General Language Understanding on 10-benchmark suite average
73.81Average AccuracyFP
Evaluation results
| Method | Links | |
|---|---|---|
| FPModel=LLaDA-Instruct, Quantization=Full Precision2026.06 | 73.81 | |
| FPModel=LLaDA-1.5, Quantization=Full Precision2026.06 | 73.53 | |
| FAIR-CalibModel=LLaDA-1.5, Quantization=W4A42026.06 | 72.75 | |
| FAIR-CalibModel=LLaDA-Instruct, Quantization=W4A42026.06 | 72.4 | |
| FlatQuantModel=LLaDA-1.5, Quantization=W4A42026.06 | 71.94 | |
| FlatQuantModel=LLaDA-Instruct, Quantization=W4A42026.06 | 71.38 | |
| QuaRotModel=LLaDA-Instruct, Quantization=W4A42026.06 | 71.34 | |
| QuaRotModel=LLaDA-1.5, Quantization=W4A42026.06 | 71.05 | |
| RTNModel=LLaDA-1.5, Quantization=W4A42026.06 | 67.4 | |
| RTNModel=LLaDA-Instruct, Quantization=W4A42026.06 | 66.83 | |
| FPModel=LLaDA-Base, Quantization=Full Precision2026.06 | 62.12 | |
| FAIR-CalibModel=LLaDA-Base, Quantization=W4A42026.06 | 61.09 | |
| FlatQuantModel=LLaDA-Base, Quantization=W4A42026.06 | 59.37 | |
| QuaRotModel=LLaDA-Base, Quantization=W4A42026.06 | 58.78 | |
| RTNModel=LLaDA-Base, Quantization=W4A42026.06 | 49.13 |