{"traceEvents": [{"name": "process_name", "ph": "M", "pid": 110, "args": {"name": "engine core"}}, {"name": "process_name", "ph": "M", "pid": 31, "args": {"name": "API server"}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 137788232.894, "dur": 24388798.345, "pid": 110, "tid": 1, "args": {"note": "8192 tokens"}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 164731469.034, "dur": 91323.364, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 164844073.078, "dur": 138398.284, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165008904.283, "dur": 110440.018, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165140772.41, "dur": 96099.374, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165260837.701, "dur": 83230.26, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165370724.907, "dur": 253245.781, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165648247.125, "dur": 69921.669, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165744506.399, "dur": 74810.764, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165846842.433, "dur": 65147.69, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 165929508.017, "dur": 94420.959, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 166044893.866, "dur": 84135.994, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 166156331.565, "dur": 70141.31, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 166306176.723, "dur": 504836.446, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 166831690.471, "dur": 76584.972, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 166937296.295, "dur": 100332.656, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167065368.174, "dur": 64379.451, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167149318.968, "dur": 103441.971, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167271985.31, "dur": 61819.241, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167374534.235, "dur": 129815.008, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167524509.112, "dur": 86230.246, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167634760.699, "dur": 59704.549, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167720529.496, "dur": 56411.534, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167798100.181, "dur": 87741.411, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167909863.342, "dur": 56910.375, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 167995123.498, "dur": 58393.057, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 168075697.311, "dur": 66817.374, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 168165606.837, "dur": 262083.883, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 168456162.519, "dur": 251824.552, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 168743092.834, "dur": 70633.043, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 168835607.524, "dur": 62935.261, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 168929995.9, "dur": 136784.395, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169092169.32, "dur": 60210.931, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169177818.639, "dur": 71378.008, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169271415.544, "dur": 64991.809, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169362462.27, "dur": 161474.374, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169549959.6, "dur": 79023.335, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169655515.892, "dur": 64509.498, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169757870.504, "dur": 64834.169, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169850701.121, "dur": 76267.513, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 169953945.629, "dur": 97426.504, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 170077771.162, "dur": 90331.784, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 170189237.186, "dur": 432934.017, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 170651417.63, "dur": 139536.727, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 170814806.357, "dur": 75675.188, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 170914758.081, "dur": 155146.383, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 171091718.156, "dur": 111573.358, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 171227370.238, "dur": 101037.079, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 171351272.716, "dur": 42750.459, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 171411057.459, "dur": 85758.376, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 171514195.81, "dur": 57014.691, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 171586385.308, "dur": 79224.558, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 173009799.922, "dur": 34832.426, "pid": 110, "tid": 1, "args": {"note": "8192 tokens"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 173549993.704, "dur": 259591.796, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 173809627.412, "dur": 234535.531, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 173051915.362, "dur": 1008070.439, "pid": 110, "tid": 1, "args": {"note": "2048 tokens"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 174064205.135, "dur": 80114667.263, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 254188595.657, "dur": 25362.777, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 254214012.078, "dur": 136.248, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 254187839.869, "dur": 41754.164, "pid": 110, "tid": 1, "args": {"note": "1024 tokens"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 254229649.647, "dur": 1803.003, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 254231785.871, "dur": 448.038, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 254232273.881, "dur": 102.057, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 254231619.2, "dur": 14475.306, "pid": 110, "tid": 1, "args": {"note": "2 tokens"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 254246160.481, "dur": 34739.215, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 254281154.692, "dur": 373.343, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 254281561.977, "dur": 98.226, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 254280994.142, "dur": 14492.267, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 254295533.312, "dur": 1807.773, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 254297483.694, "dur": 774.629, "pid": 110, "tid": 1, "args": {"note": "0 tokens"}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 254870151.798, "dur": 170167.394, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255067848.235, "dur": 71881.44, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255157685.179, "dur": 82093.58, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255272138.265, "dur": 79438.324, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255369972.761, "dur": 231369.365, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255625679.5, "dur": 64510.859, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255708048.977, "dur": 88313.669, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255814703.934, "dur": 73176.832, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 255908845.418, "dur": 95125.776, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 256021803.411, "dur": 150985.331, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 256194023.371, "dur": 435001.119, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 256648210.711, "dur": 115135.677, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 256780506.452, "dur": 59175.274, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 256898375.62, "dur": 172101.563, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257090527.538, "dur": 63201.358, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257170882.1, "dur": 54032.124, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257241092.097, "dur": 70243.24, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257327222.932, "dur": 60425.354, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257409783.672, "dur": 106824.921, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257532536.77, "dur": 108935.903, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257658201.56, "dur": 61666.712, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257735767.908, "dur": 55446.543, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257808362.885, "dur": 59261.811, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257885218.218, "dur": 92598.599, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 257994560.276, "dur": 58595.93, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 258069177.509, "dur": 55782.913, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 258141324.127, "dur": 306433.245, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 258466465.053, "dur": 280676.883, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 258769944.534, "dur": 67394.599, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 258853618.162, "dur": 