Mathematical Reasoning on GSM8K (TPS)
834.52Tokens Per Second (TPS)MBD-LLaDA2-Mini-DMax
Evaluation Results
| Method | Links | |
|---|---|---|
| MBD-LLaDA2-Mini-DMaxHardware=two H100 GPUs, Tensor Parallelism=TP=2, Decoding=single-sample2026.06 | 834.52 | |
| LLaDA2-Mini-DMaxHardware=two H100 GPUs, Tensor Parallelism=TP=2, Decoding=single-sample2026.06 | 700.82 | |
| MBD-LLaDA2-MiniHardware=two H100 GPUs, Tensor Parallelism=TP=2, Decoding=single-sample2026.06 | 687.87 | |
| LLaDA2-MiniHardware=two H100 GPUs, Tensor Parallelism=TP=2, Decoding=single-sample2026.06 | 344.05 |