Module `lib.loss`

Expand source code

from jax import Array
import jax.numpy as jnp
import optax

def cross_entropy_loss(logits: Array, labels: Array, *, mask: Array) -> Array:
    '''
    Computes the mean softmax cross entropy loss for the provided logits and labels, only considering elements where the mask is True.

    Args:
        logits (Array): The model's predictions. Typically an array of shape (batch_size, seq_len, vocab_size).
        labels (Array): The true labels. Typically an array of shape (batch_size, seq_len).
        mask (Array): Typically a boolean array of shape (batch_size, seq_len). Specifies which elements of the loss array should be included in the mean loss calculation.

    Returns:
        Array: The mean masked softmax cross entropy loss.

    Notes:
        The mask array typically specifies which elements of the sequence are actual tokens as opposed to padding tokens, so the loss is not calculated over padding tokens.
    '''
    loss = optax.softmax_cross_entropy_with_integer_labels(logits=logits, labels=labels)
    return jnp.mean(loss, where=mask)

Functions

def cross_entropy_loss(logits: jax.Array, labels: jax.Array, *, mask: jax.Array) ‑> jax.Array

Computes the mean softmax cross entropy loss for the provided logits and labels, only considering elements where the mask is True.

Args

logits : Array: The model's predictions. Typically an array of shape (batch_size, seq_len, vocab_size).
labels : Array: The true labels. Typically an array of shape (batch_size, seq_len).
mask : Array: Typically a boolean array of shape (batch_size, seq_len). Specifies which elements of the loss array should be included in the mean loss calculation.

Returns

Array: The mean masked softmax cross entropy loss.

Notes

The mask array typically specifies which elements of the sequence are actual tokens as opposed to padding tokens, so the loss is not calculated over padding tokens.

Expand source code

def cross_entropy_loss(logits: Array, labels: Array, *, mask: Array) -> Array:
    '''
    Computes the mean softmax cross entropy loss for the provided logits and labels, only considering elements where the mask is True.

    Args:
        logits (Array): The model's predictions. Typically an array of shape (batch_size, seq_len, vocab_size).
        labels (Array): The true labels. Typically an array of shape (batch_size, seq_len).
        mask (Array): Typically a boolean array of shape (batch_size, seq_len). Specifies which elements of the loss array should be included in the mean loss calculation.

    Returns:
        Array: The mean masked softmax cross entropy loss.

    Notes:
        The mask array typically specifies which elements of the sequence are actual tokens as opposed to padding tokens, so the loss is not calculated over padding tokens.
    '''
    loss = optax.softmax_cross_entropy_with_integer_labels(logits=logits, labels=labels)
    return jnp.mean(loss, where=mask)