171038.739, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259041766.262, "dur": 90512.177, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259148166.653, "dur": 61152.719, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259225929.302, "dur": 64970.327, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259309450.061, "dur": 62720.876, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259388357.551, "dur": 187441.235, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259592097.196, "dur": 75976.277, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259684386.754, "dur": 95430.535, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259796249.468, "dur": 79200.598, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259891422.036, "dur": 77207.904, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 259990781.236, "dur": 114016.449, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 260130970.343, "dur": 89781.051, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 260237294.58, "dur": 541177.142, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 260794824.946, "dur": 75789.694, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 260886270.501, "dur": 153423.855, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261055454.093, "dur": 131683.094, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261202205.76, "dur": 53573.404, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261270660.106, "dur": 56433.303, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261341605.367, "dur": 41203.06, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261397037.598, "dur": 121434.592, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261533750.837, "dur": 136788.343, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 261685193.068, "dur": 49936.826, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 266546193.184, "dur": 10.71, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 266546612.83, "dur": 9.611, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "core: _handle_client_request", "ph": "X", "ts": 266546733.468, "dur": 78.224, "pid": 110, "tid": 1, "args": {"note": "UTILITY"}}, {"name": "msgpack encode", "ph": "X", "ts": 266546909.779, "dur": 35.452, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 266547277.232, "dur": 276.757, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 266547897.0, "dur": 3.021, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 266548161.087, "dur": 5.05, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "core: _handle_client_request", "ph": "X", "ts": 266548228.551, "dur": 143.189, "pid": 110, "tid": 1, "args": {"note": "UTILITY"}}, {"name": "msgpack encode", "ph": "X", "ts": 266548530.46, "dur": 40.123, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 266548882.862, "dur": 16.291, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "render: chat template and tokenize", "ph": "X", "ts": 269866979.641, "dur": 4000.59, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "HTTP handler: create_chat_completion", "ph": "X", "ts": 269866930.638, "dur": 4387.244, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 269893709.792, "dur": 15.971, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "client: send request over ZeroMQ", "ph": "X", "ts": 269893693.311, "dur": 175.611, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "AsyncLLM._add_request", "ph": "X", "ts": 269893608.655, "dur": 292.309, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269894369.843, "dur": 186.812, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "core: preprocess_add_request (block hashes)", "ph": "X", "ts": 269894749.547, "dur": 315.389, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "Scheduler.add_request", "ph": "X", "ts": 269895272.099, "dur": 37.783, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "core: _handle_client_request", "ph": "X", "ts": 269895243.368, "dur": 167.14, "pid": 110, "tid": 1, "args": {"note": "ADD"}}, {"name": "KVCacheManager.get_computed_blocks", "ph": "X", "ts": 269895544.876, "dur": 48.274, "pid": 110, "tid": 1, "args": {"note": "prefix cache hit: 0 of 935 tokens"}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269895653.343, "dur": 189.492, "pid": 110, "tid": 1, "args": {"note": "935 new tokens, 59 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269895498.953, "dur": 532.914, "pid": 110, "tid": 1, "args": {"note": "935 tokens for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269897686.17, "dur": 556.505, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269898293.958, "dur": 129.828, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269896140.074, "dur": 49611.561, "pid": 110, "tid": 1, "args": {"note": "935 tokens"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269945914.565, "dur": 1715.118, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269895491.693, "dur": 52218.635, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269947814.854, "dur": 47.683, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269947771.041, "dur": 197.783, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269948269.143, "dur": 462.769, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269948768.514, "dur": 118.167, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269949151.718, "dur": 533.523, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269948033.198, "dur": 1718.367, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269949824.02, "dur": 871.514, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269950765.139, "dur": 21.261, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269950827.512, "dur": 152.48, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269947765.211, "dur": 3271.705, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269951123.281, "dur": 22.601, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269951105.38, "dur": 165.16, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 269951521.836, "dur": 51.003, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269952521.598, "dur": 61.264, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269952270.793, "dur": 505.681, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269952829.497, "dur": 125.988, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269953239.723, "dur": 37.213, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269951331.804, "dur": 1989.094, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269952877.391, "dur": 800.28, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269953378.202, "dur": 840.813, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269954281.578, "dur": 2112.162, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269956466.125, "dur": 79.365, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269951102.61, "dur": 5500.114, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269956672.658, "dur": 20.131, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269956661.197, "dur": 98.456, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 269956906.972, "dur": 11.841, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269957413.574, "dur": 54.984, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269957187.23, "dur": 392.655, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269957614.447, "dur": 101.146, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269957683.851, "dur": 153.3, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269957942.737, "dur": 30.792, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269956807.786, "dur": 1207.826, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269958067.365, "dur": 809.991, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269958931.539, "dur": 4131.138, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269963221.707, "dur": 127.418, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269956658.217, "dur": 6851.678, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269963632.313, "dur": 41.973, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269963607.041, "dur": 192.632, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 269964091.752, "dur": 11.45, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269964734.172, "dur": 44.843, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269964469.515, "dur": 392.305, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269964900.112, "dur": 108.107, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269965071.103, "dur": 141.569, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269965282.446, "dur": 27.772, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269963905.21, "dur": 1450.341, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269965417.364, "dur": 841.013, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269966319.361, "dur": 3195.07, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269969560.034, "dur": 67.764, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269963602.991, "dur": 6074.75, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269969743.355, "dur": 22.331, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269969729.344, "dur": 