Aurelien Chartier
Aurelien Chartier
Once you have a TRT-LLM docker image, you could build Triton inside and replace the locally installed version. See these lines in the TRT-LLM docker build: https://github.com/NVIDIA/TensorRT-LLM/blob/main/docker/Dockerfile.multi#L99-L103 I'm not familiar...
This docker is not used anymore, thanks for pointing it out. It has been removed in #813