{ "label": "", "model": "sensenova-loop-bf16", "base_url": "http://127.0.0.1:30080/v1", "host": "tux", "time": "2026-10-30T15:47:00-0700", "config": { "base_url": "http://127.0.0.1:30080/v1", "model": "sensenova-loop-bf16", "concurrency": "1,2,4,8", "num_requests": 16, "warmup": 2, "input_tokens": 1024, "output_tokens": 256, "tokenizer": "upstream/SenseNova-Loop-9B-Preview/tokenizer.json", "allow_prefix_cache": false, "no_ignore_eos": false, "timeout": 600, "label": "", "out": "eval/runs/perf_sglang_bf16.json", "prompt_text_tokens": 1024, "prompt_build": "tokenizer.json" }, "levels": [ { "concurrency": 1, "n_requests": 16, "n_ok": 16, "n_errors": 0, "errors": [], "wall_s": 712.4458575863391, "ttft_s": { "p50": 0.7494839327409863, "p90": 0.752388886641711, "mean": 0.7490631011896767 }, "e2e_s": { "p50": 44.54232191387564, "p90": 44.74176003783941, "mean": 44.52777633216465 }, "decode_tok_s_per_req": { "p50": 5.82298062979468, "p90": 5.857639877869307, "mean": 5.8248643173510715 }, "itl_s": { "p50": 0.17173366795531383, "p90": 0.17253968634587877, "mean": 0.17168066543298682 }, "output_tokens_mean": 256.0, "output_tokens_min": 256, "output_tokens_max": 256, "prompt_tokens_mean": 1051.0, "n_short_outputs": 0, "aggregate_output_tok_s": 5.7492088084793425, "requests_per_s": 0.02245784690812243, "usage_reported": true }, { "concurrency": 2, "n_requests": 16, "n_ok": 16, "n_errors": 0, "errors": [], "wall_s": 358.13540337421, "ttft_s": { "p50": 1.4381381664425135, "p90": 1.493556933477521, "mean": 1.451384060550481 }, "e2e_s": { "p50": 44.76612635422498, "p90": 45.01789597328752, "mean": 44.76674999552779 }, "decode_tok_s_per_req": { "p50": 5.888547972827849, "p90": 5.919728871447871, "mean": 5.8871749230165396 }, "itl_s": { "p50": 0.16982132055997556, "p90": 0.1708354555380841, "mean": 0.16986371961846838 }, "output_tokens_mean": 256.0, "output_tokens_min": 256, "output_tokens_max": 256, "prompt_tokens_mean": 1051.0, "n_short_outputs": 0, "aggregate_output_tok_s": 11.43701505466678, "requests_per_s": 0.04467584005729211, "usage_reported": true }, { "concurrency": 4, "n_requests": 16, "n_ok": 16, "n_errors": 0, "errors": [], "wall_s": 187.37603424675763, "ttft_s": { "p50": 2.734584681224078, "p90": 2.7569230226799846, "mean": 2.7020455568563193 }, "e2e_s": { "p50": 46.840107245370746, "p90": 46.998931530397385, "mean": 46.84355494793272 }, "decode_tok_s_per_req": { "p50": 5.772806069829697, "p90": 5.803642233569628, "mean": 5.776986593431426 }, "itl_s": { "p50": 0.17322622279947003, "p90": 0.1738359775942038, "mean": 0.1731030160166305 }, "output_tokens_mean": 256.0, "output_tokens_min": 256, "output_tokens_max": 256, "prompt_tokens_mean": 1051.0, "n_short_outputs": 0, "aggregate_output_tok_s": 21.859785945762578, "requests_per_s": 0.08538978885063507, "usage_reported": true }, { "concurrency": 8, "n_requests": 16, "n_ok": 16, "n_errors": 0, "errors": [], "wall_s": 102.59108853153884, "ttft_s": { "p50": 4.974256161600351, "p90": 5.0004629092291, "mean": 4.947074577212334 }, "e2e_s": { "p50": 51.294219580013305, "p90": 51.56152600981295, "mean": 51.29352270014351 }, "decode_tok_s_per_req": { "p50": 5.488342640760974, "p90": 5.537196663306641, "mean": 5.502280519349593 }, "itl_s": { "p50": 0.18220517470811803, "p90": 0.18259358133244163, "mean": 0.18174918959711622 }, "output_tokens_mean": 256.0, "output_tokens_min": 256, "output_tokens_max": 256, "prompt_tokens_mean": 1051.0, "n_short_outputs": 0, "aggregate_output_tok_s": 39.925495075927536, "requests_per_s": 0.15595896514034194, "usage_reported": true } ] }