Dani Azemar comments

Results 4 comments of


                                            Dani Azemar

demo error

@acoder-fin your link to the updated model seems broken. Can you re-upload it, please?

InceptionResnetV2 summary() seems to be another network as it does not look like the one in the paper.

Got an answer from google: https://stackoverflow.com/questions/64488034/inceptionresnetv2-stem-block-keras-implementation-mismatch-the-one-in-the-origin Seems that they were just changing it during internal experiments. Nonetheless, the publishers say there are no difference in performance.

add resnet18 and resnet34

Also agree. If you have all the blocks for ResNet50, ResNet101, ResNet152, why not adding a basic structure as [keras_contrib](https://github.com/keras-team/keras-contrib/blob/3fc5ef709e061416f4bc8a92ca3750c824b5d2b0/keras_contrib/applications/resnet.py#L286) does so you can create your own ResNet?

Training doesn't utilize GPU

[Tensorflow 1.14 is not tested with CUDA 10.1](https://www.tensorflow.org/install/source#tested_build_configurations), so I installed CUDA 10.0, maybe it solves your problems too. Here is what I did in case you need it: First,...