DeepSeek open-sourced DeepSeek-R1, an LLM fine-tuned with reinforcement knowing (RL) to improve thinking capability. DeepSeek-R1 attains results on par with OpenAI's o1 model on a number of standards, including MATH-500 and SWE-bench.

DeepSeek-R1 is based on DeepSeek-V3, a mixture of specialists (MoE) model just recently open-sourced by DeepSeek. This base model is fine-tuned utilizing Group Relative Policy Optimization (GRPO), a reasoning-oriented variation of RL. The research team likewise performed knowledge distillation from DeepSeek-R1 to open-source Qwen and Llama designs and released a number of variations of each; these designs exceed bigger models, including GPT-4, on math and coding criteria.

[DeepSeek-R1 is] the primary step toward enhancing language model thinking abilities utilizing pure reinforcement learning (RL). Our goal is to check out the capacity of LLMs to establish thinking abilities without any monitored information, concentrating on their self-evolution through a pure RL process...DeepSeek-R1 ... master a broad variety of tasks, including creative writing, general question answering, editing, summarization, and more. Additionally, DeepSeek-R1 demonstrates exceptional efficiency on jobs requiring long-context understanding, considerably exceeding DeepSeek-V3 on long-context criteria.

To establish the design, DeepSeek started with DeepSeek-V3 as a base. They initially tried fine-tuning it just with RL, and without any monitored fine-tuning (SFT), producing a model called DeepSeek-R1-Zero, which they have actually also released. This design exhibits strong thinking performance, but" powerful reasoning behaviors, it deals with several issues. For instance, DeepSeek-R1-Zero fights with obstacles like bad readability and language mixing."

To address this, the team utilized a brief phase of SFT to avoid the "cold start" problem of RL. They collected several thousand examples of chain-of-thought thinking to use in SFT of DeepSeek-V3 before running RL. After the RL procedure converged, they then collected more SFT information using rejection tasting, leading to a dataset of 800k samples. This dataset was utilized for additional fine-tuning and to produce the distilled designs from Llama and Qwen.

DeepSeek evaluated their design on a variety of thinking, mathematics, and coding criteria and compared it to other models, consisting of Claude-3.5- Sonnet, GPT-4o, and o1. DeepSeek-R1 outperformed all of them on several of the criteria, including AIME 2024 and MATH-500.

DeepSeek-R1 Performance. Image Source: DeepSeek-R1 Technical Report

Within a few days of its release, the LMArena revealed that DeepSeek-R1 was ranked # 3 general in the arena and # 1 in coding and mathematics. It was also tied for # 1 with o1 in "Hard Prompt with Style Control" category.

Django framework co-creator Simon Willison discussed his experiments with among the DeepSeek distilled Llama models on his blog:

Each action starts with a ... pseudo-XML tag containing the chain of idea used to assist create the reaction. [Given the prompt] "a joke about a pelican and a walrus who run a tea room together" ... It then thought for 20 paragraphs before outputting the joke! ... [T] he joke is terrible. But the process of arriving was such a fascinating insight into how these brand-new models work.

Andrew Ng's newsletter The Batch discussed DeepSeek-R1:

DeepSeek is rapidly becoming a strong home builder of open models. Not only are these designs excellent entertainers, but their license allows usage of their outputs for distillation, possibly pressing forward the state of the art for language models (and wavedream.wiki multimodal designs) of all sizes.

The DeepSeek-R1 designs are available on HuggingFace.

About the Author

Anthony Alford

Rate this Article

This content remains in the AI, ML & Data Engineering subject

Related Topics:

- AI, ML & Data Engineering - Generative AI

• Large language models
  
  - Related Editorial
  
  Related Sponsored Content
  
  - [eBook] Starting with Azure Kubernetes Service
  
  Related Sponsor
  
  Free services for AI apps. Are you prepared to experiment with innovative innovations? You can begin building intelligent apps with free Azure app, information, and AI services to minimize upfront expenses. Learn More.
  
  How could we improve? Take the InfoQ reader study
  
  Each year, we look for feedback from our readers to assist us improve InfoQ. Would you mind spending 2 minutes to share your in our brief study? Your feedback will straight help us continuously evolve how we support you. The InfoQ Team Take the study
  
  Related Content
  
  The InfoQ Newsletter
  
  A round-up of last week's material on InfoQ sent out every Tuesday. Join a neighborhood of over 250,000 senior developers.