TensorRT-LLM
TensorRT-LLM copied to clipboard
Custom sampler for executor API
Hi,
Is it currently possible to write a custom sampler to use with the executor API? I want to sample tokens in way that is not supported by the current API and exiting the generation loop at each step to do so slows things down too much
@maxilevi , do you mind to share more details (prefer sample code for illustration) on why it is not supported by current executor API and what is your request in details? thanks!
Issue has not received an update in over 14 days. Adding stale label.
This issue was closed because it has been 14 days without activity since it has been marked as stale.