bindings-java : disable flash attention by default (#3445)

This commit disables flash-attention for the Java binding test so that
the testFullTranscribe test passes.

Without this change the test was failing because the expected output
mismatches after the flash-attention change:
```console
<And so my fellow Americans ask not what your country can do for you ask what you can do for your country.>
but was:
<and so my fellow Americans ask not what your country can do for you ask what you can do for your country>
```

An alternative would also be to update the expected output but it felt
better to keep the same expected output and disable flash-attention and
not just change the expected output to match the new behavior.
This commit is contained in:
Daniel Bevenius
2025-10-01 09:13:34 +02:00
committed by GitHub
parent 8c0855fd6b
commit 2a56869669
2 changed files with 5 additions and 2 deletions

View File

@@ -20,7 +20,7 @@ public class WhisperContextParams extends Structure {
/** Use GPU for inference (default = true) */
public CBool use_gpu;
/** Use flash attention (default = false) */
/** Use flash attention (default = true) */
public CBool flash_attn;
/** CUDA device to use (default = 0) */