The practice of doing performance analysis/optimization with TensorRT-LLM

Описание: Learn best practices on TensorRT-LLM performance analysis and optimization. Hear from our experts on the analysis the performance of TensorRT-LLM by ultilizing useful tools, how to read the profile results, understand the performance bottleneck, and how to optimize the performance.

Links to the slides in this session can be found here: https://nvdam.widen.net/s/5rt2zqnqnh/...