108.287, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 269970136.899, "dur": 23.962, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269970975.922, "dur": 30.142, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269971169.154, "dur": 104.167, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269971063.717, "dur": 817.582, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269971953.573, "dur": 148.78, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269972521.599, "dur": 46.892, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269969878.043, "dur": 2747.462, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269972727.721, "dur": 1214.107, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269974019.302, "dur": 2144.124, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269976236.321, "dur": 72.595, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269969726.824, "dur": 6670.657, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269976516.829, "dur": 27.961, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269976499.967, "dur": 143.849, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 269976865.4, "dur": 12.631, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269977281.416, "dur": 25.692, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269977493.71, "dur": 112.656, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269977277.406, "dur": 789.499, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269978113.558, "dur": 179.462, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269978775.73, "dur": 49.063, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269976718.781, "dur": 2165.556, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269978966.291, "dur": 1901.169, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269981007.049, "dur": 1695.196, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269982812.952, "dur": 94.886, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269976497.107, "dur": 6476.105, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 269983053.037, "dur": 26.692, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269983035.096, "dur": 130.968, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 269983364.597, "dur": 18.421, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269983970.764, "dur": 43.233, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269984166.457, "dur": 95.236, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 269983882.269, "dur": 870.454, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 269984806.287, "dur": 160.34, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 269985410.634, "dur": 84.246, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269983232.608, "dur": 2322.626, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 269985676.811, "dur": 1382.917, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269987138.633, "dur": 2279.002, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269989625.288, "dur": 132.888, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269983031.266, "dur": 6895.501, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269990047.464, "dur": 81.785, "pid": 110, "tid": 1, "args": {"note": "0 tokens for 0 requests"}}, {"name": "msgpack encode", "ph": "X", "ts": 269990538.925, "dur": 24.211, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269990256.087, "dur": 936.199, "pid": 110, "tid": 1, "args": {"note": "0 tokens"}}, {"name": "msgpack decode", "ph": "X", "ts": 269991223.108, "dur": 32.972, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269991499.465, "dur": 116.947, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 269991379.138, "dur": 4634.249, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269996207.149, "dur": 304.439, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269990041.914, "dur": 6663.537, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 269996898.803, "dur": 25.421, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 269997610.837, "dur": 39.573, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 269997898.575, "dur": 230.785, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 269998079.777, "dur": 126.467, "pid": 110, "tid": 1, "args": {"note": "0 tokens for 0 requests"}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 269998413.177, "dur": 489.171, "pid": 110, "tid": 1, "args": {"note": "0 tokens"}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269999094.51, "dur": 69.734, "pid": 110, "tid": 1, "args": {"note": "outputs for 0 requests"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269998068.986, "dur": 1168.393, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 269999383.088, "dur": 42.702, "pid": 110, "tid": 1, "args": {"note": "outputs for 0 requests"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 269999342.215, "dur": 255.356, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 269999700.198, "dur": 15.971, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 269999985.656, "dur": 24.061, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270000398.231, "dur": 27.802, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270000613.545, "dur": 13.841, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "render: chat template and tokenize", "ph": "X", "ts": 270503364.17, "dur": 5669.425, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "HTTP handler: create_chat_completion", "ph": "X", "ts": 270503331.298, "dur": 6096.182, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 270510421.722, "dur": 54.783, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "client: send request over ZeroMQ", "ph": "X", "ts": 270510411.511, "dur": 218.364, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "AsyncLLM._add_request", "ph": "X", "ts": 270510306.604, "dur": 349.372, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270510972.286, "dur": 123.638, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "core: preprocess_add_request (block hashes)", "ph": "X", "ts": 270511166.958, "dur": 449.028, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "Scheduler.add_request", "ph": "X", "ts": 270511801.048, "dur": 30.732, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "core: _handle_client_request", "ph": "X", "ts": 270511781.777, "dur": 111.797, "pid": 110, "tid": 1, "args": {"note": "ADD"}}, {"name": "KVCacheManager.get_computed_blocks", "ph": "X", "ts": 270511979.399, "dur": 65.554, "pid": 110, "tid": 1, "args": {"note": "prefix cache hit: 16 of 2022 tokens"}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270512106.417, "dur": 275.337, "pid": 110, "tid": 1, "args": {"note": "2006 new tokens, 126 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270511948.257, "dur": 593.417, "pid": 110, "tid": 1, "args": {"note": "2006 tokens for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270514082.7, "dur": 767.288, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270514896.801, "dur": 136.809, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270512635.9, "dur": 50954.565, "pid": 110, "tid": 1, "args": {"note": "2006 tokens"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270563738.954, "dur": 1080.548, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270511943.137, "dur": 52959.79, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270564977.242, "dur": 31.012, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270564955.88, "dur": 131.648, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270565364.946, "dur": 504.911, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270565923.721, "dur": 123.198, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270566346.367, "dur": 43.463, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270565141.562, "dur": 1324.053, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270566543.259, "dur": 690.484, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270567305.277, "dur": 16273.758, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270583648.879, "dur": 133.708, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270564952.0, "dur": 18954.095, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270584014.332, "dur": 28.212, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270583983.23, "dur": 154.619, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270584301.28, "dur": 16.421, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270584836.953, "dur": 36.232, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270584718.456, "dur": 464.009, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270585224.177, "dur": 116.877, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270585667.225, "dur": 38.252, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270584207.334, "dur": 1544.246, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270585065.317, "dur": 909.497, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270585817.664, "dur": 788.05, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270586708.24, "dur": 3600.315, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270590358.518, "dur": 62.644, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270583979.76, "dur": 6519.747, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270590573.662, "dur": 22.641, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "msgpack encode", "ph": "X", "ts": 270590696.959, "dur": 22.102, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270590560.121, "dur": 312.039, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack decode", "ph": "X", "ts": 270591407.954, "dur": 20.121, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270591592.725, "dur": 69.105, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270591593.525, "dur": 717.115, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270592364.364, "dur": 156.899, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270592897.567, "dur": 44.963, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270590938.775, "dur": 2057.578, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270593084.289, "dur": 906.146, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270594060.82, "dur": 2904.121, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270597025.835, "dur": 61.964, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270590557.471, "dur": 6592.432, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270597235.548, "dur": 23.842, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270597222.747, "dur": 111.117, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270597490.274, "dur": 9.371, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270597803.653, "dur": 16.362, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270597940.832, "dur": 59.574, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270597788.363, "dur": 458.628, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270598283.964, "dur": 119.787, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270598710.86, "dur": 36.252, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270597390.428, "dur": 1403.527, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270598858.259, "dur": 725.436, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270599647.099, "dur": 4013.891, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270603707.213, "dur": 54.533, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270597219.787, "dur": 6598.322, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270603889.464, "dur": 20.521, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270603877.723, "dur": 98.756, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270604161.871, "dur": 17.221, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270604589.578, "dur": 14.901, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270604725.186, "dur": 58.614, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270604606.589, "dur": 420.566, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270605066.367, "dur": 110.198, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270605471.003, "dur": 34.782, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270604043.704, "dur": 1508.134, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270605620.502, "dur": 638.32, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270606316.886, "dur": 3967.208, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270610348.608, "dur": 55.963, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270603875.113, "dur": 6609.613, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270610552.67, "dur": 21.562, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270610540.33, "dur": 102.006, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270610776.335, "dur": 10.33, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270611068.373, "dur": 13.741, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270611190.81, "dur": 54.684, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270611081.334, "dur": 498.701, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270611617.637, "dur": 127.328, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270612042.123, "dur": 37.003, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270610688.349, "dur": 1437.92, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270612189.123, "dur": 700.953, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270612981.752, "dur": 4006.571, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270617032.335, "dur": 52.734, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270610537.689, "dur": 6603.083, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270617209.017, "dur": 21.711, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270617196.326, "dur": 99.856, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270617494.684, "dur": 7.581, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270617757.971, "dur": 12.881, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270617875.028, "dur": 51.413, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270617742.67, "dur": 394.445, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270618173.097, "dur": 108.487, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270618574.352, "dur": 35.382, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270617351.675, "dur": 1303.442, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270618712.751, "dur": 647.4, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270619414.864, "dur": 4199.393, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270623656.209, "dur": 48.834, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270617193.756, "dur": 6559.13, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270623815.73, "dur": 19.211, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270623804.429, "dur": 95.356, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270624018.392, "dur": 8.711, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270624318.051, "dur": 12.421, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270624452.089, "dur": 50.484, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270624351.523, "dur": 395.885, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270624783.36, "dur": 103.046, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270625129.962, "dur": 30.741, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270623946.918, "dur": 1255.498, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270625255.74, "dur": 629.379, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270625937.162, "dur": 4327.351, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270630327.366, "dur": 79.496, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270623801.859, "dur": 6692.088, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270630569.532, "dur": 24.241, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270630553.161, "dur": 121.247, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270630720.181, "dur": 9.391, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270631000.979, "dur": 11.09, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270631105.765, "dur": 43.403, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270631095.774, "dur": 426.097, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270631570.444, "dur": 107.997, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270631934.557, "dur": 29.312, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270630890.962, "dur": 1129.11, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270632077.406, "dur": 654.121, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270632786.92, "dur": 4090.666, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270636935.67, "dur": 56.533, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270630550.12, "dur": 6500.887, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270637118.951, "dur": 21.701, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270637104.99, "dur": 106.827, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270637327.064, "dur": 6.831, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270637582.44, "dur": 9.911, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270637683.696, "dur": 41.723, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270637624.783, "dur": 411.905, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270638079.181, "dur": 108.827, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270638466.675, "dur": 31.112, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270637266.88, "dur": 1274.99, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270638598.754, "dur": 657.831, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270639314.568, "dur": 4196.833, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270643553.813, "dur": 79.615, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270637102.04, "dur": 6588.062, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270643755.386, "dur": 43.923, "pid": 110, "tid": 1, "args": {"note": "1 new token, 1 new block"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270643743.475, "dur": 132.678, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270644028.573, "dur": 