Parameter optimization of a pytorch Deep Learning model

vision
1

Hello friends,
I have designed a deep learning model using the PyTorch framework for image classification. I know to use Keras tuner to optimize a deep learning model designed using Keras. Can anyone help me in optimizing the hyperparameters of a deep learning model designed using the Pytorch framework? I have two options

  1. Convert the PyTorch model to a Keras model and use a Keras tuner on it.
  2. Use any hyperparameter optimization method on the PyTorch model.
    Can anyone suggest me and help me with this problem of hyperparameter optimization.
2

HI,
you can use

model= Model()
optimizer = torch.optim.Adam(model.parameters(), lr=0.0001) #ADAM
or
optimizer = torch.optim.SGD(model.parameters(), lr=0.0001) # SGD

Then you can call

optimizer.step()
3

When looking into Hyperparameter optimization you could have a look into “AutoML” techniques which automate the optimization of hyperparamters. One example within PyTorch is Auto-PyTorch, you can find their github repo here: GitHub - automl/Auto-PyTorch: Automatic architecture search and hyperparameter optimization for PyTorch !