Human Preference Evaluation for Code-switched Text Generation on In-Domain Data
804Preference ScoreGold Standard
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Gold Standard2025.02 | 804 | 1 | |
| Llama3training=fine-tuned2025.02 | 573.5 | 2 | |
| NLLBtraining=fine-tuned, backbone=nllb-200-distilled-600M2025.02 | 507 | 3 | |
| Llama3 Instructtraining=fine-tuned2025.02 | 480.5 | 4 | |
| GPT-4ofsmode=5-shot2025.02 | 413.5 | 5 | |
| Llama3.3-70Bfsmode=5-shot2025.02 | 371.5 | 6 | |
| Gold Standard2025.02 | 369.5 | 1 | |
| Llama3training=fine-tuned2025.02 | 291.5 | 2 | |
| Llama3 Instructtraining=fine-tuned2025.02 | 270.5 | 3 | |
| NLLBtraining=fine-tuned, backbone=nllb-200-distilled-600M2025.02 | 259.5 | 4 | |
| GPT-4ofsmode=5-shot2025.02 | 251.5 | 5 | |
| Llama3.3-70Bfsmode=5-shot2025.02 | 207.5 | 6 |