← All models

DeepSeek-R1-Distill-Qwen-7B

DeepSeek · 2025-01-20 · 7.6B parameters

DeepSeek-R1-Distill-Qwen-7B is a dense reasoning checkpoint produced by supervised fine-tuning of Qwen2.5-Math-7B on approximately 800,000 samples curated with DeepSeek-R1, without an RL stage. Released January 20, 2025 under the MIT license as part of the original six-model R1 distillation series, its architecture config declares 131,072 positions.

Benchmark scores