Codesota · Models · Aria7 results · 7 benchmarks
Model card
Aria.
unknown
§ 02 · Benchmarks
Every benchmark Aria has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | HumanEval+ | Computer Code · Code Generation | pass-1 | 73.2% | #6 | — | source ↗ |
| 02 | DocVQA | Computer Vision · Document Understanding | anls | 92.6% | #12 | — | source ↗ |
| 03 | TextVQA | Multimodal · Visual Question Answering | accuracy | 81.1% | #13 | — | source ↗ |
| 04 | Video-MME | Multimodal · Video Understanding | accuracy | 72.1% | #15 | — | source ↗ |
| 05 | MMMU | Multimodal · Image-Text-to-Text | accuracy | 54.9% | #28 | — | source ↗ |
| 06 | MATH | Reasoning · Mathematical Reasoning | accuracy | 50.8% | #42 | — | source ↗ |
| 07 | MMLU | Reasoning · Commonsense Reasoning | accuracy | 73.3% | #53 | — | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Strengths by area
Where Aria actually performs.
§ 04 · Papers
1 paper with results for Aria.
- 2024-10-08· 7 results
Aria: An Open Multimodal Native Mixture-of-Experts Model
§ 06 · Sources & freshness
Where these numbers come from.
pwc-dump
7
results
0 of 7 rows marked verified.