A harbor authority lists three needs in order: adapt a foundation model to local notices, then speed up token generation, then put the result behind a live endpoint. Which NVIDIA sequence is that?
Select an answer to reveal the explanation.
Short Explanation
Three needs in order: adapt the model to local notices, speed up token generation, then put it behind a live endpoint. That NVIDIA sequence is NeMo, TensorRT-LLM, then Triton or NIM. RAPIDS, Megatron-as-a-server, and a job queue are the wrong stack.
Full Explanation
Associate stack story: customize with NeMo, speed up inference with TensorRT-LLM, serve with Triton or NIM. RAPIDS, Megatron-as-a-server, and Professional GPU plumbing are the wrong sequence.