Bleuscore

---
name: bleuscore
description: Compute the BLEUScore metric — provided by torchmetrics. Use when the user has predictions and ground-truth and needs to compute BLEUScore, or asks how to score with BLEUScore.
metadata:
  skill_kind: metric
  source_lib: torchmetrics
  import_path: torchmetrics.BLEUScore
  source: library_introspection
---

# bleuscore

> Metric `BLEUScore` from `torchmetrics` (torchmetrics.BLEUScore)

## When to invoke this skill

The user has predictions + ground truth and asks to evaluate with BLEUScore, or
mentions `torchmetrics.BLEUScore` directly, or wants the standard torchmetrics implementation.

## Reference signature

```python
from torchmetrics import BLEUScore

# _BLEUScore(n_gram: int = 4, smooth: bool = False, weights: Optional[collections.abc.Sequence[float]] = None, **kwargs: Any) -> None
```

## Library docstring

```
Wrapper for deprecated import.

>>> preds = ['the cat is on the mat']
>>> target = [['there is a cat on the mat', 'a cat is on the mat']]
>>> bleu = _BLEUScore()
>>> bleu(preds, target)
tensor(0.7598)
```

## Quick recipe

```python
import torchmetrics as _m
score = _m.BLEUScore(y_true, y_pred)
```

## Don'ts

- Don't reimplement when the library version handles edge cases (NaN, ties, empty inputs) better than a hand-rolled formula.
- Always check the library version's argument order — sklearn is `(y_true, y_pred)` while torchmetrics is `(preds, target)`.