10.17, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270644344.273, "dur": 11.39, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270644498.622, "dur": 43.273, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270644511.493, "dur": 345.552, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270644895.257, "dur": 104.197, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270645285.661, "dur": 28.672, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270643928.917, "dur": 1430.009, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270645413.22, "dur": 654.79, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270646121.714, "dur": 3971.178, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270650136.375, "dur": 44.392, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270643741.095, "dur": 6488.366, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270650287.174, "dur": 18.231, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270650276.573, "dur": 89.736, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270650491.437, "dur": 7.79, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270650754.943, "dur": 9.171, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270650854.33, "dur": 38.752, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270650785.585, "dur": 426.507, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270651257.165, "dur": 108.707, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270651623.848, "dur": 25.431, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270650402.631, "dur": 1291.341, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270651745.315, "dur": 614.259, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270652413.457, "dur": 4358.413, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270656823.443, "dur": 50.653, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270650274.313, "dur": 6659.627, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270657006.764, "dur": 22.122, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270656994.664, "dur": 111.636, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270657231.668, "dur": 7.111, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270657541.948, "dur": 11.74, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270657666.775, "dur": 62.754, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270657575.03, "dur": 419.766, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270658036.689, "dur": 115.287, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270658405.732, "dur": 50.803, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270657161.094, "dur": 1339.873, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270658548.77, "dur": 583.507, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270659183.51, "dur": 4146.619, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270663374.552, "dur": 41.893, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270656991.083, "dur": 6497.956, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270663578.095, "dur": 17.411, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270663568.054, "dur": 92.006, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270663846.632, "dur": 8.21, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270664145.931, "dur": 10.44, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270664250.357, "dur": 38.382, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270664062.735, "dur": 524.303, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270664625.16, "dur": 106.247, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270664984.133, "dur": 27.381, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270663697.743, "dur": 1358.904, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270665115.831, "dur": 643.45, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270665815.895, "dur": 4208.483, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270670191.198, "dur": 141.559, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270663565.914, "dur": 6932.803, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270670617.275, "dur": 34.242, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270670592.803, "dur": 159.78, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270671052.492, "dur": 10.581, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270671352.561, "dur": 8.961, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270671474.549, "dur": 36.902, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270671465.528, "dur": 371.013, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270671874.714, "dur": 97.626, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270672213.565, "dur": 24.521, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270670866.84, "dur": 1416.009, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270672348.383, "dur": 612.579, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270673017.115, "dur": 3586.794, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270676671.883, "dur": 70.055, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270670588.073, "dur": 6225.449, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270676895.557, "dur": 27.142, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270676877.536, "dur": 133.348, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270677067.098, "dur": 15.661, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270677627.873, "dur": 34.322, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270677841.787, "dur": 117.057, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270677561.039, "dur": 487.59, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270678101.873, "dur": 122.758, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270678514.209, "dur": 27.991, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270677216.588, "dur": 1371.515, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270678648.407, "dur": 648.271, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270679360.851, "dur": 3842.621, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270683269.436, "dur": 85.245, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270676873.746, "dur": 6551.599, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270683547.043, "dur": 31.492, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270683527.782, "dur": 152.279, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270683926.737, "dur": 10.64, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270684408.897, "dur": 54.003, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270684251.487, "dur": 423.917, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270684644.852, "dur": 119.957, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270684719.086, "dur": 104.237, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270685060.048, "dur": 25.221, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270683752.916, "dur": 1377.426, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270685191.056, "dur": 633.49, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270685886.059, "dur": 3924.886, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270689878.139, "dur": 66.004, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270683523.782, "dur": 6487.605, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270690082.862, "dur": 23.691, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270690067.391, "dur": 117.777, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270690398.381, "dur": 13.661, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270690968.437, "dur": 28.032, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270690727.412, "dur": 413.316, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270691183.51, "dur": 102.337, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270691219.023, "dur": 104.996, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270691551.773, "dur": 27.892, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270690235.711, "dur": 1391.417, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270691686.262, "dur": 616.058, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270692396.876, "dur": 4067.915, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270696506.993, "dur": 50.973, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270690063.611, "dur": 6551.609, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 270696678.104, "dur": 20.201, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270696665.863, "dur": 101.107, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 270696959.882, "dur": 18.501, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270697414.52, "dur": 43.983, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270697614.692, "dur": 87.836, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 270697527.357, "dur": 806.17, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 270698415.493, "dur": 167.71, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 270699031.271, "dur": 51.393, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270696812.212, "dur": 