Mathematical Reasoning on AIME 24 (Pass@256)
63.3Pass@256Qwen2.5-7B + GRPO w/ VERL.
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-7B + GRPO w/ VERL.Base Model=Qwen2.5-7B, RL Framework=GRPO, VERL Integration=true, Decoding Setting=Pass@2562025.09 | 63.3 | |
| Qwen2.5-7BBase Model=Qwen2.5-7B, RL Framework=None, Decoding Setting=Pass@2562025.09 | 60 | |
| Qwen2.5-7B + GRPOBase Model=Qwen2.5-7B, RL Framework=GRPO, Decoding Setting=Pass@2562025.09 | 56.7 | |
| Qwen2.5-7B + PPO w/ VERL.Base Model=Qwen2.5-7B, RL Framework=PPO, VERL Integration=true, Decoding Setting=Pass@2562025.09 | 56.7 | |
| Qwen2.5-7B + PPOBase Model=Qwen2.5-7B, RL Framework=PPO, Decoding Setting=Pass@2562025.09 | 53.3 | |
| Llama-3.2-3B-Instruct + GRPO w/ VERL.Base Model=Llama-3.2-3B-Instruct, RL Framework=GRPO, VERL Integration=true, Decoding Setting=Pass@2562025.09 | 50 | |
| Llama-3.2-3B-Instruct + PPOBase Model=Llama-3.2-3B-Instruct, RL Framework=PPO, Decoding Setting=Pass@2562025.09 | 46.7 | |
| Llama-3.2-3B-Instruct + PPO w/ VERL.Base Model=Llama-3.2-3B-Instruct, RL Framework=PPO, VERL Integration=true, Decoding Setting=Pass@2562025.09 | 46.7 | |
| Llama-3.2-3B-InstructBase Model=Llama-3.2-3B-Instruct, RL Framework=None, Decoding Setting=Pass@2562025.09 | 40 | |
| Llama-3.2-3B-Instruct + GRPOBase Model=Llama-3.2-3B-Instruct, RL Framework=GRPO, Decoding Setting=Pass@2562025.09 | 40 |