# EVA

**类别：** ViT · **代码：** [`timm/models/eva.py`](https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/eva.py) · **架构：** 65 · **预训练权重：** 83（其中 28 个包含 ImageNet-1k 结果）

EVA 和 EVA-02 视觉 Transformer 以及其他共享 `Eva` 类的 ViT 变体，支持旋转位置编码（RoPE）、SwiGLU MLP 及相关扩展。该模块还包含 Perception Encoder (PE)、ROPE-ViT、DINOv3、LingBot-Vision 和 Sapiens2 模型，以及带有 RoPE 的 `timm` 原始 ViT 配置。

## 论文

* [EVA：探索大规模掩码视觉表示学习的极限](https://arxiv.org/abs/2211.07636)
* [EVA-02：《新世纪福音战士》的视觉呈现](https://arxiv.org/abs/2303.11331)
* [EVA-CLIP：大规模 CLIP 的改进训练技术](https://arxiv.org/abs/2303.15389)
* [Perception Encoder：最佳的视觉嵌入并不在网络的输出端](https://arxiv.org/abs/2504.13181)
* [用于视觉 Transformer 的旋转位置嵌入](https://arxiv.org/abs/2403.13298)
* [DINOv3](https://arxiv.org/abs/2508.10104)
* [用于稠密空间感知的视觉预训练](https://arxiv.org/abs/2607.05247)
* [Sapiens2](https://arxiv.org/abs/2604.21681)

## 用法

请参阅[快速入门](../quickstart)了解如何加载预训练权重、进行图像预处理和微调，以及参阅[特征提取](../feature_extraction)了解如何将这些模型用作骨干网络。

```py
>>> import timm
>>> model = timm.create_model('eva02_small_patch14_336.mim_in22k_ft_in1k', pretrained=True)
>>> data_config = timm.data.resolve_model_data_config(model)
```

## 模型 API 支持

| 功能 | `Eva` |
|---|---|
| [`features_only=True` 特征图](../model_api#intermediate-features) | ✓（strides 16, 16, 16） |
| [`forward_intermediates()`](../model_api#intermediate-features) | ✓ |
| [`prune_intermediate_layers()`](../model_api#intermediate-features) | ✓ |
| [任意输入尺寸，无需更改](../model_api#input-size) | ✓ |
| [`set_input_size()`](../model_api#input-size) | ✓ |
| [`dynamic_img_size=True`](../model_api#input-size) | ✓ |
| [`set_grad_checkpointing()`](../model_api#training-hooks) | ✓ |
| [逐层学习率衰减（`group_matcher()`）](../model_api#training-hooks) | ✓ |
| [`reset_classifier()`](../model_api#classifier) | ✓ |

✓ 支持，✓ (square) 仅支持方形尺寸，✗ 不支持，– 无需支持（模型本身已接受任意输入尺寸）。通过在每个功能上运行随机初始化的模型（`Eva` 使用 `vit_tiny_patch16_dinov3_qkvb`，尺寸为 256）进行验证，预期同一类的其他架构表现相同。详情请参阅 [Model API](../model_api) 页面。

