Merge branch 'brain-score:master' into ReAlnet_cornet

brainscore_vision/models/mobilenet_v2_0_5_192/model.py

@@ -15,7 +15,16 @@
                                     version_id="null",
                                     sha1="e5aa083caa4833fccd48af0c578a45064824dd7f")
 MainModel = imp.load_source('MainModel',model_path.as_posix())
-model = torch.load(model_weight_path.as_posix())
+
+
+# This custom wrapper handles background class removal, and is used in related mobilenets
+class MobilenetPytorchWrapper(PytorchWrapper):
+    def __call__(self, *args, **kwargs):
+        result = super().__call__(*args, **kwargs)  # retrieve original output
+        if 'logits' in kwargs.get('layers', []):
+            result = result.isel(neuroid=slice(1, None))  # remove background class in last layer
+        return result
+
 
 def get_model(name):
     """
@@ -27,8 +36,9 @@ def get_model(name):
     :return: the model instance
     """
     assert name == 'mobilenet_v2_0_5_192'
+    model = torch.load(model_weight_path.as_posix(), weights_only=False)
     preprocessing = functools.partial(load_preprocess_images, image_size=192, preprocess_type='inception')
-    wrapper = PytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
+    wrapper = MobilenetPytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
     wrapper.image_size = 192
     return wrapper
 
@@ -70,4 +80,4 @@ def get_bibtex(name):
 if __name__ == '__main__':
     # Use this method to ensure the correctness of the BaseModel implementations.
     # It executes a mock run of brain-score benchmarks.
-    check_models.check_base_models(__name__)
+    check_models.check_base_models(__name__)

brainscore_vision/models/mobilenet_v2_0_5_224/model.py

@@ -1,8 +1,8 @@
 import functools
 from brainscore_vision.model_helpers.activations.pytorch import load_preprocess_images
-from brainscore_vision.model_helpers.activations.pytorch import PytorchWrapper
 from brainscore_vision.model_helpers.check_submission import check_models
 from brainscore_vision.model_helpers.s3 import load_weight_file
+from brainscore_vision.models.mobilenet_v2_0_5_192.model import MobilenetPytorchWrapper
 import torch
 import imp
 
@@ -15,7 +15,6 @@
                                     version_id="null",
                                     sha1="649501eadcf01f871bdb2265aa7dcac80594160a")
 MainModel = imp.load_source('MainModel',model_path.as_posix())
-model = torch.load(model_weight_path.as_posix())
 
 def get_model(name):
     """
@@ -27,17 +26,9 @@ def get_model(name):
     :return: the model instance
     """
     assert name == 'mobilenet_v2_0_5_224'
-    last_layer = model
-    while last_layer._modules:
-        last_layer = last_layer._modules[next(reversed(last_layer._modules))]
-    last_layer.register_forward_hook(lambda _layer, _input, logits: logits[:, 1:])
-
+    model = torch.load(model_weight_path.as_posix(), weights_only=False)
     preprocessing = functools.partial(load_preprocess_images, image_size=224, preprocess_type='inception')
-    wrapper = PytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
-    wrapper.image_size = 224
-    return wrapper
-    preprocessing = functools.partial(load_preprocess_images, image_size=224, preprocess_type='inception')
-    wrapper = PytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
+    wrapper = MobilenetPytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
     wrapper.image_size = 224
     return wrapper
 

brainscore_vision/models/mobilenet_v2_0_75_160/model.py

@@ -1,8 +1,8 @@
 import functools
 from brainscore_vision.model_helpers.activations.pytorch import load_preprocess_images
-from brainscore_vision.model_helpers.activations.pytorch import PytorchWrapper
 from brainscore_vision.model_helpers.check_submission import check_models
 from brainscore_vision.model_helpers.s3 import load_weight_file
+from brainscore_vision.models.mobilenet_v2_0_5_192.model import MobilenetPytorchWrapper
 import torch
 import imp
 
@@ -15,7 +15,6 @@
                                     version_id="null",
                                     sha1="9fc6f5e9864d524760c6e1dc8aa5702415457df4")
 MainModel = imp.load_source('MainModel',model_path.as_posix())
-model = torch.load(model_weight_path.as_posix())
 
 
 def get_model(name):
@@ -28,8 +27,9 @@ def get_model(name):
     :return: the model instance
     """
     assert name == 'mobilenet_v2_0_75_160'
+    model = torch.load(model_weight_path.as_posix(), weights_only=False)
     preprocessing = functools.partial(load_preprocess_images, image_size=160, preprocess_type='inception')
-    wrapper = PytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
+    wrapper = MobilenetPytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
     wrapper.image_size = 160
     return wrapper
 
