Skip to content

P
pierre-humblot
Open
Opened by Angeline De La Condamine@angelinefct897
Report abuse New issue

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 improve reasoning capability. DeepSeek-R1 attains results on par with OpenAI's o1 model on several criteria, consisting of MATH-500 and SWE-bench.

DeepSeek-R1 is based on DeepSeek-V3, a mixture of experts (MoE) design recently open-sourced by DeepSeek. This base design is fine-tuned utilizing Group Relative Policy Optimization (GRPO), it-viking.ch a reasoning-oriented variant of RL. The research team also performed knowledge distillation from DeepSeek-R1 to open-source Qwen and Llama models and launched a number of versions of each; these models outperform bigger models, consisting of GPT-4, genbecle.com on mathematics and coding criteria.

[DeepSeek-R1 is] the primary step towards improving language design thinking capabilities using pure support learning (RL). Our objective is to check out the potential of LLMs to establish reasoning abilities without any monitored data, concentrating on their self-evolution through a pure RL process...DeepSeek-R1 ... master a large range of jobs, it-viking.ch including innovative writing, general question answering, editing, summarization, and more. Additionally, DeepSeek-R1 shows impressive efficiency on jobs requiring long-context understanding, larsaluarna.se substantially exceeding DeepSeek-V3 on long-context benchmarks.

To develop the design, DeepSeek began with DeepSeek-V3 as a base. They initially tried fine-tuning it just with RL, and without any supervised fine-tuning (SFT), trademarketclassifieds.com producing a model called DeepSeek-R1-Zero, which they have actually likewise released. This model displays strong reasoning efficiency, however" powerful reasoning behaviors, it deals with a number of concerns. For example, DeepSeek-R1-Zero deals with difficulties like bad readability and language blending."

To resolve this, the group used a brief phase of SFT to prevent the "cold start" problem of RL. They gathered several thousand examples of chain-of-thought reasoning to utilize in SFT of DeepSeek-V3 before running RL. After the RL procedure assembled, they then gathered more SFT information utilizing rejection tasting, leading to a dataset of 800k samples. This dataset was used for further fine-tuning and to produce the distilled models from Llama and Qwen.

DeepSeek assessed their design on a variety of reasoning, mathematics, and coding criteria and compared it to other models, including Claude-3.5- Sonnet, GPT-4o, and o1. DeepSeek-R1 exceeded all of them on numerous of the benchmarks, including 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 total in the arena and # 1 in coding and mathematics. It was likewise tied for # 1 with o1 in "Hard Prompt with Style Control" category.

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

Each response begins with a ... pseudo-XML tag containing the chain of thought used to assist create the reaction. [Given the prompt] "a joke about a pelican and a walrus who run a tea space together" ... It then thought for 20 paragraphs before outputting the joke! ... [T] he joke is terrible. But the procedure of getting there was such an intriguing insight into how these brand-new designs work.

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

DeepSeek is rapidly becoming a strong contractor of open designs. Not just are these entertainers, however their license permits use of their outputs for distillation, potentially pushing forward the state of the art for language models (and multimodal models) of all sizes.

The DeepSeek-R1 models are available on HuggingFace.

About the Author

Anthony Alford

Rate this Article

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

Related Topics:

- AI, ML & Data Engineering