Natural Language Inference on GLUE (test)
93.15MNLI AccFLoRG
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| FLoRGBase Model=Llama-3.2-3B2026.02 | 93.15 | 93.12 | 68.73 | 73.84 | |
| FedEx-LoRABase Model=Llama-3.2-3B2026.02 | 92.74 | 93.05 | 67.89 | 73.15 | |
| FedSA-LoRABase Model=Llama-3.2-3B2026.02 | 92.38 | 93.27 | 66.81 | 72.56 | |
| FLoRGBase Model=RoBERTa-large2026.02 | 91.27 | 92.58 | 66.48 | 71.26 | |
| FFA-LoRABase Model=Llama-3.2-3B2026.02 | 91.05 | 92.53 | 65.97 | 71.33 | |
| FedEx-LoRABase Model=RoBERTa-large2026.02 | 90.96 | 92.13 | 66.11 | 70.98 | |
| FedSA-LoRABase Model=RoBERTa-large2026.02 | 90.75 | 91.54 | 63.47 | 69.93 | |
| FeDeRABase Model=Llama-3.2-3B2026.02 | 89.83 | 91.45 | 64.19 | 69.75 | |
| FFA-LoRABase Model=RoBERTa-large2026.02 | 89.28 | 90.96 | 63.72 | 68.49 | |
| FeDeRABase Model=RoBERTa-large2026.02 | 88.06 | 90.07 | 61.83 | 67.12 | |
| FLoRGBase Model=OPT-125M2026.02 | 87.35 | 89.52 | 65.28 | 68.92 | |
| FedITBase Model=Llama-3.2-3B2026.02 | 87.24 | 89.17 | 61.52 | 67.08 | |
| FedEx-LoRABase Model=OPT-125M2026.02 | 85.83 | 89.88 | 64.15 | 68.27 | |
| FedITBase Model=RoBERTa-large2026.02 | 84.91 | 87.49 | 59.34 | 64.25 | |
| FedSA-LoRABase Model=OPT-125M2026.02 | 84.61 | 88.69 | 62.83 | 67.39 | |
| FFA-LoRABase Model=OPT-125M2026.02 | 83.54 | 87.93 | 62.61 | 66.02 | |
| FeDeRABase Model=OPT-125M2026.02 | 81.15 | 86.71 | 59.34 | 64.51 | |
| FedITBase Model=OPT-125M2026.02 | 79.42 | 84.18 | 58.45 | 61.08 |