Timothy Gerdes

Please add the `--triton-output-path` option to get a triton server log. I would expect an error to be reported there that can help. For example: `--triton_output_path ~/triton_output.log`

Hi, thanks for pointing this out. The docs have been updated. Let me know if there is anything that is unclear.

https://github.com/triton-inference-server/model_analyzer/blob/main/docs/install.md#specific-version-with-local-launch-mode That will build a container based off of the TritonServer container that includes Model Analyzer and Perf Analyzer

Triton can auto-generate a default configuration for you in some cases (depending on the backend). If you don't specify a default config, Model analyzer will attempt to have Triton generate...

Any update on this?

How would we pick up the changes? Will there be a new tag/release any time soon?

@ychuang970303 unfortunately bls models are not supported yet. They are still on our roadmap and will be coming soon

Not yet. We are working on it.

FYI This feature is now available: https://github.com/triton-inference-server/model_analyzer/blob/main/docs/config_search.md#multi-model-search-mode

Hi @SeungsuBaek, there have been some fixes around request rate scheduling over the last few months. I tried to reproduce this issue and everything appears to be working correctly now....