## 架构

| 架构 | 参数（M） | GMAC | 输入大小 | 预训练权重 |
|---|---:|---:|---:|---:|
| `eva02_base_patch14_224` | 85.76 | 23.22 | 224 | 1 |
| `eva02_base_patch14_448` | 87.12 | 107.11 | 448 | 3 |
| `eva02_base_patch16_clip_224` | 86.26 | 17.62 | 224 | 1 |
| `eva02_enormous_patch14_clip_224` | 4350.56 | 1132.46 | 224 | 2 |
| `eva02_large_patch14_224` | 303.27 | 81.15 | 224 | 2 |
| `eva02_large_patch14_448` | 305.08 | 362.33 | 448 | 6 |
| `eva02_large_patch14_clip_224` | 304.11 | 81.18 | 224 | 1 |
| `eva02_large_patch14_clip_336` | 304.43 | 191.34 | 336 | 1 |
| `eva02_small_patch14_224` | 21.62 | 6.14 | 224 | 1 |
| `eva02_small_patch14_336` | 22.13 | 15.48 | 336 | 1 |
| `eva02_tiny_patch14_224` | 5.50 | 1.70 | 224 | 1 |
| `eva02_tiny_patch14_336` | 5.76 | 4.68 | 336 | 1 |
| `eva_giant_patch14_224` | 1012.56 | 267.18 | 224 | 1 |
| `eva_giant_patch14_336` | 1013.01 | 620.64 | 336 | 2 |
| `eva_giant_patch14_560` | 1014.45 | 1906.76 | 560 | 1 |
| `eva_giant_patch14_clip_224` | 1012.59 | 267.18 | 224 | 2 |
| `vit_5b_patch16_sapiens2` | 5066.90 | - | 1024x768 | 1 |
| `vit_7b_patch16_dinov3` | 6716.03 | 1775.10 | 256 | 2 |
| `vit_base_patch16_dinov3` | 85.64 | 23.60 | 256 | 1 |
| `vit_base_patch16_dinov3_qkvb` | 85.66 | 23.60 | 256 | 2 |
| `vit_base_patch16_lingbot` | 85.66 | - | 512 | 1 |
| `vit_base_patch16_rope_224` | 86.43 | 17.58 | 224 | 1 |
| `vit_base_patch16_rope_ape_224` | 86.59 | 17.58 | 224 | 1 |
| `vit_base_patch16_rope_mixed_224` | 86.44 | 17.58 | 224 | 1 |
| `vit_base_patch16_rope_mixed_ape_224` | 86.59 | 17.58 | 224 | 1 |
| `vit_base_patch16_rope_reg1_gap_256` | 86.43 | 23.22 | 256 | 1 |
| `vit_base_patch16_sapiens2` | 113.99 | - | 1024x768 | 1 |
| `vit_betwixt_patch16_rope_reg4_gap_256` | 60.23 | 16.52 | 256 | 1 |
| `vit_giant_patch16_lingbot` | 1134.47 | - | 512 | 1 |
| `vit_giant_patch16_sapiens2` | 1455.47 | - | 1024x768 | 1 |
| `vit_huge_patch16_sapiens2` | 814.28 | - | 1024x768 | 1 |
| `vit_huge_plus_patch16_dinov3` | 840.51 | 224.88 | 256 | 1 |
| `vit_huge_plus_patch16_dinov3_qkvb` | 840.59 | 224.88 | 256 | 1 |
| `vit_large_patch16_dinov3` | 303.08 | 82.43 | 256 | 2 |
| `vit_large_patch16_dinov3_qkvb` | 303.13 | 82.43 | 256 | 2 |
| `vit_large_patch16_lingbot` | 303.13 | - | 512 | 1 |
| `vit_large_patch16_rope_224` | 304.17 | 61.60 | 224 | 1 |
| `vit_large_patch16_rope_ape_224` | 304.37 | 61.60 | 224 | 1 |
| `vit_large_patch16_rope_mixed_224` | 304.20 | 61.60 | 224 | 1 |
| `vit_large_patch16_rope_mixed_ape_224` | 304.40 | 61.60 | 224 | 1 |
| `vit_large_patch16_sapiens2` | 395.45 | - | 1024x768 | 1 |
| `vit_medium_patch16_rope_reg1_gap_256` | 38.74 | 10.63 | 256 | 1 |
| `vit_mediumd_patch16_rope_reg1_gap_256` | 63.95 | 17.65 | 256 | 1 |
| `vit_pe_core_base_patch16_224` | 93.67 | 17.82 | 224 | 1 |
| `vit_pe_core_gigantic_patch14_448` | 1882.03 | 2060.12 | 448 | 1 |
| `vit_pe_core_large_patch14_336` | 317.15 | 192.33 | 336 | 1 |
| `vit_pe_core_small_patch16_384` | 23.78 | 15.69 | 384 | 1 |
| `vit_pe_core_tiny_patch16_384` | 6.14 | 4.74 | 384 | 1 |
| `vit_pe_lang_gigantic_patch14_448` | 1740.92 | 1931.99 | 448 | 2 |
| `vit_pe_lang_large_patch14_448` | 291.42 | 346.99 | 448 | 2 |
| `vit_pe_spatial_base_patch16_512` | 86.43 | 107.13 | 512 | 1 |
| `vit_pe_spatial_gigantic_patch14_448` | 1851.89 | 2055.25 | 448 | 1 |
| `vit_pe_spatial_large_patch14_448` | 303.96 | 362.05 | 448 | 1 |
| `vit_pe_spatial_small_patch16_512` | 21.98 | 31.80 | 512 | 1 |
| `vit_pe_spatial_tiny_patch16_512` | 5.68 | 10.46 | 512 | 1 |
| `vit_small_patch16_dinov3` | 21.59 | 6.26 | 256 | 1 |
| `vit_small_patch16_dinov3_qkvb` | 21.60 | 6.26 | 256 | 2 |
| `vit_small_patch16_lingbot` | 21.60 | - | 512 | 1 |
| `vit_small_patch16_rope_224` | 21.98 | 4.61 | 224 | 1 |
| `vit_small_patch16_rope_ape_224` | 22.06 | 4.61 | 224 | 1 |
| `vit_small_patch16_rope_mixed_224` | 21.99 | 4.61 | 224 | 1 |
| `vit_small_patch16_rope_mixed_ape_224` | 22.06 | 4.61 | 224 | 1 |
| `vit_small_plus_patch16_dinov3` | 28.68 | 8.11 | 256 | 1 |
| `vit_small_plus_patch16_dinov3_qkvb` | 28.69 | 8.11 | 256 | 1 |
| `vit_tiny_patch16_dinov3_qkvb` | 5.49 | - | 256 | 1 |

