Instruction Following on IFEval (Avg@k, Pass@k)
64.23Avg@kBase
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BaseBackbone=Qwen2.5-7B-Instruct2026.05 | 64.23 | 76.71 | |
| DFTBackbone=Qwen2.5-7B-Instruct2026.05 | 61.18 | 73.01 | |
| DFTBase Model=Qwen2.5-7B-Instruct, RL Stage=DAPO, k=42026.05 | 61.18 | 73.01 | |
| RPSFTBackbone=Qwen2.5-7B-Instruct2026.05 | 58.32 | 73.38 | |
| RPSFTBase Model=Qwen2.5-7B-Instruct, RL Stage=DAPO, k=42026.05 | 58.32 | 73.38 | |
| BaseBackbone=Qwen2.5-3B-Instruct2026.05 | 58.09 | 70.61 | |
| IWBackbone=Qwen2.5-7B-Instruct2026.05 | 55.91 | 69.5 | |
| IWBase Model=Qwen2.5-7B-Instruct, RL Stage=DAPO, k=42026.05 | 55.91 | 69.5 | |
| SFTBackbone=Qwen2.5-7B-Instruct2026.05 | 54.48 | 69.5 | |
| SFTBase Model=Qwen2.5-7B-Instruct, RL Stage=DAPO, k=42026.05 | 54.48 | 69.5 | |
| DFTBackbone=Qwen2.5-3B-Instruct2026.05 | 52.59 | 65.62 | |
| DFTBase Model=Qwen2.5-3B-Instruct, RL Stage=DAPO, k=42026.05 | 52.59 | 65.62 | |
| RPSFTBackbone=Qwen2.5-3B-Instruct2026.05 | 49.72 | 65.25 | |
| RPSFTBase Model=Qwen2.5-3B-Instruct, RL Stage=DAPO, k=42026.05 | 49.72 | 65.25 | |
| SFTBackbone=Qwen2.5-3B-Instruct2026.05 | 47.97 | 63.03 | |
| SFTBase Model=Qwen2.5-3B-Instruct, RL Stage=DAPO, k=42026.05 | 47.97 | 63.03 | |
| IWBackbone=Qwen2.5-3B-Instruct2026.05 | 47.92 | 62.85 | |
| IWBase Model=Qwen2.5-3B-Instruct, RL Stage=DAPO, k=42026.05 | 47.92 | 62.85 | |
| SFTBackbone=Llama-3.1-8B-Instruct2026.05 | 32.58 | 35.49 | |
| SFTBase Model=Llama-3.1-8B-Instruct, RL Stage=DAPO, k=42026.05 | 32.58 | 35.49 | |
| BaseBackbone=Llama-3.1-8B-Instruct2026.05 | 32.3 | 35.86 | |
| RPSFTBackbone=Llama-3.1-8B-Instruct2026.05 | 31.93 | 35.67 | |
| RPSFTBase Model=Llama-3.1-8B-Instruct, RL Stage=DAPO, k=42026.05 | 31.93 | 35.67 | |
| DFTBackbone=Llama-3.1-8B-Instruct2026.05 | 31.84 | 34.57 | |
| DFTBase Model=Llama-3.1-8B-Instruct, RL Stage=DAPO, k=42026.05 | 31.84 | 34.57 | |
| IWBackbone=Llama-3.1-8B-Instruct2026.05 | 31.7 | 35.3 | |
| IWBase Model=Llama-3.1-8B-Instruct, RL Stage=DAPO, k=42026.05 | 31.7 | 35.3 |