2336.996, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 270699264.115, "dur": 1246.998, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270700616.78, "dur": 2438.512, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270703153.399, "dur": 264.736, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270696663.613, "dur": 6876.28, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 270703619.488, "dur": 67.724, "pid": 110, "tid": 1, "args": {"note": "0 tokens for 0 requests"}}, {"name": "msgpack encode", "ph": "X", "ts": 270703917.586, "dur": 19.722, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 270703779.608, "dur": 577.506, "pid": 110, "tid": 1, "args": {"note": "0 tokens"}}, {"name": "msgpack decode", "ph": "X", "ts": 270704494.582, "dur": 33.333, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 270704720.527, "dur": 161.73, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 270704517.364, "dur": 5144.341, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270709744.151, "dur": 53.633, "pid": 110, "tid": 1, "args": {"note": "outputs for 0 requests"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270703614.678, "dur": 6293.863, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 270710033.719, "dur": 34.102, "pid": 110, "tid": 1, "args": {"note": "outputs for 0 requests"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 270709998.766, "dur": 119.668, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 270710253.153, "dur": 9.37, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 270710351.529, "dur": 6.66, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270710640.007, "dur": 20.101, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 270710781.186, "dur": 8.54, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "render: chat template and tokenize", "ph": "X", "ts": 271208702.842, "dur": 4213.753, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "HTTP handler: create_chat_completion", "ph": "X", "ts": 271208675.74, "dur": 4498.511, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 271213884.076, "dur": 24.911, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "client: send request over ZeroMQ", "ph": "X", "ts": 271213877.865, "dur": 151.18, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "AsyncLLM._add_request", "ph": "X", "ts": 271213804.741, "dur": 246.305, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271214271.55, "dur": 103.466, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "core: preprocess_add_request (block hashes)", "ph": "X", "ts": 271214425.01, "dur": 389.424, "pid": 110, "tid": 3, "args": {"note": ""}}, {"name": "Scheduler.add_request", "ph": "X", "ts": 271214998.335, "dur": 36.803, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "core: _handle_client_request", "ph": "X", "ts": 271214976.134, "dur": 158.72, "pid": 110, "tid": 1, "args": {"note": "ADD"}}, {"name": "KVCacheManager.get_computed_blocks", "ph": "X", "ts": 271215231.44, "dur": 190.982, "pid": 110, "tid": 1, "args": {"note": "prefix cache hit: 2000 of 2021 tokens"}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271215499.707, "dur": 138.458, "pid": 110, "tid": 1, "args": {"note": "21 new tokens, 2 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271215200.168, "dur": 570.206, "pid": 110, "tid": 1, "args": {"note": "21 tokens for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271217593.577, "dur": 699.214, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271218338.114, "dur": 175.881, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: run piecewise CUDA graphs", "ph": "X", "ts": 271218982.635, "dur": 10950.004, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271215877.84, "dur": 14207.539, "pid": 110, "tid": 1, "args": {"note": "21 tokens"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271230225.738, "dur": 1507.124, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271215194.347, "dur": 16660.592, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271231961.316, "dur": 53.953, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271231926.624, "dur": 209.333, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271232598.406, "dur": 715.065, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271233395.736, "dur": 201.912, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271234110.27, "dur": 50.534, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271232220.952, "dur": 2015.296, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271234356.016, "dur": 1195.235, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271235686.539, "dur": 29.382, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271235774.394, "dur": 192.932, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271231921.673, "dur": 4155.85, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271236198.091, "dur": 38.892, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271236167.619, "dur": 182.021, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271236578.505, "dur": 19.991, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271236941.777, "dur": 24.182, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271237101.707, "dur": 805.751, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271237213.764, "dur": 838.893, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271238094.539, "dur": 146.14, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271238707.848, "dur": 52.113, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271236455.647, "dur": 2364.988, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271238929.362, "dur": 955.98, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271239955.096, "dur": 986.612, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271240993.601, "dur": 90.295, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271236162.679, "dur": 4985.811, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271241238.906, "dur": 30.862, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271241215.105, "dur": 160.22, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271241607.789, "dur": 18.841, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271242107.58, "dur": 28.142, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271242314.073, "dur": 89.216, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271242049.637, "dur": 411.386, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271242503.325, "dur": 105.537, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271242846.027, "dur": 22.021, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271241459.69, "dur": 1451.311, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271242962.664, "dur": 612.488, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271243628.336, "dur": 3925.635, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271247600.754, "dur": 44.743, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271241210.924, "dur": 6486.486, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271247757.884, "dur": 21.041, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "msgpack encode", "ph": "X", "ts": 271247874.591, "dur": 18.941, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271247747.443, "dur": 353.992, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack decode", "ph": "X", "ts": 271248315.318, "dur": 15.641, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271248461.858, "dur": 54.733, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271248400.944, "dur": 437.517, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271248884.054, "dur": 109.107, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271249265.118, "dur": 22.451, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271248155.808, "dur": 1182.034, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271249392.586, "dur": 568.695, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271250035.076, "dur": 4019.441, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271254097.41, "dur": 38.972, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271247745.343, "dur": 6437.693, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271254238.489, "dur": 17.801, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271254229.318, "dur": 89.486, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271254479.304, "dur": 8.811, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271254988.066, "dur": 35.432, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271254709.068, "dur": 351.742, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271255142.796, "dur": 102.436, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271255271.193, "dur": 113.487, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271255493.697, "dur": 19.932, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271254362.477, "dur": 1194.724, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271255605.264, "dur": 555.045, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271256213.432, "dur": 4441.388, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271260703.783, "dur": 