GMACs 和参数量对应于每种架构的默认输入大小和分类器，数据来自[基准测试结果](https://github.com/huggingface/pytorch-image-models/tree/main/results)。

## 预训练权重

Top-1 是在所列评估规模下，来自 [`results-imagenet.csv`](https://github.com/huggingface/pytorch-image-models/blob/main/results/results-imagenet.csv) 的 ImageNet-1k 验证准确率，- 表示该权重没有 ImageNet-1k 结果（例如预训练权重或图文权重）。Classes 是预训练分类器的大小，对于没有分类器的权重则为无。

权重按来源分组：预训练配置中给出的论文或来源，否则为 `timm` 以外的 Hub 组织，否则为预训练标签的第一部分（`architecture.tag`）。仅包含单个权重的来源将列在“其他”下。

### `mim` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `eva02_base_patch14_224.mim_in22k` | 224 | - | - | none | mit | [timm](https://huggingface.co/timm/eva02_base_patch14_224.mim_in22k) |
| `eva02_base_patch14_448.mim_in22k_ft_in1k` | 448 | 88.262 | 448 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_base_patch14_448.mim_in22k_ft_in1k) |
| `eva02_base_patch14_448.mim_in22k_ft_in22k` | 448 | - | - | 21841 | mit | [timm](https://huggingface.co/timm/eva02_base_patch14_448.mim_in22k_ft_in22k) |
| `eva02_base_patch14_448.mim_in22k_ft_in22k_in1k` | 448 | 88.678 | 448 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_base_patch14_448.mim_in22k_ft_in22k_in1k) |
| `eva02_large_patch14_224.mim_in22k` | 224 | - | - | none | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_224.mim_in22k) |
| `eva02_large_patch14_224.mim_m38m` | 224 | - | - | none | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_224.mim_m38m) |
| `eva02_large_patch14_448.mim_in22k_ft_in1k` | 448 | 89.634 | 448 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_448.mim_in22k_ft_in1k) |
| `eva02_large_patch14_448.mim_in22k_ft_in22k` | 448 | - | - | 21841 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_448.mim_in22k_ft_in22k) |
| `eva02_large_patch14_448.mim_in22k_ft_in22k_in1k` | 448 | 89.956 | 448 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_448.mim_in22k_ft_in22k_in1k) |
| `eva02_large_patch14_448.mim_m38m_ft_in1k` | 448 | 89.550 | 448 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_448.mim_m38m_ft_in1k) |
| `eva02_large_patch14_448.mim_m38m_ft_in22k` | 448 | - | - | 21841 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_448.mim_m38m_ft_in22k) |
| `eva02_large_patch14_448.mim_m38m_ft_in22k_in1k` | 448 | 90.056 | 448 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_448.mim_m38m_ft_in22k_in1k) |
| `eva02_small_patch14_224.mim_in22k` | 224 | - | - | none | mit | [timm](https://huggingface.co/timm/eva02_small_patch14_224.mim_in22k) |
| `eva02_small_patch14_336.mim_in22k_ft_in1k` | 336 | 85.770 | 336 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_small_patch14_336.mim_in22k_ft_in1k) |
| `eva02_tiny_patch14_224.mim_in22k` | 224 | - | - | none | mit | [timm](https://huggingface.co/timm/eva02_tiny_patch14_224.mim_in22k) |
| `eva02_tiny_patch14_336.mim_in22k_ft_in1k` | 336 | 80.648 | 336 | 1000 | mit | [timm](https://huggingface.co/timm/eva02_tiny_patch14_336.mim_in22k_ft_in1k) |

