Question Answering on NarrativeQA (Rouge-L)
45Rouge-LH2MT
Evaluation Results
| Method | Links | |
|---|---|---|
| H2MTModel=Qwen3-4B2026.05 | 45 | |
| H2MTModel=Llama3.1-8B2026.05 | 42.83 | |
| H2MTModel=Qwen2.5-14B2026.05 | 41.67 | |
| H2MTModel=Qwen2.5-7B2026.05 | 29.26 | |
| HMTModel=Qwen2.5-14B2026.05 | 29.17 | |
| H2MTModel=SmolLM-135M2026.05 | 28.57 | |
| BackboneModel=Llama3.1-8B2026.05 | 22.84 | |
| HMTModel=Llama3.1-8B2026.05 | 21.02 | |
| HMTModel=SmolLM-135M2026.05 | 16.44 | |
| HMTModel=Qwen2.5-7B2026.05 | 14.42 | |
| BackboneModel=Qwen3-4B2026.05 | 12.15 | |
| HMTModel=Qwen3-4B2026.05 | 10.01 | |
| BackboneModel=Qwen2.5-14B2026.05 | 9.96 | |
| BackboneModel=Qwen2.5-7B2026.05 | 9.46 | |
| BackboneModel=SmolLM-135M2026.05 | 1.04 |