Codesota · Models · AIMv2 ViT-3B/14 + Llama 3.0 8B4 results · 4 benchmarks
Model card
AIMv2 ViT-3B/14 + Llama 3.0 8B.
unknown
§ 02 · Benchmarks
Every benchmark AIMv2 ViT-3B/14 + Llama 3.0 8B has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | GQA | Multimodal · Visual Question Answering | accuracy | 73.3% | #1 | — | source ↗ |
| 02 | VQA v2.0 | Multimodal · Visual Question Answering | accuracy | 80.9% | #7 | — | source ↗ |
| 03 | DocVQA | Computer Vision · Document Understanding | anls | 30.4% | #21 | — | source ↗ |
| 04 | TextVQA | Multimodal · Visual Question Answering | accuracy | 58.2% | #21 | — | 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 AIMv2 ViT-3B/14 + Llama 3.0 8B actually performs.
§ 04 · Papers
1 paper with results for AIMv2 ViT-3B/14 + Llama 3.0 8B.
- 2024-11-21· 4 results
Multimodal Autoregressive Pre-training of Large Vision Encoders
§ 06 · Sources & freshness
Where these numbers come from.
pwc-dump
4
results
0 of 4 rows marked verified.