mirrors
/
PaddleClas
mirror of https://github.com/PaddlePaddle/PaddleClas.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
PaddleClas/docs/zh_CN/models/ImageNet1k/ViT_and_DeiT.md

68 lines
5.6 KiB
Markdown
Raw Normal View History

add menu and modify documentation
2021-11-25 15:40:40 +08:00
# ViT 与 DeiT 系列
---
## 目录
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
add menu and modify documentation
2021-11-25 15:40:40 +08:00
* [1. 概述](#1)
* [2. 精度、FLOPS 和参数量](#2)
add speed to vits
2021-12-15 15:58:47 +08:00
* [3. 基于V100 GPU 的预测速度](#3)
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
add menu and modify documentation
2021-11-25 15:40:40 +08:00
<a name='1'></a>
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
add menu and modify documentation
2021-11-25 15:40:40 +08:00
## 1. 概述
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
add menu and modify documentation
2021-11-25 15:40:40 +08:00
ViT（Vision Transformer）系列模型是 Google 在 2020 年提出的，该模型仅使用标准的 Transformer 结构，完全抛弃了卷积结构，将图像拆分为多个 patch 后再输入到 Transformer 中，展示了 Transformer 在 CV 领域的潜力。[论文地址](https://arxiv.org/abs/2010.11929)。
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
add menu and modify documentation
2021-11-25 15:40:40 +08:00
DeiT（Data-efficient Image Transformers）系列模型是由 FaceBook 在 2020 年底提出的，针对 ViT 模型需要大规模数据集训练的问题进行了改进，最终在 ImageNet 上取得了 83.1%的 Top1 精度。并且使用卷积模型作为教师模型，针对该模型进行知识蒸馏，在 ImageNet 数据集上可以达到 85.2% 的 Top1 精度。[论文地址](https://arxiv.org/abs/2012.12877)。
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
add menu and modify documentation
2021-11-25 15:40:40 +08:00
<a name='2'></a>
## 2. 精度、FLOPS 和参数量
Add ViT and DeiT (#579) * Update the discription about ViT and DeiT * Fix the error file * Unified data format
2021-01-27 16:53:21 +08:00
Add SwinTransformer (#721) * Add SwinTransformer_large * Add SwinTransformer * Optimize user experience * Add the doc of SwinTransformer * Modify the format of the corner mark
2021-05-17 14:55:38 +08:00
| Models | Top1 | Top5 | Reference<br>top1 | Reference<br>top5 | FLOPS<br>(G) | Params<br>(M) |
|:--:|:--:|:--:|:--:|:--:|:--:|:--:|
docs: update metrics
2022-06-20 13:58:44 +08:00
| ViT_small_patch16_224 | 0.7553 | 0.9211 | 0.7785 | 0.9342 | 9.41 | 48.60 |
| ViT_base_patch16_224 | 0.8187 | 0.9618 | 0.8178 | 0.9613 | 16.85 | 86.42 |
add flops and params of ViT and DeiT
2021-12-15 18:13:43 +08:00
| ViT_base_patch16_384 | 0.8414 | 0.9717 | 0.8420 | 0.9722 | 49.35 | 86.42 |
| ViT_base_patch32_384 | 0.8176 | 0.9613 | 0.8166 | 0.9613 | 12.66 | 88.19 |
docs: update metrics
2022-06-20 13:58:44 +08:00
| ViT_large_patch16_224 | 0.8303 | 0.9655 | 0.8306 | 0.9644 | 59.65 | 304.12 |
add flops and params of ViT and DeiT
2021-12-15 18:13:43 +08:00
| ViT_large_patch16_384 | 0.8513 | 0.9736 | 0.8517 | 0.9736 | 174.70 | 304.12 |
