DeepSeek Open-Sources DeepSeek-R1 LLM with Performance Comparable To OpenAI's O1 Model

DeepSeek open-sourced DeepSeek-R1, an LLM fine-tuned with reinforcement learning (RL) to enhance reasoning capability. DeepSeek-R1 attains outcomes 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 professionals (MoE) model just recently open-sourced by DeepSeek. This base model is fine-tuned utilizing Group Relative Policy Optimization (GRPO), a reasoning-oriented version of RL. The research study group likewise performed knowledge distillation from DeepSeek-R1 to open-source Qwen and Llama models and released a number of versions of each; these designs exceed larger designs, including GPT-4, on mathematics and coding standards.

[DeepSeek-R1 is] the very first action towards enhancing language model reasoning abilities utilizing pure support knowing (RL). Our goal is to check out the potential of LLMs to establish thinking capabilities with no monitored information, focusing on their self-evolution through a pure RL process...DeepSeek-R1 ... master a large range of tasks, consisting of imaginative writing, basic concern answering, modifying, summarization, and more. Additionally, DeepSeek-R1 demonstrates exceptional performance on jobs requiring long-context understanding, significantly outshining DeepSeek-V3 on long-context criteria.

To establish the design, DeepSeek began with DeepSeek-V3 as a base. They first tried fine-tuning it only with RL, and without any supervised fine-tuning (SFT), producing a model called DeepSeek-R1-Zero, which they have actually likewise launched. This model shows strong thinking performance, however" powerful thinking behaviors, it faces numerous issues. For circumstances, DeepSeek-R1-Zero fights with obstacles like poor readability and language mixing."

To resolve this, the team used a brief stage of SFT to avoid the "cold start" problem of RL. They collected numerous thousand examples of chain-of-thought reasoning to utilize in SFT of DeepSeek-V3 before running RL. After the RL procedure converged, they then gathered more SFT information utilizing rejection tasting, resulting in a dataset of 800k samples. This dataset was used for additional fine-tuning and to produce the distilled designs from Llama and Qwen.

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

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

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

Django framework co-creator Simon Willison composed about his try outs 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 produce the response. [Given the timely] "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 awful. But the process of getting there was such a fascinating insight into how these new models work.

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

DeepSeek is quickly becoming a strong contractor of open designs. Not only are these models excellent entertainers, but their license allows use of their outputs for distillation, engel-und-waisen.de potentially pushing forward the cutting-edge for language designs (and multimodal models) of all sizes.

The DeepSeek-R1 models are available on HuggingFace.

About the Author

Anthony Alford

Rate this Article