Autoformalization on FLC (val)
0.64Cross-Entropy LossSFT No-Curriculum 2B
Evaluation results
| Method | Links | |
|---|---|---|
| SFT No-Curriculum 2BModel Scale=2B, Training Strategy=SFT No-Curriculum2026.03 | 0.64 | |
| SFT Curriculum 2BModel Scale=2B, Training Strategy=SFT Curriculum2026.03 | 0.68 | |
| RL (GRPO) 2BModel Scale=2B, Training Strategy=RL (GRPO)2026.03 | 0.691 |