After converting SE-ResNeXt models

osmr · Aug 19, 2018 · bf58c1d · bf58c1d
1 parent 524370c
commit bf58c1d
Show file tree

Hide file tree

Showing 3 changed files with 10 additions and 2 deletions.
diff --git a/README.md b/README.md
@@ -79,6 +79,8 @@ bottleneck block. Respectively a network without b-suffix has the stride in the
 | SE-ResNet-50 | 22.51 | 6.44 | 28,088,024 | 3,877.01M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.11/seresnet50-0644-10954a84.params.log)) |
 | SE-ResNet-101 | 21.92 | 5.89 | 49,326,872 | 7,600.01M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.11/seresnet101-0589-4c10238d.params.log)) |
 | SE-ResNet-152 | 21.48 | 5.77 | 66,821,848 | 11,324.62M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.11/seresnet152-0577-de6f099d.params.log)) |
+| SE-ResNeXt-50 (32x4d) | 21.06 | 5.58 | 27,559,896 | 4,253.33M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.12/seresnext50_32x4d-0558-a49f8fb0.params.log)) |
+| SE-ResNeXt-101 (32x4d) | 19.99 | 5.00 | 48,955,416 | 8,005.33M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.12/seresnext101_32x4d-0500-cf161260.params.log)) |
 | DenseNet-121 | 25.11 | 7.80 | 7,978,856 | 2,852.39M | Converted from Gluon Model Zoo ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.3/densenet121-0780-49b72d04.params.log)) |
 | DenseNet-161 | 22.40 | 6.18 | 28,681,000 | 7,761.25M | Converted from Gluon Model Zoo ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.3/densenet161-0618-52e30516.params.log)) |
 | DenseNet-169 | 23.89 | 6.89 | 14,149,480 | 3,381.48M | Converted from Gluon Model Zoo ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.3/densenet169-0689-281ec06b.params.log)) |
@@ -137,6 +139,8 @@ bottleneck block. Respectively a network without b-suffix has the stride in the
 | SE-ResNet-50 | 22.47 | 6.40 | 28,088,024 | 3,877.01M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.11/seresnet50-0640-8820f2af.pth.log)) |
 | SE-ResNet-101 | 21.88 | 5.89 | 49,326,872 | 7,600.01M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.11/seresnet101-0589-5e6e831b.pth.log)) |
 | SE-ResNet-152 | 21.48 | 5.76 | 66,821,848 | 11,324.62M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.11/seresnet152-0576-814cf72e.pth.log)) |
+| SE-ResNeXt-50 (32x4d) | 21.00 | 5.54 | 27,559,896 | 4,253.33M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.12/seresnext50_32x4d-0554-99e0e9aa.pth.log)) |
+| SE-ResNeXt-101 (32x4d) | 19.96 | 5.05 | 48,955,416 | 8,005.33M | From [Cadene/pretrained...pytorch] ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.12/seresnext101_32x4d-0505-0924f0a2.pth.log)) |
 | DenseNet-121 | 25.57 | 8.03 | 7,978,856 | 2,852.39M | Converted from Gluon Model Zoo ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.3/densenet121-0803-f994107a.pth.log)) |
 | DenseNet-161 | 22.86 | 6.44 | 28,681,000 | 7,761.25M | Converted from Gluon Model Zoo ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.3/densenet161-0644-c0fb22c8.pth.log)) |
 | DenseNet-169 | 24.40 | 7.19 | 14,149,480 | 3,381.48M | Converted from Gluon Model Zoo ([log](https://github.com/osmr/imgclsmob/releases/download/v0.0.3/densenet169-0719-27139105.pth.log)) |

diff --git a/gluon/models/model_store.py b/gluon/models/model_store.py
@@ -34,11 +34,13 @@
     ('preresnet152b', '0575', 'dc303191ea47ca258f5abadd203b5de24d059d1a', 'v0.0.2'),
     ('resnext101_32x4d', '0579', '9afbfdbc5a420a9f56058be0bf80d12b21a627af', 'v0.0.10'),
     ('resnext101_64x4d', '0541', '0d4fd87b8de78c5c0295e1dcb9923a578dce7adb', 'v0.0.10'),
-
     ('seresnet50', '0644', '10954a846a56a387a6a222e260d95fb8a9bd68c3', 'v0.0.11'),
     ('seresnet101', '0589', '4c10238dd485a540a464bf1c39a8752d2da040b9', 'v0.0.11'),
     ('seresnet152', '0577', 'de6f099dd39f374390639ca8854b2954af3c59b9', 'v0.0.11'),
 
+    ('seresnext50_32x4d', '0558', 'a49f8fb039973979afe2fc70974a8b07c7159bca', 'v0.0.12'),
+    ('seresnext101_32x4d', '0500', 'cf1612601f319a0e75190ae756ae380b947dcb1a', 'v0.0.12'),
+
     ('densenet121', '0780', '49b72d04bace00bb1964b38cec13d19059a14e86', 'v0.0.3'),
     ('densenet161', '0618', '52e30516e566bdef53dcb417f86849530c83d0d1', 'v0.0.3'),
     ('densenet169', '0689', '281ec06b02f407b4523245622371da669a287044', 'v0.0.3'),

diff --git a/pytorch/models/model_store.py b/pytorch/models/model_store.py
@@ -35,11 +35,13 @@
     ('preresnet152b', '0591', '2c91ab2c8d90f3990e7c30fd6ee2184f6c2c3bee', 'v0.0.2'),
     ('resnext101_32x4d', '0611', 'cf962440f11fe683fd02ec04f2102d9f47ce38a7', 'v0.0.10'),
     ('resnext101_64x4d', '0575', '651abd029bcc4ce88c62e1d900a710f284a8281e', 'v0.0.10'),
-
     ('seresnet50', '0640', '8820f2af62421ce2e1df989d6e0ce7916c78ff86', 'v0.0.11'),
     ('seresnet101', '0589', '5e6e831b7518b9b8a049dd60ed1ff82ae75ff55e', 'v0.0.11'),
     ('seresnet152', '0576', '814cf72e0deeab530332b16fb9b609e574afec61', 'v0.0.11'),
 
+    ('seresnext50_32x4d', '0554', '99e0e9aa4578af9f15045c1ceeb684a2e988628a', 'v0.0.12'),
+    ('seresnext101_32x4d', '0505', '0924f0a2c1de90dc964c482b7aff6232dbef3600', 'v0.0.12'),
+
     ('densenet121', '0803', 'f994107a83aed162916ff89e2ded4c5af5bc6457', 'v0.0.3'),
     ('densenet161', '0644', 'c0fb22c83e8077a952ce1a0c9703d1a08b2b9e3a', 'v0.0.3'),
     ('densenet169', '0719', '271391051775ba9bbf458a6bd77af4b3007dc892', 'v0.0.3'),