metadata
license: apache-2.0
arco consistently outperforms every sota model below 600m parameters on average, outperforms some base 1b models and is competitive with the best ones. arco is a merge of multiple internal models fine-tuned on a diverse set of styles and finally merged with the several models (including palmer-004-turbo), followed by a merge with base model to preserve knowledge.
benchmarks
zero-shot evaluations performed on current sota ~0.5b models.
| Parameters | Model | MMLU | ARC-C | HellaSwag | PIQA | Winogrande | Average |
|---|---|---|---|---|---|---|---|
| 0.5b | qwen2 | 44.13 | 28.92 | 49.05 | 69.31 | 56.99 | 49.68 |
| 0.5b | danube3-base | 24.81 | 36.18 | 60.46 | 73.78 | 61.01 | 51.25 |
| 0.5b | palmer-turbo | 27.36 | 35.58 | 61.79 | 73.67 | 61.17 | 51.91 |
| 0.5b | arco | 26.17 | 37.29 | 62.88 | 74.37 | 62.27 | 52.60 |
supporters
