llava batch infer, only the result corresponding to the longest prompt is correct, while other results are incorrect

[lss15151161]

version: TensorRT-LLM 0.10.0
the official script(TensorRT-LLM/examples/multimodal/run.py) use same prompt repeat to form a batch. but if I use different prompts to form a batch, the result is incorrect. how to solve it?
because the result corresponding to the longest prompt is correct, I think the reason is padding.

if i use the same prompts, the result is correct

[amukkara]

@lss15151161 This example does not support using different prompts in a batch.
Yes, the issue is pad tokens will added to end of shorter post_prompt when prompts are different.

[lss15151161]

@lss15151161 This example does not support using different prompts in a batch. Yes, the issue is pad tokens will added to end of shorter post_prompt when prompts are different.

thx for reply~
so, do you know what should I do if I want to do batch inference?

[lss15151161]

@lss15151161 This example does not support using different prompts in a batch. Yes, the issue is pad tokens will added to end of shorter post_prompt when prompts are different.

@lss15151161 This example does not support using different prompts in a batch. Yes, the issue is pad tokens will added to end of shorter post_prompt when prompts are different.

and doesn't trtllm remove pads internally?