Math on GSM8K 1,000-example (test)
5.8317PPLQwen3-VL-2B-Instruct
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen3-VL-2B-InstructAblation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 5.8317 | — | |
| Qwen3-VL-2B-InstructAblation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 5.8756 | 0.0439 | |
| Qwen3-VL-8B-InstructAblation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 6.5793 | -0.1754 | |
| Qwen3-VL-8B-InstructAblation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 6.7547 | — | |
| Qwen3-VL-4B-InstructAblation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 7.0319 | — | |
| Qwen3-VL-4B-InstructAblation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 7.0691 | 0.0372 | |
| Ministral-14B-Instruct-2512Ablation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 9.1063 | -0.2704 | |
| Ministral-14B-Instruct-2512Ablation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 9.3767 | — | |
| Ministral-3B-Instruct-2512Ablation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 16.1005 | -0.7698 | |
| Ministral-3B-Instruct-2512Ablation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.01 | 16.8703 | — |