Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[tokenizers] Converting encoding to int32 NDList #3468

Merged
merged 1 commit into from
Sep 12, 2024
Merged

[tokenizers] Converting encoding to int32 NDList #3468

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 9 additions & 3 deletions extensions/tokenizers/src/main/java/ai/djl/huggingface/tokenizers/Encoding.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,8 @@
import ai.djl.ndarray.NDList;
import ai.djl.ndarray.NDManager;

import java.util.Arrays;

/** A class holds token encoding information. */
public class Encoding {

Expand Down Expand Up @@ -58,11 +60,15 @@ protected Encoding(
* @return the {@link NDList}
*/
public NDList toNDList(NDManager manager, boolean withTokenType) {
// Converting encoding to int32 NDList because candle can't convert int64 to fp16 in cuda
NDList list = new NDList(withTokenType ? 3 : 2);
list.add(manager.create(ids));
list.add(manager.create(attentionMask));
int[] intIds = Arrays.stream(ids).mapToInt(i -> (int) i).toArray();
int[] intAttentionMask = Arrays.stream(attentionMask).mapToInt(i -> (int) i).toArray();
list.add(manager.create(intIds));
list.add(manager.create(intAttentionMask));
if (withTokenType) {
list.add(manager.create(typeIds));
int[] intTypeIds = Arrays.stream(typeIds).mapToInt(i -> (int) i).toArray();
list.add(manager.create(intTypeIds));
}
return list;
}
Expand Down
Loading