### `fb` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_pe_core_base_patch16_224.fb` | 224 | - | - | 1024 | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_core_base_patch16_224.fb) |
| `vit_pe_core_gigantic_patch14_448.fb` | 448 | - | - | 1280 | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_core_gigantic_patch14_448.fb) |
| `vit_pe_core_large_patch14_336.fb` | 336 | - | - | 1024 | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_core_large_patch14_336.fb) |
| `vit_pe_core_small_patch16_384.fb` | 384 | - | - | 512 | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_core_small_patch16_384.fb) |
| `vit_pe_core_tiny_patch16_384.fb` | 384 | - | - | 512 | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_core_tiny_patch16_384.fb) |
| `vit_pe_lang_gigantic_patch14_448.fb` | 448 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_lang_gigantic_patch14_448.fb) |
| `vit_pe_lang_gigantic_patch14_448.fb_tiling` | 448 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_lang_gigantic_patch14_448.fb_tiling) |
| `vit_pe_lang_large_patch14_448.fb` | 448 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_lang_large_patch14_448.fb) |
| `vit_pe_lang_large_patch14_448.fb_tiling` | 448 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_lang_large_patch14_448.fb_tiling) |
| `vit_pe_spatial_base_patch16_512.fb` | 512 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_spatial_base_patch16_512.fb) |
| `vit_pe_spatial_gigantic_patch14_448.fb` | 448 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_spatial_gigantic_patch14_448.fb) |
| `vit_pe_spatial_large_patch14_448.fb` | 448 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_spatial_large_patch14_448.fb) |
| `vit_pe_spatial_small_patch16_512.fb` | 512 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_spatial_small_patch16_512.fb) |
| `vit_pe_spatial_tiny_patch16_512.fb` | 512 | - | - | none | apache-2.0 | [timm](https://huggingface.co/timm/vit_pe_spatial_tiny_patch16_512.fb) |

### `naver` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_base_patch16_rope_224.naver_in1k` | 224 | 83.714 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_base_patch16_rope_224.naver_in1k) |
| `vit_base_patch16_rope_ape_224.naver_in1k` | 224 | 83.780 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_base_patch16_rope_ape_224.naver_in1k) |
| `vit_base_patch16_rope_mixed_224.naver_in1k` | 224 | 83.806 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_base_patch16_rope_mixed_224.naver_in1k) |
| `vit_base_patch16_rope_mixed_ape_224.naver_in1k` | 224 | 83.898 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_base_patch16_rope_mixed_ape_224.naver_in1k) |
| `vit_large_patch16_rope_224.naver_in1k` | 224 | 84.646 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_large_patch16_rope_224.naver_in1k) |
| `vit_large_patch16_rope_ape_224.naver_in1k` | 224 | 84.660 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_large_patch16_rope_ape_224.naver_in1k) |
| `vit_large_patch16_rope_mixed_224.naver_in1k` | 224 | 84.840 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_large_patch16_rope_mixed_224.naver_in1k) |
| `vit_large_patch16_rope_mixed_ape_224.naver_in1k` | 224 | 84.846 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_large_patch16_rope_mixed_ape_224.naver_in1k) |
| `vit_small_patch16_rope_224.naver_in1k` | 224 | 81.222 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_small_patch16_rope_224.naver_in1k) |
| `vit_small_patch16_rope_ape_224.naver_in1k` | 224 | 81.016 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_small_patch16_rope_ape_224.naver_in1k) |
| `vit_small_patch16_rope_mixed_224.naver_in1k` | 224 | 81.252 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_small_patch16_rope_mixed_224.naver_in1k) |
| `vit_small_patch16_rope_mixed_ape_224.naver_in1k` | 224 | 80.974 | 224 | 1000 | apache-2.0 | [timm](https://huggingface.co/timm/vit_small_patch16_rope_mixed_ape_224.naver_in1k) |

### `lvd1689m` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_7b_patch16_dinov3.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_7b_patch16_dinov3.lvd1689m) |
| `vit_base_patch16_dinov3.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_base_patch16_dinov3.lvd1689m) |
| `vit_base_patch16_dinov3_qkvb.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_base_patch16_dinov3_qkvb.lvd1689m) |
| `vit_huge_plus_patch16_dinov3.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_huge_plus_patch16_dinov3.lvd1689m) |
| `vit_huge_plus_patch16_dinov3_qkvb.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_huge_plus_patch16_dinov3_qkvb.lvd1689m) |
| `vit_large_patch16_dinov3.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_large_patch16_dinov3.lvd1689m) |
| `vit_large_patch16_dinov3_qkvb.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_large_patch16_dinov3_qkvb.lvd1689m) |
| `vit_small_patch16_dinov3.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_small_patch16_dinov3.lvd1689m) |
| `vit_small_patch16_dinov3_qkvb.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_small_patch16_dinov3_qkvb.lvd1689m) |
| `vit_small_plus_patch16_dinov3.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_small_plus_patch16_dinov3.lvd1689m) |
| `vit_small_plus_patch16_dinov3_qkvb.lvd1689m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_small_plus_patch16_dinov3_qkvb.lvd1689m) |

### Sapiens2

预训练配置引用了 [Sapiens2](https://arxiv.org/abs/2604.21681)。

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_5b_patch16_sapiens2.fb` | 1024x768 | - | - | none | sapiens2-license | [timm](https://huggingface.co/timm/vit_5b_patch16_sapiens2.fb) |
| `vit_base_patch16_sapiens2.fb` | 1024x768 | - | - | none | sapiens2-license | [timm](https://huggingface.co/timm/vit_base_patch16_sapiens2.fb) |
| `vit_giant_patch16_sapiens2.fb` | 1024x768 | - | - | none | sapiens2-license | [timm](https://huggingface.co/timm/vit_giant_patch16_sapiens2.fb) |
| `vit_huge_patch16_sapiens2.fb` | 1024x768 | - | - | none | sapiens2-license | [timm](https://huggingface.co/timm/vit_huge_patch16_sapiens2.fb) |
| `vit_large_patch16_sapiens2.fb` | 1024x768 | - | - | none | sapiens2-license | [timm](https://huggingface.co/timm/vit_large_patch16_sapiens2.fb) |

### `merged2b` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `eva02_base_patch16_clip_224.merged2b` | 224 | - | - | 512 | mit | [timm](https://huggingface.co/timm/eva02_base_patch16_clip_224.merged2b) |
| `eva02_large_patch14_clip_224.merged2b` | 224 | - | - | 768 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_clip_224.merged2b) |
| `eva02_large_patch14_clip_336.merged2b` | 336 | - | - | 768 | mit | [timm](https://huggingface.co/timm/eva02_large_patch14_clip_336.merged2b) |
| `eva_giant_patch14_clip_224.merged2b` | 224 | - | - | 1024 | mit | [timm](https://huggingface.co/timm/eva_giant_patch14_clip_224.merged2b) |

