEuroLLM adaptation of TorchTitan
      Python
      •
      Apache License 2.0
      •0•0•0•0•Updated Jul 14, 2026Jul 14, 2026
    ProTip! When viewing an organization's repositories, you can use the props. filter to filter by custom property.