Long Context Understanding On Longbench
Métriques
Average Score
Résultats
Résultats de performance de divers modèles sur ce benchmark
Nom du modèle | Average Score | Paper Title | Repository |
|---|---|---|---|
| GALI(Llama3-8b-ins-8k-to-16k) | 45.17 | A Training-Free Length Extrapolation Approach for LLMs: Greedy Attention Logit Interpolation (GALI) | |
| GALI(Llama3-8b-ins-4k-to-16k) | 46.22 | A Training-Free Length Extrapolation Approach for LLMs: Greedy Attention Logit Interpolation (GALI) | |
| GALI(Llama3-8b-ins-8k-to-32k) | 45.38 | A Training-Free Length Extrapolation Approach for LLMs: Greedy Attention Logit Interpolation (GALI) |
0 of 3 row(s) selected.