Open R1
Training, evaluation, and synthetic-data generation scripts for reproducing DeepSeek-R1-style reasoning models.Apache-2.0
Training, evaluation, and synthetic-data generation scripts for reproducing DeepSeek-R1-style reasoning models.
Open R1 supplies Python scripts and recipes for supervised fine-tuning, GRPO training, model evaluation, and generating synthetic reasoning data. Researchers can use its distributed training configurations and Slurm examples to run experiments on GPU infrastructure. Dataset tools compare prompts against benchmark data with n-grams and can remove contaminated samples before further training.
Open R1 README · Reviewed 2026-10-16
- 01 / Feature noteDataset decontamination
Compare dataset prompts against benchmark data with n-grams and remove contaminated samples.
Feature source ↗
GitHub stars
26,476
Detected license
Apache-2.0
Repository created
2025-01-24
Repository push
2026-10-26
2025-01-24Repository push
2026-10-26
GitHub stars
Unknown
i
Cached history uses a different month window; chart withheld until aligned.
Latest attempt: ok · Checked 2026-10-17T00:32:23.054453+00:00 · Last successful fetch 2026-10-17T00:32:23.054453+00:00 · Provider source
Hacker News stories
3 · 2024-11 – 2026-10
i
OpenSSF Scorecard
Unknown
i
No result returned.
Checked 2026-10-31T03:25:05Z · Latest attempt: not found · Provider source
GitHub release
Unknown
i
No result returned.
Checked 2026-11-04T04:26:56Z · Latest attempt: not found · Provider source
Unknown
