Example Plots#

Here are the list of sample plots that gets created by default from running the genai-perf with --generate-plots flag:

Distribution of Input Sequence Lengths to Output Sequence Lengths#

../../../_images/distribution_of_input_sequence_lengths_to_output_sequence_lengths.jpeg

Request Latency Analysis#

../../../_images/request_latency.jpeg

Time to First Token Analysis#

../../../_images/time_to_first_token.jpeg

Time to First Token vs. Input Sequence Lengths#

../../../_images/time_to_first_token_vs_input_sequence_lengths.jpeg

Token-to-Token Latency vs. Output Token Position#

../../../_images/token-to-token_latency_vs_output_token_position.jpeg