### `sbb` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_base_patch16_rope_reg1_gap_256.sbb_in1k` | 256 | 83.872 | 256 | 1000 | mit | [timm](https://huggingface.co/timm/vit_base_patch16_rope_reg1_gap_256.sbb_in1k) |
| `vit_betwixt_patch16_rope_reg4_gap_256.sbb_in1k` | 256 | 83.906 | 256 | 1000 | mit | [timm](https://huggingface.co/timm/vit_betwixt_patch16_rope_reg4_gap_256.sbb_in1k) |
| `vit_medium_patch16_rope_reg1_gap_256.sbb_in1k` | 256 | 83.810 | 256 | 1000 | mit | [timm](https://huggingface.co/timm/vit_medium_patch16_rope_reg1_gap_256.sbb_in1k) |
| `vit_mediumd_patch16_rope_reg1_gap_256.sbb_in1k` | 256 | 84.316 | 256 | 1000 | mit | [timm](https://huggingface.co/timm/vit_mediumd_patch16_rope_reg1_gap_256.sbb_in1k) |

### 面向稠密空间感知的视觉预训练

预训练配置引用了[Vision Pretraining for Dense Spatial Perception](https://arxiv.org/abs/2607.05247)。

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_base_patch16_lingbot.robbyant` | 512 | - | - | none | apache-2.0 | [belfner/vit_base_patch16_lingbot.robbyant](https://huggingface.co/belfner/vit_base_patch16_lingbot.robbyant) |
| `vit_giant_patch16_lingbot.robbyant` | 512 | - | - | none | apache-2.0 | [belfner/vit_giant_patch16_lingbot.robbyant](https://huggingface.co/belfner/vit_giant_patch16_lingbot.robbyant) |
| `vit_large_patch16_lingbot.robbyant` | 512 | - | - | none | apache-2.0 | [belfner/vit_large_patch16_lingbot.robbyant](https://huggingface.co/belfner/vit_large_patch16_lingbot.robbyant) |
| `vit_small_patch16_lingbot.robbyant` | 512 | - | - | none | apache-2.0 | [belfner/vit_small_patch16_lingbot.robbyant](https://huggingface.co/belfner/vit_small_patch16_lingbot.robbyant) |

### `eupe` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_base_patch16_dinov3_qkvb.eupe_lvd1689m` | 256 | - | - | none | fair-noncommercial-research-license | [timm](https://huggingface.co/timm/vit_base_patch16_dinov3_qkvb.eupe_lvd1689m) |
| `vit_small_patch16_dinov3_qkvb.eupe_lvd1689m` | 256 | - | - | none | fair-noncommercial-research-license | [timm](https://huggingface.co/timm/vit_small_patch16_dinov3_qkvb.eupe_lvd1689m) |
| `vit_tiny_patch16_dinov3_qkvb.eupe_lvd1689m` | 256 | - | - | none | fair-noncommercial-research-license | [timm](https://huggingface.co/timm/vit_tiny_patch16_dinov3_qkvb.eupe_lvd1689m) |

### `sat493m` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `vit_7b_patch16_dinov3.sat493m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_7b_patch16_dinov3.sat493m) |
| `vit_large_patch16_dinov3.sat493m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_large_patch16_dinov3.sat493m) |
| `vit_large_patch16_dinov3_qkvb.sat493m` | 256 | - | - | none | dinov3-license | [timm](https://huggingface.co/timm/vit_large_patch16_dinov3_qkvb.sat493m) |

### `clip` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `eva_giant_patch14_224.clip_ft_in1k` | 224 | 88.896 | 224 | 1000 | mit | [timm](https://huggingface.co/timm/eva_giant_patch14_224.clip_ft_in1k) |
| `eva_giant_patch14_336.clip_ft_in1k` | 336 | 89.462 | 336 | 1000 | mit | [timm](https://huggingface.co/timm/eva_giant_patch14_336.clip_ft_in1k) |

### `laion2b` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `eva02_enormous_patch14_clip_224.laion2b` | 224 | - | - | 1024 | mit | [timm](https://huggingface.co/timm/eva02_enormous_patch14_clip_224.laion2b) |
| `eva02_enormous_patch14_clip_224.laion2b_plus` | 224 | - | - | 1024 | mit | [timm](https://huggingface.co/timm/eva02_enormous_patch14_clip_224.laion2b_plus) |

