DeepSeek-R1 · GitHub Models · GitHub

DeepSeek-R1 stands out at reasoning jobs utilizing a step-by-step training process, such as language, clinical thinking, and coding jobs. It features 671B overall criteria with 37B active specifications, and 128k context length.

DeepSeek-R1 develops on the development of earlier reasoning-focused designs that enhanced performance by extending Chain-of-Thought (CoT) thinking. DeepSeek-R1 takes things further by combining support learning (RL) with fine-tuning on thoroughly chosen datasets. It progressed from an earlier version, DeepSeek-R1-Zero, which relied exclusively on RL and showed strong reasoning abilities however had concerns like hard-to-read outputs and language inconsistencies. To deal with these restrictions, DeepSeek-R1 integrates a percentage of cold-start data and follows a refined training pipeline that mixes reasoning-oriented RL with supervised fine-tuning on curated datasets, leading to a design that attains modern performance on thinking criteria.

Usage Recommendations

We advise adhering to the following setups when making use of the DeepSeek-R1 series designs, including benchmarking, to attain the efficiency:

– Avoid including a system timely; all guidelines should be contained within the user timely.
– For mathematical problems, it is suggested to include a directive in your prompt such as: “Please reason action by action, and put your last response within boxed .”.
– When assessing design performance, it is recommended to perform multiple tests and balance the results.

Additional suggestions

The design’s reasoning output (contained within the tags) may include more hazardous material than the design’s last action. Consider how your application will use or display the thinking output; you may wish to suppress the thinking output in a production setting.