Skip to content

Choose Your Path

OGAL = 4.6M Active Learning experiments archived at DOI:10.25532/OPARA-862.

Pick the question that matches your goal:

  • "I want to analyze the published dataset for my own research"

    Mine the 4.6M pre-computed results—no experiments needed.

    Analyze OPARA

  • "I want to reproduce the paper or run experiments from scratch"

    Run the exact scripts for the paper's figures, or recompute on HPC/SLURM.

    Reproduce & Run

  • "I want to extend the dataset with new strategies/hyperparameters"

    Add your experiments and integrate them with the shared benchmark.

    Extend the Benchmark

  • "I want research ideas"

    Open questions and unexplored directions using OGAL data.

    Research Ideas

Getting Started

# Setup
conda create --name ogal --file conda-linux-64.lock && conda activate ogal && poetry install
cp .server_access_credentials.cfg.example .server_access_credentials.cfg
# edit .server_access_credentials.cfg → set OUTPUT_PATH and DATASETS_PATH under [LOCAL]

# Download archived results
wget -c -O full_exp_jan.zip \
  "https://opara.zih.tu-dresden.de/bitstreams/38951489-5076-4544-a99b-c20dddfc2c6b/download"
unzip full_exp_jan.zip -d /path/to/results/

# Generate leaderboard
python -m eva_scripts.final_leaderboard --EXP_TITLE full_exp_jan

📄 Paper ・ 📦 Dataset (DOI) ・ 💻 GitHub