@@ -71,4 +71,4 @@ def get_bibtex(name):
 if __name__ == '__main__':
     # Use this method to ensure the correctness of the BaseModel implementations.
     # It executes a mock run of brain-score benchmarks.
-    check_models.check_base_models(__name__)
+    check_models.check_base_models(__name__)

brainscore_vision/models/mobilenet_v2_0_75_192/model.py

@@ -1,8 +1,8 @@
 import functools
 from brainscore_vision.model_helpers.activations.pytorch import load_preprocess_images
-from brainscore_vision.model_helpers.activations.pytorch import PytorchWrapper
 from brainscore_vision.model_helpers.check_submission import check_models
 from brainscore_vision.model_helpers.s3 import load_weight_file
+from brainscore_vision.models.mobilenet_v2_0_5_192.model import MobilenetPytorchWrapper
 import torch
 import imp
 
@@ -15,7 +15,7 @@
                                     version_id="null",
                                     sha1="af063236e83cb92fd78ed3eb7d9d2d4a65d794ab")
 MainModel = imp.load_source('MainModel', model_path.as_posix())
-model = torch.load(model_weight_path.as_posix())
+
 
 def get_model(name):
     """
@@ -28,7 +28,8 @@ def get_model(name):
     """
     assert name == 'mobilenet_v2_0_75_192'
     preprocessing = functools.partial(load_preprocess_images, image_size=192, preprocess_type='inception')
-    wrapper = PytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
+    model = torch.load(model_weight_path.as_posix(), weights_only=False)
+    wrapper = MobilenetPytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
     wrapper.image_size = 192
     return wrapper
 
@@ -68,4 +69,4 @@ def get_bibtex(name):
 if __name__ == '__main__':
     # Use this method to ensure the correctness of the BaseModel implementations.
     # It executes a mock run of brain-score benchmarks.
-    check_models.check_base_models(__name__)
+    check_models.check_base_models(__name__)

brainscore_vision/models/mobilenet_v2_0_75_224/__init__.py

@@ -0,0 +1,7 @@
+from brainscore_vision.model_helpers.brain_transformation import ModelCommitment
+from brainscore_vision import model_registry
+from .model import get_layers,get_model
+
+
+model_registry['mobilenet_v2_0_75_224'] = \
+    lambda: ModelCommitment(identifier='mobilenet_v2_0_75_224', activations_model=get_model('mobilenet_v2_0_75_224'), layers=get_layers('mobilenet_v2_0_75_224'))

