# deepseek-ai/deepseek-r1

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/deepseek-ai-deepseek-r1).**

91,996 stars · 11,716 forks · MIT

## Links

- GitHub: https://github.com/deepseek-ai/DeepSeek-R1
- awesome-repositories: https://awesome-repositories.com/repository/deepseek-ai-deepseek-r1.md

## Description

DeepSeek-R1 is an open-weights large language model focused on advanced reasoning. It uses chain-of-thought processing and internal monologues to solve complex mathematical and logical problems by breaking tasks into sequential, verifiable thought processes.

The model is developed using reinforcement learning to optimize reasoning patterns and verify logical steps. It employs a distillation process to transfer these high-performance logic capabilities from a large teacher model into smaller, computationally efficient versions.

The training framework incorporates group relative policy optimization, cold-start supervised fine-tuning, and multi-stage model distillation. These methods are supported by large-scale compute orchestration across GPU clusters.

## Tags

### Artificial Intelligence & ML

- [Complex Problem Solving](https://awesome-repositories.com/f/artificial-intelligence-ml/complex-problem-solving.md) — Provides advanced reasoning capabilities for solving intricate logical and mathematical challenges. ([source](https://github.com/deepseek-ai/deepseek-r1#readme))
- [Advanced Reasoning Models](https://awesome-repositories.com/f/artificial-intelligence-ml/advanced-reasoning-models.md) — Generates internal monologues for self-correction and logical verification before producing final responses.
- [Chain of Thought Implementations](https://awesome-repositories.com/f/artificial-intelligence-ml/artificial-intelligence-tooling/language-model-integrations/model-orchestration-management/reasoning-engines/chain-of-thought-implementations.md) — Produces long-form internal monologues during inference to verify logic and self-correct errors.
- [Chain-of-Thought Prompting](https://awesome-repositories.com/f/artificial-intelligence-ml/chain-of-thought-prompting.md) — Implements core chain-of-thought methodology to break down complex problems into verifiable logical steps. ([source](https://github.com/deepseek-ai/deepseek-r1#readme))
- [Reinforcement Learning Alignment](https://awesome-repositories.com/f/artificial-intelligence-ml/large-language-models/reinforcement-learning-alignment.md) — Trained via reinforcement learning to optimize reasoning patterns and verify logical steps.
- [Model Distillation Pipelines](https://awesome-repositories.com/f/artificial-intelligence-ml/model-distillation-pipelines.md) — Transfers reasoning capabilities from large teacher models to smaller student models using generated rationales.
- [Open-Weights Models](https://awesome-repositories.com/f/artificial-intelligence-ml/open-weights-models.md) — Provides a pre-trained model with publicly available weights for advanced reasoning.
- [Reasoning Optimization](https://awesome-repositories.com/f/artificial-intelligence-ml/reasoning-models/reasoning-optimization.md) — Optimizes reasoning performance by rewarding correct final answers to generate internal chain-of-thought sequences.
- [Reasoning Process Controllers](https://awesome-repositories.com/f/artificial-intelligence-ml/reasoning-workflows/reasoning-process-controllers.md) — Structures thinking processes to ensure high accuracy and verifiable logic in final answers.
- [Group Relative Policy Optimization](https://awesome-repositories.com/f/artificial-intelligence-ml/group-relative-policy-optimization.md) — Employs group relative policy optimization to stabilize training and improve reasoning accuracy.
- [Knowledge Distillation](https://awesome-repositories.com/f/artificial-intelligence-ml/knowledge-distillation.md) — Transfers complex reasoning capabilities from large models to smaller versions to reduce computational costs.
- [Large Scale Training Suites](https://awesome-repositories.com/f/artificial-intelligence-ml/large-scale-training-suites.md) — Orchestrates distributed training across massive GPU clusters to handle high-parameter model updates.
- [Reasoning-Distilled Models](https://awesome-repositories.com/f/artificial-intelligence-ml/model-distillation-pipelines/reasoning-distilled-models.md) — Provides compact model versions that retain high-performance logic with lower computational overhead.
- [Cold-Start Initializations](https://awesome-repositories.com/f/artificial-intelligence-ml/supervised-fine-tuning/cold-start-initializations.md) — Uses a high-quality reasoning dataset for initial supervised fine-tuning to establish a baseline for reinforcement learning.

### Part of an Awesome List

- [Frontier Reasoning Models](https://awesome-repositories.com/f/awesome-lists/ai/frontier-reasoning-models.md) — Reinforcement learning framework for incentivizing reasoning capabilities.
- [LLM Providers and Models](https://awesome-repositories.com/f/awesome-lists/ai/llm-providers-and-models.md) — High-performance reasoning model for complex tasks.
- [Model Architectures](https://awesome-repositories.com/f/awesome-lists/ai/model-architectures.md) — Reasoning-focused model architecture and technical implementation.
- [Model Implementations](https://awesome-repositories.com/f/awesome-lists/ai/model-implementations.md) — First-generation reasoning models from a leading research lab.
- [Reasoning Models](https://awesome-repositories.com/f/awesome-lists/ai/reasoning-models.md) — State-of-the-art reasoning model for complex logical tasks.
- [Reward Shaping Techniques](https://awesome-repositories.com/f/awesome-lists/ai/reward-shaping-techniques.md) — Incentivizing reasoning capabilities through reinforcement learning.