r""" a helper function to call a method on the given RRef

r""" a helper function to call a method on the given RRef

r""" Each trainer runs a forward pass which involves an embedding lookup on the parameter server and running nn.Linear locally. During the bac

apply activation checkpointing to model returns None as model is updated directly

Override of Tracer.call_module (see https://pytorch.org/docs/stable/fx.html#torch.fx.Tracer.call_module). This override:

Create torch.compiled optimizers on each worker

Performs a graph attention layer operation. Args: h (torch.Tensor): Input tensor representing node features.

Performs a forward pass through the network. Args: input_tensor (torch.Tensor): Input tensor representing node features.

forward of both actor and critic

r"""Inputs of forward function Args: x: the sequence fed to the positional encoder model (required). Shape: x:

Forward pass of the attention module. Args: x (torch.Tensor): Input tensor. freqs_cis (torch.Tensor): Precom

Perform a forward pass through the TransformerBlock. Args: x (torch.Tensor): Input tensor. freqs_cis (torch.

Perform a forward pass through the Transformer model. Args: tokens (torch.Tensor): Input token indices. Returns