[ROCm] [CI] Bugfix Existing CI cases (#3946)

Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
This commit is contained in:
TJian authored and GitHub committed 2026-06-01 22:02:15 +08:00
1 parent c067485030
commit 29343540f7
7 files changed
+90 -33

No files matched your search

@@ -53,7 +53,7 @@ python examples/offline_inference/ming_flash_omni/end2end.py \
--image-path ./3_0.png
```
### Omni-Speech (Thinker + Talker)
### Omni-Speech (Thinker + Talker) {#omni-speech-thinker--talker}
The default deploy YAML already runs thinker+talker, so spoken output only requires requesting `audio` (or `text,audio`) modalities.
The thinker processes your multimodal input, generates text, then the talker synthesises the response as speech.