List your company

Mathematical Reasoning on GSM8K (Composite Score)

88.16Composite ScoreSFTKey-Tag
30.252845.286460.3275.3536Dec 24, 2025

Evaluation results

MethodLinks
SFTKey-TagModel=Qwen3-8B-Base2025.1288.16
SFT-TagModel=Qwen3-8B-Base2025.1288.05
SFTKey-TagModel=Qwen2.5-7B2025.1284.2
SFT(baseline)Model=Qwen2.5-7B2025.1283.05
SFT-TagModel=Qwen2.5-7B2025.1283.02
SFT(baseline)Model=Qwen3-8B-Base2025.1282.18
SFTKey-TagModel=SmolLM3-3B-Base2025.1277.98
SFT-TagModel=SmolLM3-3B-Base2025.1277.75
SFT(baseline)Model=SmolLM3-3B-Base2025.1276.85
Key-TagModel=Qwen3-8B-Base2025.1273.6
Key-TagModel=SmolLM3-3B-Base2025.1258.5
Key-TagModel=Qwen2.5-7B2025.1255.19
Key-TagModel=Qwen2.5-3B2025.1250.3
SFT(baseline)Model=Qwen2.5-3B2025.1247.49
SFTKey-TagModel=Qwen2.5-3B2025.1247.39
SFT-TagModel=Qwen2.5-3B2025.1238.66
SFTKey-TagModel=Qwen2.5-1.5B2025.1236.88
Key-TagModel=Qwen2.5-1.5B2025.1234.3
SFT-TagModel=Qwen2.5-1.5B2025.1233.32
SFT(baseline)Model=Qwen2.5-1.5B2025.1232.48