List your company
ResearchFormerly WizWand

Code Reasoning on LeetCodeDataset

74.5Pass@4Hybrid-LoRA
-2.9817.13537.2557.365May 12, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Hybrid-LoRABackbone=Qwen-2.5 7B Instruct, Post-training Approach=GSPO, #Tunable Params=449M2026.0574.5
Hybrid-LoRABackbone=Qwen-2.5 7B Instruct, Post-training Approach=GRPO, #Tunable Params=449M2026.0573.2
AutoLoRABackbone=Qwen-2.5 7B Instruct, Post-training Approach=GSPO, #Tunable Params=453M2026.0572.5
AutoLoRABackbone=Qwen-2.5 7B Instruct, Post-training Approach=GRPO, #Tunable Params=453M2026.0571.5
Full-FT (ref.)Backbone=Qwen-2.5 3B Instruct, Post-training Approach=GSPO, #Tunable Params=3086M2026.0565.1
Hybrid-LoRABackbone=Qwen-2.5 3B Instruct, Post-training Approach=GSPO, #Tunable Params=192M2026.0564
Full-FT (ref.)Backbone=Qwen-2.5 3B Instruct, Post-training Approach=GRPO, #Tunable Params=3086M2026.0563.8
Hybrid-LoRABackbone=Qwen-2.5 3B Instruct, Post-training Approach=GRPO, #Tunable Params=192M2026.0562.7
AutoLoRABackbone=Qwen-2.5 3B Instruct, Post-training Approach=GSPO, #Tunable Params=194M2026.0562.6
AutoLoRABackbone=Qwen-2.5 3B Instruct, Post-training Approach=GRPO, #Tunable Params=201M2026.0559.6
Full-FT (ref.)Backbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GSPO, #Tunable Params=1544M2026.0549.1
Full-FT (ref.)Backbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GRPO, #Tunable Params=1544M2026.0548.3
Hybrid-LoRABackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GRPO, #Tunable Params=105M2026.0548
Hybrid-LoRABackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GSPO, #Tunable Params=105M2026.0547.6
LoRA-DropBackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GRPO, #Tunable Params=105M2026.0546.3
AutoLoRABackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GRPO, #Tunable Params=106M2026.0545.7
AdaLoRABackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GRPO, #Tunable Params=105M2026.0545.4
AutoLoRABackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GSPO, #Tunable Params=107M2026.0545.2
AdaLoRABackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GSPO, #Tunable Params=106M2026.0544.8
LoRA-DropBackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GSPO, #Tunable Params=105M2026.0544.7
LoRA-FFNBackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=GSPO, #Tunable Params=122M2026.0544.5
Full-FTBackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=SFT, #Tunable Params=1544M2026.0540.6
CoTBackbone=Qwen-2.5 3B Instruct, Post-training Approach=-, #Tunable Params=0.02026.0540.2
COTBackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=COT, #Tunable Params=0.02026.0514.6
IOBackbone=Qwen-2.5 1.5B Instruct, Post-training Approach=IO, #Tunable Params=0.02026.050