{ "config": { "model": "llama-3.2-3b-instruct", "host": "172.27.128.1", "port": 1269, "timeoutSec": 60, "maxTokens": 12000, "context": 32768, "outputCap": 12000, "reasoningCounted": true }, "rows": [ { "n": 1, "wallMs": 60008, "totalTokens": 5696, "totalReasoningTokens": 0, "totalAllTokens": 5696, "combinedTokPerSec": 95, "combinedTokPerSecAll": 95, "meanAgentTokPerSec": 96.1, "meanAgentTokPerSecAll": 97.6, "ttftMeanMs": 117, "ttftMaxMs": 117, "ok": 0, "timeout": 1, "errors": 0, "scaleEffPct": 100, "perAgentRelPct": 103 }, { "n": 2, "wallMs": 60032, "totalTokens": 9131, "totalReasoningTokens": 0, "totalAllTokens": 9131, "combinedTokPerSec": 152, "combinedTokPerSecAll": 152, "meanAgentTokPerSec": 103.1, "meanAgentTokPerSecAll": 105.1, "ttftMeanMs": 118, "ttftMaxMs": 128, "ok": 0, "timeout": 2, "errors": 0, "scaleEffPct": 160, "perAgentRelPct": 111 }, { "n": 3, "wallMs": 60035, "totalTokens": 10791, "totalReasoningTokens": 0, "totalAllTokens": 10791, "combinedTokPerSec": 180, "combinedTokPerSecAll": 180, "meanAgentTokPerSec": 84.1, "meanAgentTokPerSecAll": 85.9, "ttftMeanMs": 2312, "ttftMaxMs": 2322, "ok": 0, "timeout": 3, "errors": 0, "scaleEffPct": 189, "perAgentRelPct": 90 }, { "n": 4, "wallMs": 60037, "totalTokens": 12879, "totalReasoningTokens": 0, "totalAllTokens": 12879, "combinedTokPerSec": 215, "combinedTokPerSecAll": 215, "meanAgentTokPerSec": 98.0, "meanAgentTokPerSecAll": 98.7, "ttftMeanMs": 2002, "ttftMaxMs": 2037, "ok": 0, "timeout": 4, "errors": 0, "scaleEffPct": 226, "perAgentRelPct": 104 }, { "n": 5, "wallMs": 60036, "totalTokens": 13437, "totalReasoningTokens": 0, "totalAllTokens": 13437, "combinedTokPerSec": 224, "combinedTokPerSecAll": 224, "meanAgentTokPerSec": 76.6, "meanAgentTokPerSecAll": 76.9, "ttftMeanMs": 5387, "ttftMaxMs": 5423, "ok": 0, "timeout": 5, "errors": 0, "scaleEffPct": 236, "perAgentRelPct": 81 }, { "n": 6, "wallMs": 60040, "totalTokens": 14751, "totalReasoningTokens": 0, "totalAllTokens": 14751, "combinedTokPerSec": 246, "combinedTokPerSecAll": 246, "meanAgentTokPerSec": 73.0, "meanAgentTokPerSecAll": 73.0, "ttftMeanMs": 5425, "ttftMaxMs": 5446, "ok": 0, "timeout": 6, "errors": 0, "scaleEffPct": 259, "perAgentRelPct": 77 }, { "n": 7, "wallMs": 60038, "totalTokens": 15621, "totalReasoningTokens": 0, "totalAllTokens": 15621, "combinedTokPerSec": 260, "combinedTokPerSecAll": 260, "meanAgentTokPerSec": 76.8, "meanAgentTokPerSecAll": 76.8, "ttftMeanMs": 5789, "ttftMaxMs": 5813, "ok": 0, "timeout": 7, "errors": 0, "scaleEffPct": 274, "perAgentRelPct": 81 }, { "n": 8, "wallMs": 60040, "totalTokens": 16191, "totalReasoningTokens": 0, "totalAllTokens": 16191, "combinedTokPerSec": 270, "combinedTokPerSecAll": 270, "meanAgentTokPerSec": 54.6, "meanAgentTokPerSecAll": 54.6, "ttftMeanMs": 