osmr / imgclsmob

TOP 1 ACCURACY TOP 5 ACCURACY
SPEED
MODEL CODE PAPER
ε-REPR
CODE PAPER
ε-REPR
PAPER
GLOBAL RANK
108-MENet-8x1
(g=3)
56.1% -- 79.2% -- #376
1.0-SqNxt-23
57.5% -- 80.9% -- #374
1.0-SqNxt-23v5
59.2% -- 82.2% -- #362
128-MENet-8x1
(g=4)
57.6% -- 80.4% -- #373
1.5-SqNxt-23
65.1% -- 86.5% -- #345
1.5-SqNxt-23v5
66.2% -- 87.0% -- #334
160-MENet-8x1
(g=8)
56.2% -- 79.2% -- #378
2.0-SqNxt-23
69.4% -- 89.0% -- #312
2.0-SqNxt-23v5
70.4% -- 89.3% -- #299
228-MENet-12x1
(g=3)
65.9% -- 86.8% -- #341
256-MENet-12x1
(g=4)
67.3% -- 87.5% -- #329
348-MENet-12x1
(g=3)
71.8% -- 90.4% -- #286
352-MENet-12x1
(g=8)
68.4% -- 88.0% -- #324
456-MENet-24x1
(g=3)
74.7% -- 92.0% -- #248
AlexNet
59.0% -- 81.8% -- 417.5 #366
AlexNet-b
58.4% -- 81.0% -- 418.1 #369
BAM-ResNet-50
76.9% -- 93.4% -- #197
BN-Inception
74.6% -- 92.3% -- 426.1 #246
BN-VGG-11
71.0% -- 90.4% -- 420.9 #284
BN-VGG-11b
70.4% -- 90.0% -- 415.6 #298
BN-VGG-13
72.2% -- 90.9% -- 406.0 #280
BN-VGG-13b
71.6% -- 90.4% -- 394.5 #286
BN-VGG-16
74.3% -- 92.2% -- 400.7 #238
BN-VGG-16b
72.8% -- 91.3% -- 394.4 #275
BN-VGG-19
75.9% -- 92.9% -- 395.8 #221
BN-VGG-19b
73.9% -- 91.6% -- 385.3 #261
CBAM-ResNet-50
77.6% -- 93.9% -- #174
CondenseNet-74
(C=G=4)
73.8% -- 91.7% -- #263
CondenseNet-74
(C=G=8)
71.1% -- 89.9% -- #297
DarkNet-53
78.3% -- 94.4% -- #125
DARTS
73.3% -- 91.3% -- #273
DenseNet-121
76.5% 76.4% 93.0% 93.3% #211
DenseNet-161
77.1% -- 93.6% -- #187
DenseNet-169
77.6% 77.9% 93.7% 94.1% #177
DenseNet-201
76.9% 78.5% 93.4% 94.5% #201
DiracNetV2-18
68.5% -- 88.3% -- #321
DiracNetV2-34
71.2% -- 90.1% -- #293
DLA-102
78.0% -- 94.0% -- #164
DLA-169
78.7% -- 94.3% -- #128
DLA-34
74.6% -- 92.1% -- #246
DLA-46-C
65.7% -- 86.8% -- #336
DLA-60
77.0% -- 93.3% -- #204
DLA-X-102
78.5% -- 94.2% -- #136
DLA-X2-102
79.5% -- 94.6% -- #97
DLA-X-46-C
66.7% -- 87.3% -- #330
DLA-X-60
78.2% -- 94.0% -- #157
DLA-X-60-C
69.0% -- 89.1% -- #307
DPN-131
79.5% -- 94.5% -- #105
DPN-68
76.8% -- 93.2% -- #207
DPN-98
79.2% -- 94.5% -- #109
DRN-C-26
75.1% -- 92.5% -- #231
DRN-C-42
77.1% -- 93.4% -- #195
DRN-C-58
78.3% -- 94.0% -- #160
DRN-D-105
79.4% -- 94.5% -- #104
DRN-D-22
74.2% -- 91.8% -- #255
DRN-D-38
76.2% -- 93.1% -- #212
DRN-D-54
78.8% -- 94.1% -- #145
EfficientNet-B0
75.2% 76.3% 92.5% 93.2% #230
EfficientNet-B0b
76.1% -- 93.0% -- #217
EfficientNet-B1
77.0% 78.8% 93.6% 94.4% #197
EfficientNet-B1b
78.4% -- 94.1% -- #152
EfficientNet-B2b
79.7% -- 94.7% -- 368.5 #88
EfficientNet-B3b
81.4% -- 95.6% -- 328.4 #53
EfficientNet-B4b
82.8% -- 96.2% -- 221.8 #36
EfficientNet-B5b
83.6% -- 96.7% -- 115.4 #28
EfficientNet-B6b
84.1% -- 96.9% -- 68.4 #18
EfficientNet-B7b
84.3% -- 96.9% -- 39.5 #17
ESPNetv2 x0.5
