GitLab

DeepSeek open-sourced DeepSeek-R1, an LLM fine-tuned with reinforcement learning (RL) to improve reasoning capability. DeepSeek-R1 attains outcomes on par with OpenAI's o1 model on numerous standards, including MATH-500 and SWE-bench.

DeepSeek-R1 is based upon DeepSeek-V3, a mix of experts (MoE) design just recently open-sourced by DeepSeek. This base design is fine-tuned using Group Relative Policy Optimization (GRPO), a reasoning-oriented variant of RL. The research group also performed knowledge distillation from DeepSeek-R1 to open-source Qwen and Llama designs and launched several versions of each; these designs surpass bigger models, including GPT-4, on math and coding standards.

[DeepSeek-R1 is] the first action towards improving language model reasoning capabilities utilizing pure support (RL). Our goal is to explore the capacity of LLMs to develop thinking capabilities with no monitored data, focusing on their self-evolution through a pure RL process...DeepSeek-R1 ... excels in a wide variety of jobs, including creative writing, basic concern answering, modifying, summarization, and more. Additionally, DeepSeek-R1 demonstrates impressive efficiency on jobs requiring long-context understanding, considerably surpassing DeepSeek-V3 on long-context criteria.

To establish the model, DeepSeek started with DeepSeek-V3 as a base. They first tried fine-tuning it just with RL, and with no supervised fine-tuning (SFT), producing a design called DeepSeek-R1-Zero, trademarketclassifieds.com which they have also launched. This design displays strong thinking performance, however" effective reasoning habits, it faces a number of concerns. For example, DeepSeek-R1-Zero battles with obstacles like poor readability and language blending."

To resolve this, the group utilized a brief phase of SFT to avoid the "cold start" problem of RL. They gathered several thousand examples of chain-of-thought reasoning to use in SFT of DeepSeek-V3 before running RL. After the RL procedure converged, they then collected more SFT information utilizing rejection sampling, 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 examined their model on a variety of thinking, mathematics, and coding standards and compared it to other designs, consisting of Claude-3.5- Sonnet, GPT-4o, and o1. DeepSeek-R1 exceeded all of them on several of the standards, 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 overall in the arena and # 1 in coding and math. It was likewise tied for # 1 with o1 in "Hard Prompt with Style Control" classification.

Django structure co-creator Simon Willison composed about his explores among the DeepSeek distilled Llama models on his blog:

Each action begins with a ... pseudo-XML tag containing the chain of thought utilized to assist generate the action. [Given the prompt] "a joke about a pelican and a walrus who run a tea room together" ... It then believed for 20 paragraphs before outputting the joke! ... [T] he joke is dreadful. But the process of arriving was such an interesting insight into how these new designs work.

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

DeepSeek is rapidly emerging as a strong contractor of open models. Not only are these designs fantastic entertainers, but their license permits usage of their outputs for distillation, possibly pushing forward the state of the art for language models (and multimodal designs) of all sizes.

The DeepSeek-R1 designs are available on HuggingFace.

About the Author

Anthony Alford

Rate this Article

This material remains in the AI, ML & Data Engineering topic

Related Topics:

- AI, ML & Data Engineering

• Generative AI
• Large language models
  
  - Related Editorial
  
  Related Sponsored Content
  
  - [eBook] Getting Going with Azure Kubernetes Service
  
  Related Sponsor
  
  Free services for AI apps. Are you prepared to try out cutting-edge innovations? You can begin building smart apps with complimentary Azure app, data, and AI services to lessen upfront costs. Learn More.
  
  How could we enhance? Take the InfoQ reader survey
  
  Each year, we seek feedback from our readers to help us enhance InfoQ. Would you mind spending 2 minutes to share your feedback in our short survey? Your feedback will straight assist us continuously develop how we support you. The InfoQ Team Take the survey
  
  Related Content
  
  The InfoQ Newsletter
  
  A round-up of last week's material on InfoQ sent every Tuesday. Join a community of over 250,000 senior developers.