![]() Therefore, we must reshape the output ourselves before passing to the output layer, as a Reshape layer is no longer included following the GlobalAveragePooling2D layer. The new version is keeping the output of this layer as ( None, 1, 1, 1024), whereas the old version had a Reshape layer following the pooling layer to reshape the output to ( None, 1024). ![]() In the MobileNet model, the newer version of TensorFlow sets this parameter to True for the GlobalAveragePooling2D layer by default, causing a difference in the output shape shown in the corresponding video. When it is set to True, the output shape of the layer will keep the same number of dimensions from the previous layer. This is the last layer that we grab from the original MobileNet model when constructing our fine-tuned version.The new optional parameter for this layer is called keepdims. On Windows, open the main Settings pane and choose Apps, Apps. If you’re running Backup and Sync or Drive File Stream, get rid of these programs first. In a later version of TensorFlow, a new parameter was introduced to the GlobalAveragePooling2D layer. How to install and use Drive for Desktop.
0 Comments
Leave a Reply. |