You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I found that trt-llm-dev supports paged KV cache for encoder-decoder models (like Nougat). Does the encoder-decoder model work with inflight batching? Or can we use other batching methods other than static batching to improve performance? Thanks
The text was updated successfully, but these errors were encountered:
@Oldpan for now we only support static batching for encoder-decoder models. But we plan inflight batching support in the coming weeks. Please keep an eye out. Feel free to reopen the issue if you have more questions.
I found that trt-llm-dev supports paged KV cache for encoder-decoder models (like Nougat). Does the encoder-decoder model work with inflight batching? Or can we use other batching methods other than static batching to improve performance? Thanks
The text was updated successfully, but these errors were encountered: