o }oi @sFddlZddlmmZda d dejdejfddZd d d ZdS) NlogitstargetscCstj|||ddS)a Computes the cross-entropy loss between logits and targets. Args: logits (torch.Tensor): Model predictions of shape (sequence_length, num_classes). targets (torch.Tensor): Ground-truth labels of shape (sequence_length,). ignore_index (int, optional): Target value that is ignored when computing the loss. Defaults to -100. Returns: torch.Tensor: The sum of cross-entropy losses over the sequence. sum) ignore_index reduction)F cross_entropyfloat)rrrr R/home/ubuntu/.local/lib/python3.10/site-packages/nemo/automodel/loss/chunked_ce.pycompute_cross_entropysr Tc Cs|j|jkr ||j}|durs