brainscore_vision/models/mobilenet_v2_0_75_224/model.py

@@ -0,0 +1,73 @@
+import functools
+from brainscore_vision.model_helpers.activations.pytorch import load_preprocess_images
+from brainscore_vision.model_helpers.check_submission import check_models
+from brainscore_vision.model_helpers.s3 import load_weight_file
+from brainscore_vision.models.mobilenet_v2_0_5_192.model import MobilenetPytorchWrapper
+import torch
+import imp
+
+model_path = load_weight_file(bucket="brainscore-storage", folder_name="brainscore-vision/models",
+                                    relative_path="mobilenet_v2_0.75_224/mobilenet.py",
+                                    version_id="null",
+                                    sha1="e2dc4cd2fc8f922755b45e1e6149c3d8cc19622e")
+model_weight_path = load_weight_file(bucket="brainscore-storage", folder_name="brainscore-vision/models",
+                                    relative_path="mobilenet_v2_0.75_224/mobilenet.pth",
+                                    version_id="null",
+                                    sha1="df37e28722d56b91d2151c2d06b5b4bc1aef2390")
+MainModel = imp.load_source('MainModel', model_path.as_posix())
+
+def get_model(name):
+    """
+    This method fetches an instance of a base model. The instance has to be callable and return a xarray object,
+    containing activations. There exist standard wrapper implementations for common libraries, like pytorch and
+    keras. Checkout the examples folder, to see more. For custom implementations check out the implementation of the
+    wrappers.
+    :param name: the name of the model to fetch
+    :return: the model instance
+    """
+    assert name == 'mobilenet_v2_0_75_224'
+    preprocessing = functools.partial(load_preprocess_images, image_size=224, preprocess_type='inception')
+    model = torch.load(model_weight_path.as_posix(), weights_only=False)
+    wrapper = MobilenetPytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
+    wrapper.image_size = 224
+    return wrapper
+
+
+def get_layers(name):
+    assert name == 'mobilenet_v2_0_75_224'
+    layer_names = (['MobilenetV2_Conv_Conv2D'] +
+                   [f'MobilenetV2_expanded_conv_{i}_expand_Conv2D' for i in range(1, 17)] +
+                   ['MobilenetV2_Conv_1_Conv2D'])
+    return layer_names
+
+
+def get_bibtex(name):
+    """
+    A method returning the bibtex reference of the requested model as a string.
+    """
+    return '''
+@article{DBLP:journals/corr/abs-1801-04381,
+  author       = {Mark Sandler and
+                  Andrew G. Howard and
+                  Menglong Zhu and
+                  Andrey Zhmoginov and
+                  Liang{-}Chieh Chen},
+  title        = {Inverted Residuals and Linear Bottlenecks: Mobile Networks for Classification,
+                  Detection and Segmentation},
+  journal      = {CoRR},
+  volume       = {abs/1801.04381},
+  year         = {2018},
+  url          = {http://arxiv.org/abs/1801.04381},
+  eprinttype    = {arXiv},
+  eprint       = {1801.04381},
+  timestamp    = {Tue, 12 Jan 2021 15:30:06 +0100},
+  biburl       = {https://dblp.org/rec/journals/corr/abs-1801-04381.bib},
+  bibsource    = {dblp computer science bibliography, https://dblp.org}
+}
+'''
+
+
+if __name__ == '__main__':
+    # Use this method to ensure the correctness of the BaseModel implementations.
+    # It executes a mock run of brain-score benchmarks.
+    check_models.check_base_models(__name__)

brainscore_vision/models/mobilenet_v2_0_75_224/region_layer_map/mobilenet_v2_0_75_224.json

@@ -0,0 +1,6 @@
+{
+    "V1": "MobilenetV2_expanded_conv_9_expand_Conv2D",
+    "V2": "MobilenetV2_expanded_conv_7_expand_Conv2D",
+    "V4": "MobilenetV2_expanded_conv_7_expand_Conv2D",
+    "IT": "MobilenetV2_expanded_conv_14_expand_Conv2D"
+}

brainscore_vision/models/mobilenet_v2_0_75_224/requirements.txt

@@ -0,0 +1,2 @@
+imp
+torch

brainscore_vision/models/mobilenet_v2_0_75_224/test.py

@@ -0,0 +1,8 @@
+import brainscore_vision
+import pytest
+
+
+@pytest.mark.travis_slow
+def test_has_identifier():
+    model = brainscore_vision.load_model('mobilenet_v2_0_75_224')
+    assert model.identifier == 'mobilenet_v2_0_75_224'

brainscore_vision/models/mobilenet_v2_1_0_128/__init__.py

@@ -0,0 +1,7 @@
+from brainscore_vision.model_helpers.brain_transformation import ModelCommitment
+from brainscore_vision import model_registry
+from .model import get_layers,get_model
+
+
+model_registry['mobilenet_v2_1_0_128'] = \
+    lambda: ModelCommitment(identifier='mobilenet_v2_1_0_128', activations_model=get_model('mobilenet_v2_1_0_128'), layers=get_layers('mobilenet_v2_1_0_128'))

