Update semantic_segmentation.md (#26419)

docs/source/en/tasks/semantic_segmentation.md

@@ -206,7 +206,7 @@ The transform is applied on the fly which is faster and consumes less disk space
 
 ## Evaluate
 
-Including a metric during training is often helpful for evaluating your model's performance. You can quickly load a evaluation method with the 🤗 [Evaluate](https://huggingface.co/docs/evaluate/index) library. For this task, load the [mean Intersection over Union](https://huggingface.co/spaces/evaluate-metric/accuracy) (IoU) metric (see the 🤗 Evaluate [quick tour](https://huggingface.co/docs/evaluate/a_quick_tour) to learn more about how to load and compute a metric):
+Including a metric during training is often helpful for evaluating your model's performance. You can quickly load an evaluation method with the 🤗 [Evaluate](https://huggingface.co/docs/evaluate/index) library. For this task, load the [mean Intersection over Union](https://huggingface.co/spaces/evaluate-metric/accuracy) (IoU) metric (see the 🤗 Evaluate [quick tour](https://huggingface.co/docs/evaluate/a_quick_tour) to learn more about how to load and compute a metric):
 
 ```py
 >>> import evaluate