FuseAI/FuseO1-QwQ-DeepSeekR1-LightR1-32B
TEXT GENERATIONConcurrency Cost:2Model Size:32.8BQuant:FP8Ctx Length:32kPublished:Mar 7, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

FuseAI/FuseO1-QwQ-DeepSeekR1-LightR1-32B is a 32.8 billion parameter language model developed by the FuseAI Team, created by merging Qwen/QwQ-32B, deepseek-ai/DeepSeek-R1-Distill-Qwen-32B, and qihoo360/Light-R1-32B using the SCE merging method. This model is specifically optimized for System-II reasoning capabilities, excelling in complex mathematical, coding, and scientific tasks. It demonstrates significant performance improvements on benchmarks like AIME24, LiveCodeBench, and GPQA-Diamond compared to its constituent models.

Loading preview...