Source code for pytext.metrics.language_model_metrics

#!/usr/bin/env python3
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved

import math
from typing import NamedTuple

Language model metric utilities.

[docs]class LanguageModelMetric(NamedTuple): """ Class for language model metrics. Attributes: perplexity_per_word: Average perplexity per word of the dataset. """ perplexity_per_word: float
[docs] def print_metrics(self): print(f"Perplexity per word : {self.perplexity_per_word: 0.2f}")
[docs]def compute_language_model_metric(loss_per_word: float) -> LanguageModelMetric: return LanguageModelMetric(perplexity_per_word=math.exp(loss_per_word))