57.7% -- 79.8% -- #375
ESPNetv2 x1.0
66.1% -- 86.5% -- #342
ESPNetv2 x1.25
67.9% -- 87.8% -- #325
ESPNetv2 x1.5
69.2% -- 88.7% -- #314
ESPNetv2 x2.0
72.1% -- 90.4% -- #285
FBNet-Cb
75.1% -- 92.4% -- 407.7 #236
FD-MobileNet x0.25
44.1% -- 69.0% -- #384
FD-MobileNet x0.5
56.9% -- 79.8% -- #374
FD-MobileNet x0.75
61.6% -- 83.6% -- #355
FD-MobileNet x1.0
65.8% -- 86.6% -- #339
IBN(b)-ResNet-50
76.4% -- 93.1% -- #213
IBN-DenseNet-121
75.5% -- 92.8% -- #223
IBN-DenseNet-169
76.8% -- 93.5% -- #191
IBN-ResNet-101
78.7% -- 94.4% -- #120
IBN-ResNet-50
77.2% -- 93.6% -- #191
IBN-ResNeXt-101
(32x4d)
79.1% -- 94.6% -- #102
IGCV3 x0.25
46.3% -- 71.3% -- #383
IGCV3 x0.5
60.2% -- 82.7% -- #359
IGCV3 x0.75
68.9% -- 88.6% -- #317
IGCV3 x1.0
72.1% -- 90.8% -- #279
InceptionResNetV
80.1% -- 95.1% -- #68
InceptionV3
78.9% -- 94.4% -- #126
InceptionV4
79.4% -- 94.7% -- #93
i-RevNet-301
74.0% -- 91.6% -- #259
MixNet-L
78.9% 78.9% 94.2% 94.2% 400.3 #143
MixNet-M
77.1% 77.0% 93.4% 93.3% 404.0 #200
MixNet-S
76.0% 75.8% 92.8% 92.8% 413.1 #218
MnasNet
68.4% -- 88.3% -- #322
MobileNetV2 x0.25
51.7% -- 75.5% -- #382
MobileNetV2 x0.5
64.0% -- 85.1% -- #354
MobileNetV2 x0.75
69.8% -- 89.2% -- #305
MobileNetV2 x1.0
73.0% -- 91.1% -- #273
MobileNetV3 L/224/1.0
75.1% -- 92.2% -- #233
MobileNet x0.25
53.7% -- 77.5% -- #380
MobileNet x0.5
65.8% -- 86.5% -- #346
MobileNet x0.75
69.9% -- 89.2% -- #303
MobileNet x1.0
73.4% -- 91.0% -- #276
74.3% -- 91.8% -- #253
81.9% -- 95.8% -- #46
PeleeNet
68.2% -- 88.5% -- #323
PNASNet-5-Large
82.1% -- 95.7% -- #49
PolyNet
80.9% 81.3% 95.5% 95.8% #59
PrepResNet-10
64.9% -- 85.8% -- #351
PreResNet-101
78.3% -- 94.1% -- #155
PreResNet-101b
79.0% -- 94.4% -- #113
PreResNet-12
66.1% -- 86.5% -- #343
PreResNet-14
67.4% -- 87.6% -- #326
PreResNet-152
79.1% -- 94.5% -- #113
PreResNet-152b
79.9% -- 94.8% -- #83
PreResNet-16
69.5% -- 88.9% -- #311
PreResNet-18
71.6% -- 90.3% -- #288
PreResNet-200b
78.7% -- 94.1% -- #137
PreResNet-26
73.7% -- 91.5% -- #265
PreResNet-269b
79.1% -- 94.2% -- #110
PreResNet-34
75.1% -- 92.3% -- #238
PreResNet-50
77.6% -- 93.5% -- #190
PreResNet-50b
77.5% -- 93.4% -- #183
PreResNet-BC-26b
74.5% -- 92.0% -- #249
PreResNet-BC-38b
77.1% -- 93.4% -- #194
ProxylessNAS CPU
75.3% -- 92.4% -- #232
ProxylessNAS GPU
75.2% -- 92.6% -- #231
ProxylessNAS Mob-14
76.7% -- 93.4% -- #209
ProxylessNAS Mobile
74.6% -- 92.2% -- #241
PyramidNet-101
(a=360)
78.0% -- 93.8% -- #171
ResNet-10
65.3% -- 85.6% -- #352
ResNet-101
78.1% 78.2% 93.8% 94.0% #164
ResNet-101b
79.4% -- 94.7% -- #98
ResNet-12
66.4% -- 86.7% -- #337
ResNet-14
67.5% -- 87.5% -- #327
ResNet-152
79.0% 78.6%
94.5% 94.3% #120
ResNet-152b
80.1% -- 95.0% -- #76
ResNet-16
