[🚀 ICLR 2026 Oral] NextStep-1: SOTA Autogressive Image Generation with Continuous Tokens. A research project developed by the StepFun’s Multimodal Intelligence team.
EfficientSAM3 compresses SAM3 into lightweight, edge-friendly models via progressive knowledge distillation for fast promptable concept segmentation and tracking.
[ECCV 2026 Oral] DreamID-V: Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformer
关键参数与规格
关键参数与规格
参数
值
语言
Python
类别
Models
子类别
Vision models
星标数
697
分叉数
161
贡献者
3
创建日期
2019-07-09
更新日期
2026-09-23
1 天星标数
0
7 天星标数
0
排名
14,865
上次排名
14,705
排名变化
-160
7 天星标排名
14,692
上次 7 天星标排名
14,704
7 天星标排名变化
12
总星标排名
14,937
上次总星标排名
14,908
总星标排名变化
-29
国家
USA
README 摘录
PyTorch models trained on CIFAR-10 dataset I modified TorchVision official implementation of popular CNN models, and trained those on CIFAR-10 dataset. I changed *number of class, filter size, stride, and padding* in the the original code so that it works with CIFAR-10. I also share the **weights** of these models, so you can just load the weights and use them. The code is highly re-producible and readable by using PyTorch-Lightning. 483 M | 103 MB | | 10
| densenet169 | 94. 493 M | 49 MB | | 11 | mobilenet_v2 | 93. 491 M | 22 MB | | 13 | inception_v3 | 93. 640 M | 83 MB | Details Report & Run Logs Weight and Biases' details report for this project WandB Report Weight and Biases' run logs for this project WandB Run Log. You can see each run hyper-parameters, training accuracy, validation accuracy, loss, time taken. How To Cite How to use pretrained models **Automatically download and extract the weights from Box (933 MB)** Or
use Google Drive backup link (you have to download and extract manually) **Load model and run** If you use your own images, all models expect data to be in range [0, 1] then normalized by How to train models from scratch Check the to see all available hyper-parameter choices. To reproduce the same accuracy use the default hyper-parameters How to test pretrained models Output Requirements **Just to use pretrained models** pytorch = 1.
I modified TorchVision official implementation of popular CNN models, and trained those on CIFAR-10 dataset.
I changed number of class, filter size, stride, and padding in the the original code so that it works with CIFAR-10.
I also share the weights of these models, so you can just load the weights and use them.
The code is highly re-producible and readable by using PyTorch-Lightning.
Statistics of supported models
No.
Model
Val. Acc.
No. Params
Size
1
vgg11_bn
92.39%
28.150 M
108 MB
2
vgg13_bn
94.22%
28.334 M
109 MB
3
vgg16_bn
94.00%
33.647 M
129 MB
4
vgg19_bn
93.95%
38.959 M
149 MB
5
resnet18
93.07%
11.174 M
43 MB
6
resnet34
93.34%
21.282 M
82 MB
7
resnet50
93.65%
23.521 M
91 MB
8
densenet121
94.06%
6.956 M
28 MB
9
densenet161
94.07%
26.483 M
103 MB
10
densenet169
94.05%
12.493 M
49 MB
11
mobilenet_v2
93.91%
2.237 M
9 MB
12
googlenet
92.85%
5.491 M
22 MB
13
inception_v3
93.74%
21.640 M
83 MB
Details Report & Run Logs
Weight and Biases' details report for this project WandB Report
Weight and Biases' run logs for this project WandB Run Log. You can see each run hyper-parameters, training accuracy, validation accuracy, loss, time taken.
How To Cite
How to use pretrained models
Automatically download and extract the weights from Box (933 MB)
python train.py --download_weights 1
Or use Google Drive backup link (you have to download and extract manually)
Load model and run
from cifar10_models.vgg import vgg11_bn, vgg13_bn, vgg16_bn, vgg19_bn
# Untrained model
my_model = vgg11_bn()
# Pretrained model
my_model = vgg11_bn(pretrained=True)
my_model.eval() # for evaluation
If you use your own images, all models expect data to be in range [0, 1] then normalized by
mean = [0.4914, 0.4822, 0.4465]
std = [0.2471, 0.2435, 0.2616]
How to train models from scratch
Check the train.py to see all available hyper-parameter choices.
To reproduce the same accuracy use the default hyper-parameters
