Fix flakey test dependent on model or linux optimized model name

Author: kderusso

x-pack/plugin/inference/src/test/java/org/elasticsearch/xpack/inference/services/elasticsearch/MultilingualE5SmallModelTests.java

@@ -11,7 +11,7 @@
 import org.elasticsearch.test.ESTestCase;
 import org.elasticsearch.xpack.inference.chunking.SentenceBoundaryChunkingSettings;
 
-import static org.hamcrest.Matchers.equalTo;
+import static org.hamcrest.Matchers.containsString;
 
 public class MultilingualE5SmallModelTests extends ESTestCase {
 
@@ -29,7 +29,7 @@ public void testHugeChunkingSettings() {
 
         assertThat(
             expectedException.getMessage(),
-            equalTo(".multilingual-e5-small does not support chunk sizes larger than 300. Requested chunk size: 10000")
+            containsString("does not support chunk sizes larger than 300. Requested chunk size: 10000")
         );
     }
 }