69.5% -- 88.8% -- #310
ResNet-18
71.5% 72.1% 90.2% -- 412.9 #293
ResNet-26
73.7% -- 91.5% -- #266
ResNet-34
75.2% -- 92.2% -- #232
ResNet-50
77.7% 77.1%
93.7% 93.3%
#177
ResNet-50b
77.6% -- 93.6% -- #175
ResNet-BC-26b
74.9% -- 92.0% -- #247
ResNet-BC-38b
76.3% -- 93.0% -- #214
ResNeXt-101
(32x4d)
78.2% -- 93.9% -- #167
ResNeXt-101
(64x4d)
79.0% -- 94.3% -- #132
ResNeXt-14
(16x4d)
68.1% -- 87.5% -- #328
ResNeXt-14
(32x2d)
67.4% -- 87.2% -- #328
ResNeXt-14
(32x4d)
69.7% -- 88.5% -- #307
ResNeXt-26
(32x2d)
73.4% -- 91.1% -- #275
ResNeXt-26
(32x4d)
75.9% -- 92.5% -- #223
SENet-154
81.4% 82.7% 95.4% 96.2% #59
SENet-16
74.4% -- 91.8% -- #254
SENet-28
78.1% -- 94.0% -- #165
SE-PreResNet-10
66.0% -- 86.6% -- #339
SE-PreResNet-18
71.9% -- 90.4% -- #287
SE-PreResNet-BC-26b
76.8% -- 93.4% -- #198
SE-PreResNet-BC-38b
78.4% -- 94.2% -- #137
SE-ResNet-10
66.1% -- 86.3% -- #348
SE-ResNet-101
78.1% -- 94.1% -- #149
SE-ResNet-152
78.5% -- 94.2% -- #141
SE-ResNet-18
71.8% -- 90.4% -- #283
SE-ResNet-26
74.3% -- 91.8% -- #256
SE-ResNet-50
77.5% -- 93.6% -- #181
SE-ResNet-50b
79.2% -- 94.6% -- #105
SE-ResNet-BC-26b
76.4% -- 93.0% -- #212
SE-ResNet-BC-38b
78.4% -- 94.1% -- #154
SE-ResNeXt-101
(32x4d)
80.0% -- 94.9% -- #78
SE-ResNeXt-50
(32x4d)
79.0% -- 94.5% -- #119
ShuffleNetV2b x0.5
59.7% -- 81.8% -- #365
ShuffleNetV2b x1.0
69.4% -- 88.8% -- #313
ShuffleNetV2b x1.5
72.7% -- 90.9% -- #277
ShuffleNetV2b x2.0
74.4% -- 91.7% -- #250
ShuffleNetV2 x0.5
59.0% -- 81.4% -- #367
ShuffleNetV2 x1.0
68.6% -- 88.4% -- #320
ShuffleNetV2 x1.5
72.5% -- 90.6% -- #281
ShuffleNetV2 x2.0
74.1% -- 91.6% -- #264
ShuffleNet x0.25
(g=1)
37.6% -- 62.7% -- #386
ShuffleNet x0.25
(g=3)
38.3% -- 63.5% -- #385
ShuffleNet x0.5
(g=1)
53.4% -- 77.4% -- #381
ShuffleNet x0.5
(g=3)
55.8% -- 79.2% -- #379
ShuffleNet x0.75
(g=1)
60.4% -- 82.9% -- #357
ShuffleNet x0.75
(g=3)
61.8% -- 83.5% -- #356
ShuffleNet x1.0
(g=1)
65.1% -- 86.1% -- #350
ShuffleNet x1.0
(g=2)
65.8% -- 86.4% -- #347
ShuffleNet x1.0
(g=3)
65.6% -- 86.5% -- #347
ShuffleNet x1.0
(g=4)
65.8% -- 86.6% -- #338
ShuffleNet x1.0
(g=8)
65.9% -- 86.6% -- #340
SPNASNet
74.1% -- 91.8% -- 412.2 #253
SqueezeNet v1.0
60.7% -- 82.3% -- #360
SqueezeNet v1.1
60.7% -- 82.3% -- #361
SqueezeResNet v1.0
60.2% -- 81.9% -- #361
SqueezeResNet v1.1
59.9% -- 81.8% -- #364
VGG-11
70.1% 70.4% 89.6% 89.6% 411.2 #301
VGG-13
71.2% 71.3% 90.2% 90.1% 406.1 #292
VGG-16
73.0% 74.4% 91.3% 91.9% 396.4 #274
VGG-19
73.8% 74.5% 91.6% 92.0% 390.5 #261
WRN-50-2
77.5% -- 93.6% -- #185
Xception
79.0% 79.0% 94.5% 94.5% #106
ZFNet
60.2% 64.0% 82.7% 85.3% 428.2 #362
ZFNet-b
63.6% -- 85.1% -- 426.3 #354
See Full Build Details +get badge code
