tensorhydra/qwen3-8b-aimo3-tir
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Mar 28, 2026License:qwen-researchArchitecture:Transformer Cold

The tensorhydra/qwen3-8b-aimo3-tir is an 8 billion parameter Qwen-8B model, fine-tuned using LoRA for tool-integrated reasoning tasks. It is specifically optimized for mathematical reasoning and problem-solving based on the AIMO3 competition dataset, which was synthesized by GPT-OSS-120B. The LoRA adapters are merged into the base model, provided in SafeTensors format, making it ready for straightforward deployment without PEFT dependencies. This model excels at complex reasoning problems requiring tool use and step-by-step solutions.

Loading preview...