8652, "ttftMaxMs": 8679, "ok": 0, "timeout": 8, "errors": 0, "scaleEffPct": 284, "perAgentRelPct": 57 }, { "n": 9, "wallMs": 60043, "totalTokens": 17569, "totalReasoningTokens": 0, "totalAllTokens": 17569, "combinedTokPerSec": 293, "combinedTokPerSecAll": 293, "meanAgentTokPerSec": 53.0, "meanAgentTokPerSecAll": 53.0, "ttftMeanMs": 9637, "ttftMaxMs": 9690, "ok": 0, "timeout": 9, "errors": 0, "scaleEffPct": 308, "perAgentRelPct": 56 }, { "n": 10, "wallMs": 60041, "totalTokens": 18737, "totalReasoningTokens": 0, "totalAllTokens": 18737, "combinedTokPerSec": 312, "combinedTokPerSecAll": 312, "meanAgentTokPerSec": 56.5, "meanAgentTokPerSecAll": 56.5, "ttftMeanMs": 11195, "ttftMaxMs": 11262, "ok": 0, "timeout": 10, "errors": 0, "scaleEffPct": 328, "perAgentRelPct": 59 }, { "n": 11, "wallMs": 60043, "totalTokens": 19319, "totalReasoningTokens": 0, "totalAllTokens": 19319, "combinedTokPerSec": 322, "combinedTokPerSecAll": 322, "meanAgentTokPerSec": 52.1, "meanAgentTokPerSecAll": 52.1, "ttftMeanMs": 11290, "ttftMaxMs": 11347, "ok": 0, "timeout": 11, "errors": 0, "scaleEffPct": 339, "perAgentRelPct": 55 }, { "n": 12, "wallMs": 60042, "totalTokens": 19959, "totalReasoningTokens": 0, "totalAllTokens": 19959, "combinedTokPerSec": 332, "combinedTokPerSecAll": 332, "meanAgentTokPerSec": 42.5, "meanAgentTokPerSecAll": 42.5, "ttftMeanMs": 12112, "ttftMaxMs": 12174, "ok": 0, "timeout": 12, "errors": 0, "scaleEffPct": 349, "perAgentRelPct": 45 }, { "n": 13, "wallMs": 60044, "totalTokens": 20456, "totalReasoningTokens": 0, "totalAllTokens": 20456, "combinedTokPerSec": 341, "combinedTokPerSecAll": 341, "meanAgentTokPerSec": 46.4, "meanAgentTokPerSecAll": 46.4, "ttftMeanMs": 12941, "ttftMaxMs": 12994, "ok": 0, "timeout": 13, "errors": 0, "scaleEffPct": 359, "perAgentRelPct": 49 }, { "n": 14, "wallMs": 60042, "totalTokens": 21708, "totalReasoningTokens": 0, "totalAllTokens": 21708, "combinedTokPerSec": 362, "combinedTokPerSecAll": 362, "meanAgentTokPerSec": 45.6, "meanAgentTokPerSecAll": 45.6, "ttftMeanMs": 12768, "ttftMaxMs": 12827, "ok": 0, "timeout": 14, "errors": 0, "scaleEffPct": 381, "perAgentRelPct": 48 }, { "n": 15, "wallMs": 60043, "totalTokens": 21301, "totalReasoningTokens": 0, "totalAllTokens": 21301, "combinedTokPerSec": 355, "combinedTokPerSecAll": 355, "meanAgentTokPerSec": 35.0, "meanAgentTokPerSecAll": 35.0, "ttftMeanMs": 14205, "ttftMaxMs": 14356, "ok": 0, "timeout": 15, "errors": 0, "scaleEffPct": 374, "perAgentRelPct": 37 }, { "n": 16, "wallMs": 60046, "totalTokens": 22483, "totalReasoningTokens": 0, "totalAllTokens": 22483, "combinedTokPerSec": 374, "combinedTokPerSecAll": 374, "meanAgentTokPerSec": 34.9, "meanAgentTokPerSecAll": 34.9, "ttftMeanMs": 14963, "ttftMaxMs": 15039, "ok": 0, "timeout": 16, "errors": 