[![SotaBench](https://img.shields.io/endpoint.svg?url=https://sotabench.com/api/v0/badge/gh/osmr/imgclsmob)](https://sotabench.com/user/osemery/repos/osmr/imgclsmob)

How the Repository is Evaluated

The full sotabench.py file - source
from torchbench.image_classification import ImageNet
from pytorch.pytorchcv.model_provider import trained_model_metainfo_list
from pytorch.pytorchcv.model_provider import get_model as ptcv_get_model
import torchvision.transforms as transforms
import torch
import math
# import os

for model_metainfo in trained_model_metainfo_list:
    net = ptcv_get_model(model_metainfo[0], pretrained=True)
    input_image_size = model_metainfo[3]
    resize_inv_factor = model_metainfo[4]
    batch_size = model_metainfo[5]
    model_description = model_metainfo[6]
    assert (not hasattr(net, "in_size")) or (input_image_size == net.in_size[0])
    ImageNet.benchmark(
        model=net,
        model_description=model_description,
        paper_model_name=model_metainfo[1],
        paper_arxiv_id=model_metainfo[2],
        input_transform=transforms.Compose([
            transforms.Resize(int(math.ceil(float(input_image_size) / resize_inv_factor))),
            transforms.CenterCrop(input_image_size),
            transforms.ToTensor(),
            transforms.Normalize(
                mean=[0.485, 0.456, 0.406],
                std=[0.229, 0.224, 0.225]),
        ]),
        batch_size=batch_size,
        num_gpu=1,
        # data_root=os.path.join("..", "imgclsmob_data", "imagenet")
    )
    torch.cuda.empty_cache()
STATUS
BUILD
COMMIT MESSAGE
RUN TIME
Prep for workin on MnasNet
osmr   fb220bf  ·  yesterday
0h:41m:15s
Update PIP-packages (gluoncv2: 0.0.51, pytorchcv: 0.0.51, chaine…
osmr   458c23b  ·  yesterday
0h:42m:28s
Cosmetic refactoring
osmr   abdece4  ·  yesterday
0h:41m:18s
Update info about SPNASNet model
osmr   2a0143a  ·  5 days ago
0h:46m:53s
Update info about FBNet-Cb model
osmr   51d2d15  ·  5 days ago
unknown
Update info about MixNet-L model
osmr   e40d1f7  ·  5 days ago
0h:12m:35s
Experimenting with Sotabench, 6
osmr   ceb2e0f  ·  Nov 04 2019
1h:29m:18s
Experimenting with Sotabench, 5
osmr   b55ef4e  ·  Nov 04 2019
0h:23m:05s
Experimenting with Sotabench, 4
osmr 5cfd753    abb0a9e  ·  Nov 04 2019
1h:26m:57s
Experimenting with Sotabench, 2
osmr   6baac27  ·  Nov 02 2019
0h:42m:43s
Update info about MixNet-M model
osmr   cc191ac  ·  Nov 02 2019
6h:34m:35s
Experimenting with Sotabench
osmr   a7c8968  ·  Oct 29 2019
0h:52m:21s
0h:15m:25s
unknown
unknown
Update README
osmr   dcf351b  ·  Oct 29 2019
0h:04m:34s
Update info about MixNet-S model
osmr   6cc5e11  ·  Oct 29 2019
unknown
Add sotabench
osmr   d9e9877  ·  Oct 29 2019
unknown
unknown