45.663, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271254227.038, "dur": 6579.161, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271260871.734, "dur": 20.841, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271260860.203, "dur": 105.906, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271261127.07, "dur": 5.94, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271261389.706, "dur": 14.471, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271261537.895, "dur": 51.684, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271261325.402, "dur": 1371.506, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271262803.364, "dur": 196.693, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271263590.804, "dur": 49.533, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271261018.853, "dur": 2721.76, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271263873.231, "dur": 1331.164, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271265377.476, "dur": 1844.455, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271267291.085, "dur": 39.142, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271260857.703, "dur": 6532.368, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271267467.506, "dur": 18.911, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271267456.975, "dur": 101.457, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271267777.346, "dur": 19.161, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271268159.589, "dur": 10.201, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271268264.736, "dur": 38.572, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271268481.31, "dur": 1106.969, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271269702.426, "dur": 211.363, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271270544.128, "dur": 59.724, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271267634.737, "dur": 3064.221, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271270825.396, "dur": 1327.403, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271272279.777, "dur": 1542.947, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271273884.737, "dur": 42.263, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271267454.825, "dur": 6539.609, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271274056.568, "dur": 19.371, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "msgpack encode", "ph": "X", "ts": 271274150.784, "dur": 14.181, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271274045.687, "dur": 300.379, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack decode", "ph": "X", "ts": 271274588.981, "dur": 17.132, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271274745.141, "dur": 65.824, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271274609.573, "dur": 414.996, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271275060.921, "dur": 106.507, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271275421.384, "dur": 47.603, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271274402.71, "dur": 1111.099, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271275567.803, "dur": 612.998, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271276233.654, "dur": 4287.868, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271280609.728, "dur": 97.806, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271274043.727, "dur": 6790.145, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271280922.377, "dur": 28.852, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271280899.956, "dur": 141.439, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271281346.904, "dur": 19.951, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271281845.075, "dur": 9.811, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271281950.502, "dur": 39.742, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271281953.082, "dur": 771.398, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271282771.183, "dur": 141.959, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271283352.859, "dur": 49.324, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271281140.091, "dur": 2337.136, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271283564.153, "dur": 963.36, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271284590.607, "dur": 2399.82, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271287096.774, "dur": 121.977, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271280896.586, "dur": 6475.235, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271287529.43, "dur": 39.503, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271287501.679, "dur": 201.482, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271288014.231, "dur": 10.221, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271288304.779, "dur": 9.881, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271288409.416, "dur": 59.623, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271288423.466, "dur": 909.737, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271289403.408, "dur": 220.053, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271290090.471, "dur": 49.863, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271287813.458, "dur": 2404.371, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271290321.215, "dur": 1144.052, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271291562.843, "dur": 1971.363, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271293598.57, "dur": 89.536, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271287496.808, "dur": 6253.922, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271293836.735, "dur": 29.322, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271293816.204, "dur": 150.239, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271294104.632, "dur": 7.68, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271294381.349, "dur": 13.021, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271294520.138, "dur": 50.523, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271294452.213, "dur": 422.597, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271294915.622, "dur": 109.057, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271295278.915, "dur": 28.212, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271294031.587, "dur": 1318.603, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271295407.153, "dur": 644.59, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271296108.437, "dur": 4031.502, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271300208.373, "dur": 100.207, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271293812.083, "dur": 6579.772, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271300545.864, "dur": 59.364, "pid": 110, "tid": 1, "args": {"note": "1 new token, 1 new block"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271300528.593, "dur": 181.902, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271300879.105, "dur": 8.831, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271301197.625, "dur": 12.881, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271301332.734, "dur": 48.493, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271301487.283, "dur": 476.16, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271302002.526, "dur": 116.287, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271302390.11, "dur": 26.982, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271300785.519, "dur": 1703.617, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271302552.75, "dur": 676.932, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271303294.256, "dur": 3422.314, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271306766.953, "dur": 62.074, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271300525.443, "dur": 6371.038, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271306974.206, "dur": 26.872, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271306957.905, "dur": 123.458, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271307243.893, "dur": 10.851, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271307628.897, "dur": 20.922, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271307793.057, "dur": 68.435, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271307613.826, "dur": 669.562, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271308332.811, "dur": 196.443, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271308967.901, "dur": 46.733, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271307145.627, "dur": 1929.731, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271309147.672, "dur": 1010.224, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271310237.21, "dur": 3089.864, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271313391.128, "dur": 89.015, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271306954.745, "dur": 6597.543, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271313629.842, "dur": 25.662, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271313612.261, "dur": 127.988, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271313942.932, "dur": 