Compare kernel distribution before and after annotation.

manually process each sample with per sample gradient

Create a conv+transformer model for benchmarking.

Convert a batch of tensors to their original scale. Args: batch (torch.Tensor): Batch of normalized tensors. mean (torch.Ten

Prunes tensor corresponding to parameter called `name` in `module` by removing every other entry in the tensors. Modifies module in place (and

Gets the stats of an environment.

takes torch.benchmark objects and compares delta of second vs first.

Initialize a NCCL group for one rank of the spawned demo.

Update online action value (Q) function with a batch of experiences

Mark end of episode

Sample experiences from memory

Run the style transfer.

Training with the normal model.