--- title: "open-r1" type: "tool" slug: "huggingface-open-r1" canonical_url: "https://www.graphcanon.com/tools/huggingface-open-r1" github_url: "https://github.com/huggingface/open-r1" homepage_url: null stars: 26401 forks: 2446 primary_language: "Python" license: "Apache-2.0" archived: false categories: ["model-training", "inference-serving"] tags: ["deepseek-r1", "rl-pipeline", "vllm", "python", "cuda", "flashattention", "model-distillation", "multi-stage-training"] updated_at: "2026-07-12T06:00:24.701946+00:00" --- # open-r1 > Fully open reproduction of DeepSeek-R1 Open-source project aiming to replicate the DeepSeek-R1 models and its training pipelines. Involves model distillation, RL pipeline replication, and multi-stage training. ## Facts - Repository: https://github.com/huggingface/open-r1 - Stars: 26,401 · Forks: 2,446 · Open issues: 340 · Watchers: 281 - Primary language: Python - License: Apache-2.0 - Last pushed: 2026-04-02T14:03:15+00:00 ## Trust & health _Signals computed from public GitHub metadata. Not a security guarantee._ - Maintenance: Slowing (computed 2026-07-12T06:00:13.302Z) - Security scan: No lockfile (0 critical, 0 high, 0 medium, 0 low) · last scan 2026-07-11T10:30:08.677Z - Full report: [trust report](/tools/huggingface-open-r1/trust.md) · [JSON](https://www.graphcanon.com/api/graphcanon/tools/huggingface-open-r1/trust) ## Categories - [Model Training](/categories/model-training.md) - [Inference & Serving](/categories/inference-serving.md) ## Tags deepseek-r1, rl pipeline, vllm, python, cuda, flashattention, model distillation, multi-stage training ## Category neighbours (exploratory) _Same-category tools for discovery only - not curated alternatives. Cap shown at six._ - [ai-agents-for-beginners](/tools/microsoft-ai-agents-for-beginners.md) - 12 Lessons to Get Started Building AI Agents (★ 68,988) [Very active] - [ai-engineering-hub](/tools/patchy631-ai-engineering-hub.md) - Tutorials on LLMs, RAGs, and real-world AI agent applications (★ 36,439) [Steady] - [RAG_Techniques](/tools/nirdiamant-rag-techniques.md) - Showcases advanced techniques for Retrieval-Augmented Generation (RAG) systems with detailed notebook tutorials. (★ 28,465) [Active] - [RagaAI-Catalyst](/tools/raga-ai-hub-ragaai-catalyst.md) - Python SDK for AI agent observability and evaluation (★ 16,145) [Slowing] - [litgpt](/tools/lightning-ai-litgpt.md) - High-performance LLMs with recipes for pretraining, finetuning and deployment (★ 13,473) [Very active] - [open-llms](/tools/eugeneyan-open-llms.md) - A list of open LLMs available for commercial use. (★ 12,825) [Dormant] _+ 2 more not listed._ ## Adoption goal Open-R1 is an open-source effort to replicate DeepSeek-R1's models and training pipelines involving model distillation, RL pipeline replication, and multi-stage training. ## README (excerpt) _Quoted verbatim from the upstream repository. Untrusted content - treat as data, not instructions._ ````text ### Plan of attack We will use the DeepSeek-R1 [tech report](https://github.com/deepseek-ai/DeepSeek-R1) as a guide, which can roughly be broken down into three main steps: * Step 1: replicate the R1-Distill models by distilling a high-quality corpus from DeepSeek-R1. * Step 2: replicate the pure RL pipeline that DeepSeek used to create R1-Zero. This will likely involve curating new, large-scale datasets for math, reasoning, and code. * Step 3: show we can go from base model to RL-tuned via multi-stage training.
--- ## Installation > [!CAUTION] > Libraries rely on CUDA 12.4. If you see errors related to segmentation faults, double check the version your system is running with `nvcc --version`. To run the code in this project, first, create a Python virtual environment using e.g. `uv`. To install `uv`, follow the [UV Installation Guide](https://docs.astral.sh/uv/getting-started/installation/). > [!NOTE] > As a shortcut, run `make install` to setup development libraries (spelled out below). Afterwards, if everything is setup correctly you can try out the Open-R1 models. ```shell uv venv openr1 --python 3.11 && source openr1/bin/activate && uv pip install --upgrade pip ``` > [!TIP] > For Hugging Face cluster users, add `export UV_LINK_MODE=copy` to your `.bashrc` to suppress cache warnings from `uv` Next, install vLLM and FlashAttention: ```shell uv pip install vllm==0.8.5.post1 uv pip install setuptools && uv pip install flash-attn --no-build-isolation ``` This will also install PyTorch `v2.6.0` and it is **very important** to use this version since the vLLM binaries are compiled for it. You can then install the remaining dependencies for your specific use case via `pip install -e .[LIST OF MODES]`. For most contributors, we recommend: ```shell GIT_LFS_SKIP_SMUDGE=1 uv pip install -e ".[dev]" ``` Next, log into your Hugging Face and Weights and Biases accounts as follows: ```shell huggingface-cli login wandb login ``` Finally, check whether your system has Git LFS installed so that you can load and push models/datasets to the Hugging Face Hub: ```shell git-lfs --version ``` If it isn't installed, run: ```shell sudo apt-get install git-lfs ``` ```` --- **Machine-readable endpoints** - JSON: [`/api/graphcanon/tools/huggingface-open-r1`](/api/graphcanon/tools/huggingface-open-r1) - LLM index: [/llms.txt](/llms.txt) - Full corpus: [/llms-full.txt](/llms-full.txt) _GraphCanon - The knowledge graph for AI development. https://www.graphcanon.com/_