Langdonholmes Cohen Weighted Kappa

---
name: langdonholmes-cohen-weighted-kappa
description: Compute langdonholmes/cohen_weighted_kappa via the HuggingFace `evaluate` library. Use when the user has predictions + references and wants the canonical implementation of langdonholmes/cohen_weighted_kappa.
metadata:
  skill_kind: metric
  source_lib: huggingface-evaluate
  hf_module: langdonholmes/cohen_weighted_kappa
  source: library_introspection
---

# langdonholmes-cohen-weighted-kappa

> Metric `langdonholmes/cohen_weighted_kappa` from the HuggingFace `evaluate` library.

## When to invoke

User asks to compute `langdonholmes/cohen_weighted_kappa` or wants HF evaluate's canonical version.

## Recipe

```python
import evaluate
metric = evaluate.load("langdonholmes/cohen_weighted_kappa")
result = metric.compute(predictions=preds, references=refs)
print(result)
```

## Don'ts

- Don't assume your in-house `langdonholmes/cohen_weighted_kappa` matches HF — version conventions vary.
- Many evaluate metrics have task-specific arguments (`average=`, `lang=`, `model_type=`); read the metric card before reporting numbers.