replace get_model() with self.model and rm pytorch dir.

README.md

@@ -14,6 +14,7 @@ __init__(self, model_name: str = 'efficientnet-b7', framework: str = 'pytorch',
 
 **Args:**
 
+
 - model_name:
   - the model name for embedding
   - supported types: `str`, for example 'efficientnet-b7'

efficientnet_image_embedding.py

@@ -11,23 +11,20 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-
+import timm
 from typing import NamedTuple
-from PIL import Image
 import torch
-from torchvision import transforms
-import sys
 import towhee
-from pathlib import Path
 import numpy
+import torch.nn as nn
 
-from towhee.operator import Operator
+from towhee.operator import NNOperator
 from towhee.utils.pil_utils import to_pil
 from timm.data import resolve_data_config
 from timm.data.transforms_factory import create_transform
-import os
 
-class EfficientnetImageEmbedding(Operator):
+
+class EfficientnetImageEmbedding(NNOperator):
     """
     Embedding extractor using efficientnet.
     Args:
@@ -37,22 +34,28 @@ class EfficientnetImageEmbedding(Operator):
             Path to local weights.
     """
 
-    def __init__(self, model_name: str = '', framework: str = 'pytorch', weights_path: str = None) -> None:
+    def __init__(self, model_name: str = '', num_classes: int = 1000, framework: str = 'pytorch',
+                 weights_path: str = None) -> None:
+        super().__init__(framework=framework)
         model_name = model_name.replace('efficientnet-b', 'tf_efficientnet_b')
-        super().__init__()
-        if framework == 'pytorch':
-            import importlib.util
-            path = os.path.join(str(Path(__file__).parent), 'pytorch', 'model.py')
-            opname = os.path.basename(str(Path(__file__))).split('.')[0]
-            spec = importlib.util.spec_from_file_location(opname, path)
-            module = importlib.util.module_from_spec(spec)
-            spec.loader.exec_module(module)
-        self.model = module.Model(model_name, weights_path)
-        config = resolve_data_config({}, model=self.model._model)
+        self.device = 'cuda' if torch.cuda.is_available() else 'cpu'
+        if weights_path:
+            self.model = timm.create_model(model_name, checkpoint_path=weights_path, num_classes=num_classes)
+        else:
+            self.model = timm.create_model(model_name, pretrained=True, num_classes=num_classes)
+        self.model.eval()
+        config = resolve_data_config({}, model=self.model)
         self.tfms = create_transform(**config)
 
     def __call__(self, image: 'towhee.types.Image') -> NamedTuple('Outputs', [('feature_vector', numpy.ndarray)]):
         Outputs = NamedTuple('Outputs', [('feature_vector', numpy.ndarray)])
-        img = self.tfms(to_pil(image)).unsqueeze(0)
-        features = self.model(img)
-        return Outputs(features.flatten().detach().numpy())
+        self.model.to(self.device)
+        self.model.eval()
+        img_tensor = self.tfms(to_pil(image)).unsqueeze(0)
+        features = self.model.forward_features(img_tensor)
+        if features.dim() == 4:  # if the shape of feature map is [N, C, H, W], where H > 1 and W > 1
+            global_pool = nn.AdaptiveAvgPool2d(1)
+            features = global_pool(features)
+        features = features.to('cpu')
+        features = features.flatten().detach().numpy()
+        return Outputs(features)

pytorch/__init__.py

@@ -1,22 +0,0 @@
-# Copyright 2021 Zilliz. All rights reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
-
-import os
-
-# For requirements.
-try:
-    import efficientnet_pytorch
-except ModuleNotFoundError:
-    os.system('pip install efficientnet_pytorch')

pytorch/model.py

@@ -1,39 +0,0 @@
-# Copyright 2021 Zilliz. All rights reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
-
-import torch
-import timm
-
-
-class Model():
-    """
-    PyTorch model class
-    """
-    def __init__(self, model_name: str, weights_path: str):
-        super().__init__()
-        if weights_path:
-            self._model = timm.create_model(model_name, checkpoint_path=weights_path, num_classes=0)
-        else:
-            self._model = timm.create_model(model_name, pretrained=True, num_classes=0)
-        self._model.eval()
-
-    def __call__(self, img_tensor: torch.Tensor):
-        return self._model(img_tensor)
-
-    def train(self):
-        """
-        For training model
-        """
-        pass