Skip to content

Instantly share code, notes, and snippets.

@nerdalert
Last active April 10, 2025 15:00
Show Gist options
  • Save nerdalert/c59711765002840243cedae00713d046 to your computer and use it in GitHub Desktop.
Save nerdalert/c59711765002840243cedae00713d046 to your computer and use it in GitHub Desktop.
date backend model_id tokenizer_id num_prompts framework request_rate burstiness max_concurrency duration completed total_input_tokens total_output_tokens request_throughput request_goodput: output_throughput total_token_throughput mean_ttft_ms median_ttft_ms std_ttft_ms p99_ttft_ms mean_tpot_ms median_tpot_ms std_tpot_ms p99_tpot_ms mean_itl_ms median_itl_ms std_itl_ms p99_itl_ms
2025-04-10 01:08:42 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 120 vllm 1 1 nan 102.546 120 120000 12000 1.1702 nan 117.02 1287.22 418.944 358.51 117.766 750.236 56.8231 54.1962 11.0061 84.6333 56.8231 37.1144 79.7653 320.912
2025-04-10 01:16:56 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 1200 vllm 10 1 nan 483.386 1200 1200000 120000 2.48249 nan 248.249 2730.74 174863 175294 102874 350122 100.18 77.0298 228.431 627.818 100.166 43.5764 2296.79 466.211
2025-04-10 01:25:57 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 2400 vllm 20 1 nan 529.077 1308 1308000 130800 2.47223 nan 247.223 2719.45 223113 228206 126454 401587 106.944 79.7957 286.196 895.216 106.93 43.643 2870.69 325.799
2025-04-10 01:34:58 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 3600 vllm 30 1 nan 528.253 1308 1308000 130800 2.47609 nan 247.609 2723.7 229363 236590 128320 401989 110.216 77.4306 316.987 1212.2 110.202 43.5953 3168.12 317.885
2025-04-10 01:43:58 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 4200 vllm 35 1 nan 527.925 1308 1308000 130800 2.47763 nan 247.763 2725.39 232598 241346 129180 401945 106.615 77.1056 292.932 1074.84 106.601 43.6504 2932.13 317.435
2025-04-10 01:51:01 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 2000 vllm inf 1 nan 411.181 1017 1017000 101700 2.47337 nan 247.337 2720.7 201785 202555 116776 401101 103.271 76.7273 269.71 817.872 103.259 43.47 2703.82 174.292
2025-04-10 02:34:40 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 120 sgl 1 1 nan 102.572 120 120000 12000 1.16991 nan 116.991 1286.9 414.314 355.465 118.829 767.049 56.7146 54.2361 10.9518 84.6905 56.7146 37.1214 77.1924 320.552
2025-04-10 02:42:55 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 1200 sgl 10 1 nan 483.417 1200 1200000 120000 2.48233 nan 248.233 2730.56 174895 175335 102881 350161 100.205 77.0311 228.438 627.861 100.192 43.5786 2296.88 466.479
2025-04-10 02:49:21 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 2400 sgl 20 1 nan 375.05 902 902000 90200 2.40501 nan 240.501 2645.51 157595 157513 92251.6 312789 82.4376 77.8888 88.1399 116.955 82.4358 43.6637 936.806 325.905
2025-04-10 03:06:31 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 120 sgl 1 1 nan 102.575 120 120000 12000 1.16987 nan 116.987 1286.86 411.621 351.711 117.848 760.864 56.6146 54.1877 10.958 84.8155 56.6146 37.086 76.7083 321.25
2025-04-10 03:14:46 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 1200 sgl 10 1 nan 483.418 1200 1200000 120000 2.48232 nan 248.232 2730.56 174889 175324 102880 350160 100.202 77.0516 228.439 627.851 100.189 43.5777 2296.87 466.509
2025-04-10 03:23:47 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 2400 sgl 20 1 nan 529.097 1308 1308000 130800 2.47214 nan 247.214 2719.35 223124 228232 126451 401526 106.957 79.7714 286.196 895.276 106.943 43.6914 2870.68 325.283
2025-04-10 03:32:47 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 3600 sgl 30 1 nan 528.263 1308 1308000 130800 2.47604 nan 247.604 2723.65 229370 236604 128318 401994 110.222 77.2302 316.996 1214.58 110.208 43.6342 3168.11 318.624
2025-04-10 03:41:48 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 4200 sgl 35 1 nan 527.982 1308 1308000 130800 2.47736 nan 247.736 2725.09 232607 241355 129183 401978 106.643 77.1012 292.979 1075.34 106.629 43.6526 2932.5 318.157
2025-04-10 03:48:50 vllm meta-llama/Llama-3.1-8B-Instruct meta-llama/Llama-3.1-8B-Instruct 2000 sgl inf 1 nan 411.147 1017 1017000 101700 2.47357 nan 247.357 2720.93 201752 202539 116795 401080 103.388 76.9119 269.689 817.844 103.375 43.5 2703.83 175.695
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment