summaryrefslogtreecommitdiffstats
path: root/linguistics-components
diff options
context:
space:
mode:
Diffstat (limited to 'linguistics-components')
-rw-r--r--linguistics-components/src/main/java/com/yahoo/language/huggingface/Encoding.java3
-rw-r--r--linguistics-components/src/main/java/com/yahoo/language/huggingface/HuggingFaceTokenizer.java2
2 files changed, 5 insertions, 0 deletions
diff --git a/linguistics-components/src/main/java/com/yahoo/language/huggingface/Encoding.java b/linguistics-components/src/main/java/com/yahoo/language/huggingface/Encoding.java
index ddb098c911d..107900ff73c 100644
--- a/linguistics-components/src/main/java/com/yahoo/language/huggingface/Encoding.java
+++ b/linguistics-components/src/main/java/com/yahoo/language/huggingface/Encoding.java
@@ -2,6 +2,8 @@
package com.yahoo.language.huggingface;
+import com.yahoo.api.annotations.Beta;
+
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
@@ -9,6 +11,7 @@ import java.util.List;
/**
* @author bjorncs
*/
+@Beta
public record Encoding(
List<Long> ids, List<Long> typeIds, List<String> tokens, List<Long> wordIds, List<Long> attentionMask,
List<Long> specialTokenMask, List<CharSpan> charTokenSpans, List<Encoding> overflowing) {
diff --git a/linguistics-components/src/main/java/com/yahoo/language/huggingface/HuggingFaceTokenizer.java b/linguistics-components/src/main/java/com/yahoo/language/huggingface/HuggingFaceTokenizer.java
index 56fba370470..dd53bd1c695 100644
--- a/linguistics-components/src/main/java/com/yahoo/language/huggingface/HuggingFaceTokenizer.java
+++ b/linguistics-components/src/main/java/com/yahoo/language/huggingface/HuggingFaceTokenizer.java
@@ -2,6 +2,7 @@
package com.yahoo.language.huggingface;
+import com.yahoo.api.annotations.Beta;
import com.yahoo.component.AbstractComponent;
import com.yahoo.component.annotation.Inject;
import com.yahoo.language.Language;
@@ -25,6 +26,7 @@ import static com.yahoo.yolean.Exceptions.uncheck;
*
* @author bjorncs
*/
+@Beta
public class HuggingFaceTokenizer extends AbstractComponent implements Embedder, Segmenter, AutoCloseable {
private final Map<Language, ai.djl.huggingface.tokenizers.HuggingFaceTokenizer> models = new EnumMap<>(Language.class);