0, "scaleEffPct": 394, "perAgentRelPct": 37 }, { "n": 17, "wallMs": 60046, "totalTokens": 20385, "totalReasoningTokens": 0, "totalAllTokens": 20385, "combinedTokPerSec": 339, "combinedTokPerSecAll": 339, "meanAgentTokPerSec": 29.2, "meanAgentTokPerSecAll": 29.2, "ttftMeanMs": 16215, "ttftMaxMs": 16311, "ok": 0, "timeout": 17, "errors": 0, "scaleEffPct": 357, "perAgentRelPct": 31 }, { "n": 18, "wallMs": 60046, "totalTokens": 22609, "totalReasoningTokens": 0, "totalAllTokens": 22609, "combinedTokPerSec": 377, "combinedTokPerSecAll": 377, "meanAgentTokPerSec": 28.1, "meanAgentTokPerSecAll": 28.1, "ttftMeanMs": 13247, "ttftMaxMs": 13361, "ok": 0, "timeout": 18, "errors": 0, "scaleEffPct": 397, "perAgentRelPct": 30 }, { "n": 19, "wallMs": 60047, "totalTokens": 22781, "totalReasoningTokens": 0, "totalAllTokens": 22781, "combinedTokPerSec": 379, "combinedTokPerSecAll": 379, "meanAgentTokPerSec": 27.3, "meanAgentTokPerSecAll": 27.3, "ttftMeanMs": 13783, "ttftMaxMs": 13930, "ok": 0, "timeout": 19, "errors": 0, "scaleEffPct": 399, "perAgentRelPct": 29 }, { "n": 20, "wallMs": 60047, "totalTokens": 22689, "totalReasoningTokens": 0, "totalAllTokens": 22689, "combinedTokPerSec": 378, "combinedTokPerSecAll": 378, "meanAgentTokPerSec": 25.9, "meanAgentTokPerSecAll": 25.9, "ttftMeanMs": 15803, "ttftMaxMs": 15976, "ok": 0, "timeout": 20, "errors": 0, "scaleEffPct": 398, "perAgentRelPct": 27 }, { "n": 21, "wallMs": 60046, "totalTokens": 23067, "totalReasoningTokens": 0, "totalAllTokens": 23067, "combinedTokPerSec": 384, "combinedTokPerSecAll": 384, "meanAgentTokPerSec": 25.2, "meanAgentTokPerSecAll": 25.2, "ttftMeanMs": 15645, "ttftMaxMs": 15824, "ok": 0, "timeout": 21, "errors": 0, "scaleEffPct": 404, "perAgentRelPct": 27 }, { "n": 22, "wallMs": 60051, "totalTokens": 23090, "totalReasoningTokens": 0, "totalAllTokens": 23090, "combinedTokPerSec": 385, "combinedTokPerSecAll": 385, "meanAgentTokPerSec": 24.3, "meanAgentTokPerSecAll": 24.3, "ttftMeanMs": 16679, "ttftMaxMs": 16873, "ok": 0, "timeout": 22, "errors": 0, "scaleEffPct": 405, "perAgentRelPct": 26 }, { "n": 23, "wallMs": 60045, "totalTokens": 24114, "totalReasoningTokens": 0, "totalAllTokens": 24114, "combinedTokPerSec": 402, "combinedTokPerSecAll": 402, "meanAgentTokPerSec": 24.0, "meanAgentTokPerSecAll": 24.0, "ttftMeanMs": 15453, "ttftMaxMs": 15666, "ok": 0, "timeout": 23, "errors": 0, "scaleEffPct": 423, "perAgentRelPct": 25 }, { "n": 24, "wallMs": 60049, "totalTokens": 24255, "totalReasoningTokens": 0, "totalAllTokens": 24255, "combinedTokPerSec": 404, "combinedTokPerSecAll": 404, "meanAgentTokPerSec": 23.1, "meanAgentTokPerSecAll": 23.1, "ttftMeanMs": 16062, "ttftMaxMs": 16268, "ok": 0, "timeout": 24, "errors": 0, "scaleEffPct": 425, "perAgentRelPct": 24 } ] }