Free-form generation on FREE-FORM GENERATION ablation
48Rouge-LGPT-3.5-TURBO
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-3.5-TURBOInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 48 | |
| GPT-4Inference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 47 | |
| Tulu-65BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 39 | |
| GPT-4Inference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 39 | |
| GPT-4Inference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 39 | |
| GPT-3.5-TURBOInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 38 | |
| GPT-4Inference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 38 | |
| Tulu-30BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 37 | |
| GPT-3.5-TURBOInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 37 | |
| LLAMA-2-CHAT-70BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 35 | |
| GPT-3.5-TURBOInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 31 | |
| Tulu-30BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 30 | |
| LLAMA-2-CHAT-70BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 27 | |
| LLAMA-2-CHAT-70BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 26 | |
| LLAMA-2-CHAT-70BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 24 | |
| VICUNA-13BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 23 | |
| VICUNA-13BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 22 | |
| Tulu-13BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 21 | |
| VICUNA-7BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 20 | |
| Tulu-7BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 19 | |
| Tulu-65BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 19 | |
| VICUNA-13BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 18 | |
| Tulu-13BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 17 | |
| Tulu-30BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 17 | |
| Tulu-7BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 16 | |
| LLAMA-2-CHAT-13BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 16 | |
| VICUNA-7BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 16 | |
| VICUNA-13BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 16 | |
| Tulu-13BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 15 | |
| Tulu-30BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 15 | |
| Tulu-65BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 15 | |
| LLAMA-2-CHAT-13BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 15 | |
| LLAMA-2-CHAT-7BInference Mode=SINGLE-TASK, Sub-task Position=1ST, Shots=12024.02 | 14 | |
| VICUNA-7BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 14 | |
| LLAMA-2-CHAT-13BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 13 | |
| Tulu-7BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 12 | |
| Tulu-13BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 12 | |
| Tulu-65BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 11 | |
| LLAMA-2-CHAT-7BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 11 | |
| Tulu-7BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 10 | |
| LLAMA-2-CHAT-7BInference Mode=SINGLE-TASK, Sub-task Position=FINAL, Shots=12024.02 | 10 | |
| VICUNA-7BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 9 | |
| LLAMA-2-CHAT-7BInference Mode=MULTI-TASK, Sub-task Position=FINAL, Shots=12024.02 | 8 | |
| LLAMA-2-CHAT-13BInference Mode=MULTI-TASK, Sub-task Position=1ST, Shots=12024.02 | 7 |