Amit S

Does removing ```[0]``` from the this line ```best_der_idx = np.argmax(derivatives[n_skip_beginning:-n_skip_end])[0]``` from [get_best_lr](https://github.com/surmenok/keras_lr_finder/blob/f1013c997080792259478a25908c80c08a2f2fbd/keras_lr_finder/lr_finder.py#L145) method work? I tried and I didn't get any errors. Will it still return the best learning rate?

Here is a trick you can do to understand if selenium is necessary: try running it without the selenium if it runs without any problems then you don't need selenium,...

It uses accuracy during training, while evaluating in `cinc17_eval.ipynb` it uses f1 score.

use opencv dnn to load the model.

> > use opencv dnn to load the model. > > Hi @amitjslearn could you please show us your CMakeLists for using CPU? > > Thank you! If you have...

They don't have good documentation about how to perceive the model's output, so I wasn't able to make anything out of it. You can try to localize the face first...

> I am not being able to load the model with opencv either. Please, comment with the code you are using to load the model, then someone will be able...