__[Slides]__ - [here](https://yadi.sk/i/eyCRxQHx3NH2dg)

- Russian: [lecture I](https://yadi.sk/i/vNQYuGJP3FzDPU), [lecture II](https://yadi.sk/i/0IUoVMfI3GJcqb), [seminar](https://yadi.sk/i/YrGhi-803FzDTp)

- English: [lecture I](https://www.youtube.com/watch?v=GxZrEKZfW2o), [lecture II](https://www.youtube.com/watch?v=pA4BsUK3oP4)

- http://cs231n.github.io/transfer-learning/

- lasagne [recipes](https://github.com/Lasagne/Recipes)

- [a few words on soft-targets](http://www.kdnuggets.com/2015/05/dark-knowledge-neural-network.html)

- If you want to learn more about graphical models - [first video](https://www.youtube.com/watch?v=uAsys22y5mY)

- **PyTorch** [finetuning tutorial](https://pytorch.org/tutorials/beginner/finetuning_torchvision_models_tutorial.html)

Your goal for this week is to use pre-trained convolutuional network from the model zoo to classify cats versus dogs.

As usual, follow the instructions in seminar_lasagne or seminar_tf depending on your framework.