### `m30m` 标签

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `eva_giant_patch14_336.m30m_ft_in22k_in1k` | 336 | 89.566 | 336 | 1000 | mit | [timm](https://huggingface.co/timm/eva_giant_patch14_336.m30m_ft_in22k_in1k) |
| `eva_giant_patch14_560.m30m_ft_in22k_in1k` | 560 | 89.790 | 560 | 1000 | mit | [timm](https://huggingface.co/timm/eva_giant_patch14_560.m30m_ft_in22k_in1k) |

### 其他

| 权重 | 输入大小 | Top-1（%） | 评估大小 | 类 | 许可证 | Hub |
|---|---:|---:|---:|---:|---|---|
| `eva_giant_patch14_clip_224.laion400m` | 224 | - | - | 1024 | mit | [timm](https://huggingface.co/timm/eva_giant_patch14_clip_224.laion400m) |

## 引用

```bibtex
@article{EVA,
  title={EVA: Exploring the Limits of Masked Visual Representation Learning at Scale},
  author={Fang, Yuxin and Wang, Wen and Xie, Binhui and Sun, Quan and Wu, Ledell and Wang, Xinggang and Huang,
  Tiejun and Wang, Xinlong and Cao, Yue},
  journal={arXiv preprint arXiv:2211.07636},
  year={2022}
}

@article{EVA02,
  title={EVA-02: A Visual Representation for Neon Genesis},
  author={Fang, Yuxin and Sun, Quan and Wang, Xinggang and Huang, Tiejun and Wang, Xinlong and Cao, Yue},
  journal={arXiv preprint arXiv:2303.11331},
  year={2023}
}

@article{bolya2025perception,
  title={Perception encoder: The best visual embeddings are not at the output of the network},
  author={Bolya, Daniel and Huang, Po-Yao and Sun, Peize and Cho, Jang Hyun and Madotto, Andrea and Wei, Chen and Ma,
    Tengyu and Zhi, Jiale and Rajasegaran, Jathushan and Rasheed, Hanoona and others},
  journal={arXiv preprint arXiv:2504.13181},
  year={2025}
}

@inproceedings{heo2024rotary,
  title={Rotary position embedding for vision transformer},
  author={Heo, Byeongho and Park, Song and Han, Dongyoon and Yun, Sangdoo},
  booktitle={European Conference on Computer Vision},
  pages={289--305},
  year={2024},
  organization={Springer}
}

@article{simeoni2025dinov3,
  title={{DINOv3}},
  author={Sim{'e}oni, Oriane and Vo, Huy V. and Seitzer, Maximilian and Baldassarre, Federico and Oquab, Maxime
    and Jose, Cijo and Khalidov, Vasil and Szafraniec, Marc and Yi, Seungeun and Ramamonjisoa, Micha{"e}l
    and Massa, Francisco and Haziza, Daniel and Wehrstedt, Luca and Wang, Jianyuan and Darcet, Timoth{'e}e
    and Moutakanni, Th{'e}o and Sentana, Leonel and Roberts, Claire and Vedaldi, Andrea and Tolan, Jamie
    and Brandt, John and Couprie, Camille and Mairal, Julien and J{'e}gou, Herv{'e} and Labatut, Patrick
    and Bojanowski, Piotr},
  year={2025},
  eprint={2508.10104},
  url={https://arxiv.org/abs/2508.10104},
}

@article{lingbot-vision2026,
  title={Vision Pretraining for Dense Spatial Perception},
  author={Fu, Zelin and Tan, Bin and Sun, Changjiang and Liu, Shaohui and Zheng, Kecheng and Xu, Yinghao
    and Zhu, Xing and Shen, Yujun and Xue, Nan},
  journal={arXiv preprint arXiv:2607.05247},
  year={2026}
}

@article{khirodkarsapiens2,
  title={Sapiens2},
  author={Khirodkar, Rawal and Wen, He and Martinez, Julieta and Dong, Yuan and Su, Zhaoen and Saito, Shunsuke},
  journal={arXiv preprint arXiv:2604.21681},
  year={2026}
}
```