valley2_dpo: benchmark results
Provider: ByteDance. Released 2025-02-09. Access: Open.
Aggregate Rating 1542 ± 29, rank #1816 of 10916 rated models, from 148 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMBench V1.1 CN - Function Reasoning | 96.7 | Accuracy (%) | 98.9 |
| OpenVLM MMMU - Architecture and Engineering | 56.7 | Accuracy (%) | 95.4 |
| OpenVLM MMMU - Electronics | 83.3 | Accuracy (%) | 95.2 |
| OpenVLM MMBench V1.1 CN - AR | 90.2 | Accuracy (%) | 93.5 |
| OpenVLM MMBench V1.1 EN - Nature Relation | 92.4 | Accuracy (%) | 91.2 |
| OpenVLM MMMU - Music | 43.3 | Accuracy (%) | 91.2 |
| OpenVLM MMBench V1.1 CN - Image Topic | 98.9 | Accuracy (%) | 90.3 |
| OpenVLM AI2D - Parts of the Earth | 84.6 | Accuracy (%) | 89.8 |
| OpenVLM MMBench V1.1 EN - Future Prediction | 82.7 | Accuracy (%) | 89.6 |
| OpenVLM MMBench V1.1 EN - Function Reasoning | 95.6 | Accuracy (%) | 88.7 |
| OpenVLM MMBench V1.1 CN - Image Emotion | 84.4 | Accuracy (%) | 88.3 |
| OpenVLM MMBench V1.1 CN - Future Prediction | 82.7 | Accuracy (%) | 88 |
Interactive version: theaggregate.ai/model?slug=valley2-dpo · How It Works · Data refreshed daily, snapshot 2026-11-08.