| ViT_large_patch32_384 | 0.8153 | 0.9608 | 0.815 | - | 44.24 | 306.48 |
Add SwinTransformer (#721) * Add SwinTransformer_large * Add SwinTransformer * Optimize user experience * Add the doc of SwinTransformer * Modify the format of the corner mark
2021-05-17 14:55:38 +08:00
| Models | Top1 | Top5 | Reference<br>top1 | Reference<br>top5 | FLOPS<br>(G) | Params<br>(M) |
|:--:|:--:|:--:|:--:|:--:|:--:|:--:|
docs: update metrics
2022-06-20 13:52:39 +08:00
| DeiT_tiny_patch16_224 | 0.7208 | 0.9112 | 0.722 | 0.911 | 1.07 | 5.68 |
| DeiT_small_patch16_224 | 0.7982 | 0.9495 | 0.799 | 0.950 | 4.24 | 21.97 |
| DeiT_base_patch16_224 | 0.8180 | 0.9558 | 0.818 | 0.956 | 16.85 | 86.42 |
| DeiT_base_patch16_384 | 0.8289 | 0.9624 | 0.829 | 0.972 | 49.35 | 86.42 |
| DeiT_tiny_distilled_patch16_224 | 0.7449 | 0.9192 | 0.745 | 0.919 | 1.08 | 5.87 |
| DeiT_small_distilled_patch16_224 | 0.8117 | 0.9538 | 0.812 | 0.954 | 4.26 | 22.36 |
| DeiT_base_distilled_patch16_224 | 0.8330 | 0.9647 | 0.834 | 0.965 | 16.93 | 87.18 |
| DeiT_base_distilled_patch16_384 | 0.8520 | 0.9720 | 0.852 | 0.972 | 49.43 | 87.18 |
add speed to vits
2021-12-15 15:58:47 +08:00
<a name='3'></a>
## 3. 基于 V100 GPU 的预测速度
| Models | Crop Size | Resize Short Size | FP32<br/>Batch Size=1<br/>(ms) | FP32<br/>Batch Size=4<br/>(ms) | FP32<br/>Batch Size=8<br/>(ms) |
| -------------------------- | --------- | ----------------- | ------------------------------ | ------------------------------ | ------------------------------ |
| ViT_small_<br/>patch16_224 | 256 | 224 | 3.71 | 9.05 | 16.72 |
| ViT_base_<br/>patch16_224 | 256 | 224 | 6.12 | 14.84 | 28.51 |
| ViT_base_<br/>patch16_384 | 384 | 384 | 14.15 | 48.38 | 95.06 |
| ViT_base_<br/>patch32_384 | 384 | 384 | 4.94 | 13.43 | 24.08 |
| ViT_large_<br/>patch16_224 | 256 | 224 | 15.53 | 49.50 | 94.09 |
| ViT_large_<br/>patch16_384 | 384 | 384 | 39.51 | 152.46 | 304.06 |
| ViT_large_<br/>patch32_384 | 384 | 384 | 11.44 | 36.09 | 70.63 |
| Models | Crop Size | Resize Short Size | FP32<br/>Batch Size=1<br/>(ms) | FP32<br/>Batch Size=4<br/>(ms) | FP32<br/>Batch Size=8<br/>(ms) |
| ------------------------------------ | --------- | ----------------- | ------------------------------ | ------------------------------ | ------------------------------ |
| DeiT_tiny_<br>patch16_224 | 256 | 224 | 3.61 | 3.94 | 6.10 |
| DeiT_small_<br>patch16_224 | 256 | 224 | 3.61 | 6.24 | 10.49 |
| DeiT_base_<br>patch16_224 | 256 | 224 | 6.13 | 14.87 | 28.50 |
| DeiT_base_<br>patch16_384 | 384 | 384 | 14.12 | 48.80 | 97.60 |
| DeiT_tiny_<br>distilled_patch16_224 | 256 | 224 | 3.51 | 4.05 | 6.03 |
| DeiT_small_<br>distilled_patch16_224 | 256 | 224 | 3.70 | 6.20 | 10.53 |
| DeiT_base_<br>distilled_patch16_224 | 256 | 224 | 6.17 | 14.94 | 28.58 |
| DeiT_base_<br>distilled_patch16_384 | 384 | 384 | 14.12 | 48.76 | 97.09 |