search

stefanonardo / pytorch-esn

An Echo State Network module for PyTorch.
MIT License
205 stars 43 forks source link

Why is the accuracy in the training set and test set both about 10%? #18

Closed Cro-allen closed 5 months ago

Cro-allen commented 10 months ago

I have not changed any code of mnist in the examples. Why is the accuracy in the training set and test set both about 10%?

stefanonardo commented 9 months ago

@Cro-allen the model is not "trained". I added the code to show an example of working with bigger datasets. If that's what you are asking.

Cro-allen commented 9 months ago

Thank you for your reply, and where can I find classification examples for the bigger datasets?

stefanonardo commented 9 months ago

Not in this repository I'm afraid. You can take a look at my thesis, I trained ESNs on bigger datasets. You find the link in the README.

Shadab442 commented 5 months ago

If I want to improve the accuracy (say around 95%) for classification on the MNIST dataset, do I need to tune the hyperparameters? Or do I need to change anything in the implementation? Can you give some insights about what I may need to improve?