apyskir

Hi, I have a proposal: let's make it possible to dump adapted input of a step to disk. It's very handy when you are working on a 5th or 10th...

Hi! Let's assume I'm training a huge network for a lot of epochs and it saves checkpoints in `checkpoints` folder. I suggest to prepare a possibility to run `transform` on...

When model outputs a single tensor, [this](https://github.com/neptune-ml/steppy-toolkit/blob/master/toolkit/pytorch_transformers/models.py#L96) line gets `len(outputs_batch)` equal to first dimension of the tensor, but it should be 1. On the other hand if model outputs a...

Hi, these 2 lines cannot work: https://github.com/neptune-ml/steppy-toolkit/blob/master/toolkit/pytorch_transformers/architectures/unet.py#L251 https://github.com/neptune-ml/steppy-toolkit/blob/master/toolkit/pytorch_transformers/architectures/unet.py#L299 There is no place, where `self.conv_stride` is declared. These UNets work only on assumption, that stride of convolutional layers is equal to...

Hi, while trying to use UNet from scratch I get `ImportError: cannot import name 'get_downsample_pad'`. This line https://github.com/neptune-ml/steppy-toolkit/blob/master/toolkit/pytorch_transformers/architectures/unet.py#L6: `from toolkit.pytorch_transformers.utils import get_downsample_pad, get_upsample_pad` should probably be changed to: `from toolkit.pytorch_transformers.architectures.utils...