List your company

Mathematical Reasoning on GSM-8K (train)

0.056Train LossMGUP-AdamW
0.0550.061750.06850.07525Jun 16, 2026

Evaluation results

MethodLinks
MGUP-AdamWModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning2026.060.056
MicroAdamWModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning, m=102026.060.057
AdamWModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning2026.060.064
AdamW-8bModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning2026.060.069
GALoreModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning, rank=5122026.060.07
LDAdamWModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning, rank=5122026.060.073
C-AdamWModel=LLaMA2-7B, Evaluation Protocol=Fine-tuning2026.060.081