Language Understanding and Reasoning on ARC-C, ARC-E, HellaSwag, LAMBADA, PIQA, WinoGrande (test)
56.81Mean ScoreOLMo-1B
Evaluation Results
| Method | Links | |
|---|---|---|
| OLMo-1BParams=1B, Reported tokens=3T, Hardware in public card=not listed in HF card, Evaluation mode=Zero-shot2026.06 | 56.81 | |
| Qwen2.5-0.5BParams=0.49B, Reported tokens=not listed in card, Hardware in public card=not listed in card, Evaluation mode=Zero-shot2026.06 | 53.63 | |
| SmolLM2-135MParams=135M, Reported tokens=2T, Hardware in public card=64x H100, Evaluation mode=Zero-shot2026.06 | 49.17 | |
| SmolLM-135MParams=135M, Reported tokens=600B, Hardware in public card=64x H100, Evaluation mode=Zero-shot2026.06 | 47.67 | |
| L20-Edu-135M finalParams=134.5M, Reported tokens=13B, Hardware in public card=1x NVIDIA L20, Evaluation mode=Zero-shot2026.06 | 41.5 |