Open-book Question Answering on OpenBookQA (Score)
43.36ScoreDCLM
Evaluation results
| Method | Links | |
|---|---|---|
| DCLMPre-training Curation Strategy=Human Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 43.36 | |
| Nemotron-CC_ASIPre-training Curation Strategy=AI Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 42.2 | |
| Nemotron-CC_ASI+Pre-training Curation Strategy=AI Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 41.44 | |
| Fineweb-EduPre-training Curation Strategy=Human Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 39.92 | |
| Ultra-FinewebPre-training Curation Strategy=Human Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 39.84 | |
| Nemotron-CCPre-training Curation Strategy=AI Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 39.8 |