Multi-task Language Proficiency on English Ability Average
44.9Average ScoreMistral-7B-Instruct
Evaluation results
| Method | Links | |
|---|---|---|
| Mistral-7B-InstructTraining Strategy=Two-phase Continual Fine-tuning (CFT), Phase 2 (P2) Dataset=MULTI, Evaluation Phase=P12024.10 | 44.9 | |
| Llama-3-8B-InstructTraining Strategy=Two-phase Continual Fine-tuning (CFT), Evaluation Phase=P12024.10 | 37.2 | |
| Mistral-7BTraining Strategy=Two-phase Continual Fine-tuning (CFT), Phase 1 (P1) Dataset=ALPACA, Evaluation Phase=P22024.10 | 37.1 | |
| Mistral-7BTraining Strategy=Dataset Mixture, Dataset Mixture=ALPACA MULTIALPACA2024.10 | 37.1 | |
| Mistral-7BTraining Strategy=Two-phase Continual Fine-tuning (CFT), Phase 1 (P1) Dataset=ALPACA, Evaluation Phase=P12024.10 | 34.3 | |
| Llama-3-8BTraining Strategy=Dataset Mixture, Dataset Mixture=ALPACA MULTIALPACA2024.10 | 33.5 | |
| Mistral-7B-InstructTraining Strategy=Two-phase Continual Fine-tuning (CFT), Phase 2 (P2) Dataset=MULTI, Evaluation Phase=P22024.10 | 33.2 | |
| Llama-3-8BTraining Strategy=Two-phase Continual Fine-tuning (CFT), Phase 1 (P1) Dataset=ALPACA, Phase 2 (P2) Dataset=ALPACA, Evaluation Phase=P22024.10 | 26.5 | |
| Llama-3-8BTraining Strategy=Two-phase Continual Fine-tuning (CFT), Phase 1 (P1) Dataset=ALPACA, Phase 2 (P2) Dataset=ALPACA, Evaluation Phase=P12024.10 | 24.7 | |
| Llama-3-8B-InstructTraining Strategy=Two-phase Continual Fine-tuning (CFT), Evaluation Phase=P22024.10 | 17.8 |