brainscore_vision/models/mobilenet_v2_1_0_128/model.py

@@ -0,0 +1,73 @@
+import functools
+from brainscore_vision.model_helpers.activations.pytorch import load_preprocess_images
+from brainscore_vision.model_helpers.check_submission import check_models
+from brainscore_vision.model_helpers.s3 import load_weight_file
+from brainscore_vision.models.mobilenet_v2_0_5_192.model import MobilenetPytorchWrapper
+import torch
+import imp
+
+model_path = load_weight_file(bucket="brainscore-storage", folder_name="brainscore-vision/models",
+                                    relative_path="mobilenet_v2_1.0_128/mobilenet_v2_1.py",
+                                    version_id="null",
+                                    sha1="c5f5fe31e92fc8fee3580f1703fcd3db74f6a753")
+model_weight_path = load_weight_file(bucket="brainscore-storage", folder_name="brainscore-vision/models",
+                                    relative_path="mobilenet_v2_1.0_128/mobilenet_v2_1.0_128_frozen.pth",
+                                    version_id="null",
+                                    sha1="5f7f2f0117fc74d16e839bddf63036e6ef6b3e42")
+MainModel = imp.load_source('MainModel',model_path.as_posix())
+
+def get_model(name):
+    """
+    This method fetches an instance of a base model. The instance has to be callable and return a xarray object,
+    containing activations. There exist standard wrapper implementations for common libraries, like pytorch and
+    keras. Checkout the examples folder, to see more. For custom implementations check out the implementation of the
+    wrappers.
+    :param name: the name of the model to fetch
+    :return: the model instance
+    """
+    assert name == 'mobilenet_v2_1_0_128'
+    preprocessing = functools.partial(load_preprocess_images, image_size=128, preprocess_type='inception')
+    model = torch.load(model_weight_path.as_posix(), weights_only=False)
+    wrapper = MobilenetPytorchWrapper(identifier=name, model=model, preprocessing=preprocessing)
+    wrapper.image_size = 128
+    return wrapper
+
+
+def get_layers(name):
+    assert name == 'mobilenet_v2_1_0_128'
+    layer_names = (['MobilenetV2_Conv_Conv2D'] +
+                   [f'MobilenetV2_expanded_conv_{i}_expand_Conv2D' for i in range(1, 17)] +
+                   ['MobilenetV2_Conv_1_Conv2D'])
+    return layer_names
+
+
+def get_bibtex(name):
+    """
+    A method returning the bibtex reference of the requested model as a string.
+    """
+    return '''
+@article{DBLP:journals/corr/abs-1801-04381,
+  author       = {Mark Sandler and
+                  Andrew G. Howard and
+                  Menglong Zhu and
+                  Andrey Zhmoginov and
+                  Liang{-}Chieh Chen},
+  title        = {Inverted Residuals and Linear Bottlenecks: Mobile Networks for Classification,
+                  Detection and Segmentation},
+  journal      = {CoRR},
+  volume       = {abs/1801.04381},
+  year         = {2018},
+  url          = {http://arxiv.org/abs/1801.04381},
+  eprinttype    = {arXiv},
+  eprint       = {1801.04381},
+  timestamp    = {Tue, 12 Jan 2021 15:30:06 +0100},
+  biburl       = {https://dblp.org/rec/journals/corr/abs-1801-04381.bib},
+  bibsource    = {dblp computer science bibliography, https://dblp.org}
+}
+'''
+
+
+if __name__ == '__main__':
+    # Use this method to ensure the correctness of the BaseModel implementations.
+    # It executes a mock run of brain-score benchmarks.
+    check_models.check_base_models(__name__)

brainscore_vision/models/mobilenet_v2_1_0_128/region_layer_map/mobilenet_v2_1_0_128.json

@@ -0,0 +1,6 @@
+{
+    "V1": "MobilenetV2_expanded_conv_4_expand_Conv2D",
+    "V2": "MobilenetV2_expanded_conv_7_expand_Conv2D",
+    "V4": "MobilenetV2_expanded_conv_5_expand_Conv2D",
+    "IT": "MobilenetV2_expanded_conv_14_expand_Conv2D"
+}

brainscore_vision/models/mobilenet_v2_1_0_128/requirements.txt

@@ -0,0 +1,2 @@
+imp
+torch

brainscore_vision/models/mobilenet_v2_1_0_128/test.py

@@ -0,0 +1,8 @@
+import brainscore_vision
+import pytest
+
+
+@pytest.mark.travis_slow
+def test_has_identifier():
+    model = brainscore_vision.load_model('mobilenet_v2_1_0_128')
+    assert model.identifier == 'mobilenet_v2_1_0_128'

brainscore_vision/models/mobilenet_v2_1_0_160/__init__.py

@@ -0,0 +1,7 @@
+from brainscore_vision.model_helpers.brain_transformation import ModelCommitment
+from brainscore_vision import model_registry
+from .model import get_layers,get_model
+
+
+model_registry['mobilenet_v2_1_0_160'] = \
+    lambda: ModelCommitment(identifier='mobilenet_v2_1_0_160', activations_model=get_model('mobilenet_v2_1_0_160'), layers=get_layers('mobilenet_v2_1_0_160'))