Language Tutoring on AFRILANGTUTOR
83.7Amharic PerformanceAFRILANGTUTOR (Llama-3-8B-IT)
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AFRILANGTUTOR (Llama-3-8B-IT)Judge Criteria=Instr. Align.2026.04 | 83.7 | 85.5 | 87 | 87.8 | 84.4 | 86.9 | 87.5 | 82.9 | 86.1 | 86.8 | 85.9 | |
| AFRILANGTUTOR (Llama-3-8B-IT)Judge Criteria=Coher. Natu.2026.04 | 78.1 | 80.2 | 79.4 | 79.8 | 80.8 | 78.4 | 78.4 | 78.7 | 78 | 79.5 | 79.1 | |
| AFRILANGTUTOR (Gemma-3-12B-IT)Judge Criteria=Instr. Align.2026.04 | 74.5 | 79.3 | 79.8 | 79.7 | 77.4 | 75.6 | 78.6 | 78 | 79.1 | 80.6 | 78.3 | |
| Gemma-3-12B-ITJudge Criteria=Instr. Align.2026.04 | 69.7 | 71.8 | 68.2 | 71.2 | 68.8 | 69.6 | 71.6 | 69 | 72.1 | 73.2 | 70.5 | |
| Llama-3-8B-ITJudge Criteria=Instr. Align.2026.04 | 68.2 | 70.8 | 72.3 | 74.6 | 72.1 | 71.1 | 73.3 | 69.7 | 71.8 | 72.2 | 71.6 | |
| AFRILANGTUTOR (Gemma-3-12B-IT)Judge Criteria=Pedag. Compl.2026.04 | 66.7 | 69.1 | 68.8 | 68.5 | 67.2 | 67 | 67.2 | 68.1 | 67 | 69.4 | 67.9 | |
| AFRILANGTUTOR (Llama-3-8B-IT)Judge Criteria=Pedag. Compl.2026.04 | 63.8 | 66 | 67.8 | 68.8 | 67.3 | 65.7 | 66 | 65.1 | 66.8 | 65.9 | 66.3 | |
| Llama-3-8B-ITJudge Criteria=Coher. Natu.2026.04 | 62.4 | 70.8 | 68.5 | 71.5 | 70.6 | 67.5 | 70.3 | 63.6 | 69.2 | 71.6 | 68.6 | |
| Gemma-3-12B-ITJudge Criteria=Coher. Natu.2026.04 | 61.2 | 65.1 | 63.7 | 63.3 | 60.9 | 62 | 64.5 | 60.4 | 64.9 | 66.2 | 63.2 | |
| AFRILANGTUTOR (Gemma-3-12B-IT)Judge Criteria=Coher. Natu.2026.04 | 59.8 | 69.3 | 72.3 | 68.7 | 66.7 | 61.4 | 69.4 | 65.8 | 71.8 | 70.7 | 67.6 | |
| Gemma-3-12B-ITJudge Criteria=Pedag. Compl.2026.04 | 57.4 | 57.9 | 56.4 | 58.8 | 56.9 | 57 | 58.3 | 56.8 | 59 | 60.9 | 57.9 | |
| AFRILANGTUTOR (Gemma-3-12B-IT)Judge Criteria=Ling. Cult. Acc.2026.04 | 51.7 | 50 | 52.2 | 50.2 | 45.1 | 51.5 | 45.7 | 50.8 | 48.5 | 52.2 | 49.8 | |
| Gemma-3-12B-ITJudge Criteria=Ling. Cult. Acc.2026.04 | 49.9 | 45.3 | 44.4 | 49 | 47.4 | 49 | 45.5 | 50.2 | 47.2 | 52.5 | 48 | |
| Llama-3-8B-ITJudge Criteria=Pedag. Compl.2026.04 | 48.8 | 51.4 | 52.2 | 51.6 | 48.5 | 49.3 | 53.2 | 50 | 51.7 | 50.8 | 50.8 | |
| AFRILANGTUTOR (Llama-3-8B-IT)Judge Criteria=Ling. Cult. Acc.2026.04 | 48.5 | 48.7 | 49.1 | 53 | 54 | 46.4 | 45.7 | 51.1 | 43.8 | 46 | 48.6 | |
| Llama-3-8B-ITJudge Criteria=Ling. Cult. Acc.2026.04 | 32.7 | 47.7 | 41.5 | 45.2 | 46.8 | 40.5 | 47.3 | 33.3 | 45.2 | 47.7 | 42.8 |