Snippet Collection

Full collection of supported code snippets for the PyTorch and Fastai library. → refers to pressing Tab to trigger the snippet.

PyTorch Snippets

Trigger	Content
`pytorch:imports→`	insert the most common pytorch imports
`pytorch:metric→`	custom metric function
`pytorch:loss_class→`	selection of a classification loss function
`pytorch:loss_reg→`	selection of a regression loss function
`pytorch:init→`	creates an parameter initialization function and applies it to the neural net
`pytorch:optimizer→`	selection of an optimizer
`pytorch:scheduler→`	selection of a learning rate scheduler
`pytorch:sequential→`	builds a sequential network
`pytorch:device→`	check the available device (cpu/gpu)
`pytorch:module→`	creates a pytorch module class
`pytorch:dataset`	Template for a custom pytorch dataset
`pytorch:function→`	creates a pytorch function class
`pytorch:train→`	creates a training loop
`pytorch:freeze→`	freezes all layers of the model
`pytorch:unfreeze→`	unfreeze all layers of the model
`pytorch:container→`	stores modules or parameters in some kind of container
`pytorch:checkpoint→`	load a model from local checkpoint or url
`pytorch:github→`	load a model from a github repo
`pytorch:sampler→`	select a sampler

Trigger	Content
`pytorch:layer:activation→`	selection of a nonlinearity
`pytorch:layer:linear→`	selection of a linear layer
`pytorch:layer:conv→`	selection of a convolutional layer
`pytorch:layer:recurrent→`	selection of a recurrent layer
`pytorch:layer:norm→`	selection of a normalization layer
`pytorch:layer:sparse→`	selection of an embedding layer
`pytorch:layer:pooling→`	selection of a pooling layer
`pytorch:layer:padding→`	selection of a padding layer
`pytorch:layer:dropout→`	selection of a dropout layer
`pytorch:layer:vision→`	selection of a vision layer
`pytorch:layer:distance→`	selection of a distance layer
`pytorch:layer:resnet:block→`	creates a Resnet BasicBlock
`pytorch:layer:resnet:bottleneck→`	creates a Resnet BottleneckBlock

Trigger	Content
`pytorch:F:activation→`	applies a nonlinearity function
`pytorch:F:linear→`	applies a linear function
`pytorch:F:conv→`	applies a convolutional function
`pytorch:F:norm→`	applies a normalization function
`pytorch:F:sparse→`	applies an embedding function
`pytorch:F:pooling→`	applies a pooling function
`pytorch:F:dropout→`	applies dropout
`pytorch:F:vision→`	applies a vision function
`pytorch:F:loss→`	applies a loss function
`pytorch:F:distance→`	applies a distance function