17.221, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271314290.404, "dur": 13.821, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271314415.122, "dur": 89.335, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271314613.994, "dur": 1017.624, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271315701.582, "dur": 202.083, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271316480.251, "dur": 57.553, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271313802.573, "dur": 2809.696, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271316731.796, "dur": 1163.623, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271317996.106, "dur": 1877.287, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271319945.217, "dur": 113.658, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271313608.711, "dur": 6532.249, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271320241.526, "dur": 33.552, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271320218.934, "dur": 155.54, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271320616.83, "dur": 8.52, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271321052.687, "dur": 23.481, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271321215.057, "dur": 79.695, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271320973.322, "dur": 574.396, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271321585.36, "dur": 121.427, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271322077.761, "dur": 45.303, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271320470.75, "dur": 1704.637, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271322246.901, "dur": 786.829, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271323096.854, "dur": 3383.922, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271326539.19, "dur": 55.173, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271320214.844, "dur": 6443.703, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271326729.352, "dur": 23.891, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271326715.631, "dur": 114.847, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271327004.699, "dur": 10.3, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271327355.211, "dur": 14.221, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271327523.141, "dur": 57.044, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271327302.577, "dur": 439.288, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271327782.567, "dur": 107.187, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271328152.01, "dur": 27.172, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271326885.822, "dur": 1339.523, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271328284.559, "dur": 622.679, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271328986.743, "dur": 4108.557, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271333158.994, "dur": 73.614, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271326712.24, "dur": 6582.472, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271333371.677, "dur": 25.412, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271333355.456, "dur": 142.379, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271333687.617, "dur": 17.281, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271334112.813, "dur": 21.082, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271334275.504, "dur": 76.254, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271334267.533, "dur": 868.494, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271335218.363, "dur": 156.119, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271335874.124, "dur": 48.712, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271333562.809, "dur": 2453.713, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271336130.949, "dur": 1059.207, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271337297.162, "dur": 2364.208, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271339737.065, "dur": 74.485, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271333351.616, "dur": 6541.239, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271339977.53, "dur": 38.313, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271339959.549, "dur": 151.319, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271340313.101, "dur": 9.661, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271340626.831, "dur": 18.101, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271340752.549, "dur": 65.424, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271340684.924, "dur": 625.379, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271341351.466, "dur": 169.991, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271341922.502, "dur": 45.812, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271340179.613, "dur": 1837.944, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271342089.432, "dur": 933.968, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271343092.695, "dur": 3236.352, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271346504.788, "dur": 131.578, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271339955.949, "dur": 6837.727, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271346925.304, "dur": 41.093, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271346894.582, "dur": 207.484, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271347537.583, "dur": 8.49, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271348016.223, "dur": 21.631, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271348167.132, "dur": 71.494, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271348094.608, "dur": 961.05, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271349130.002, "dur": 152.88, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271349795.914, "dur": 49.763, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271347275.016, "dur": 2639.395, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271350018.968, "dur": 1080.168, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271351200.122, "dur": 1605.12, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271352880.067, "dur": 100.916, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271346889.642, "dur": 6174.946, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271353147.604, "dur": 28.701, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271353126.502, "dur": 139.929, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271353457.843, "dur": 10.17, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271353747.861, "dur": 18.081, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271353879.239, "dur": 62.734, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271353827.626, "dur": 671.102, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271354536.63, "dur": 137.269, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271355070.294, "dur": 47.923, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271353337.265, "dur": 1829.305, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271355237.914, "dur": 898.286, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271356198.404, "dur": 3278.855, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271359544.633, "dur": 63.284, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271353122.362, "dur": 6561.43, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271359765.387, "dur": 28.202, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271359750.066, "dur": 134.119, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271360193.894, "dur": 11.561, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271360519.874, "dur": 16.032, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271360644.062, "dur": 61.834, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271360459.691, "dur": 584.406, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271361092.9, "dur": 122.218, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271361511.896, "dur": 27.112, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271359950.619, "dur": 1651.153, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271361658.196, "dur": 601.717, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271362314.377, "dur": 3661.509, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271366038.499, "dur": 74.075, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271359746.076, "dur": 6426.822, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271366247.873, "dur": 25.091, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271366232.412, "dur": 123.497, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271366530.41, "dur": 9.301, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271366812.578, "dur": 11.061, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271366917.794, "dur": 49.214, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271366852.27, "dur": 464.129, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271367356.392, "dur": 145.779, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271367799.4, "dur": 36.162, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271366411.233, "dur": 1467.302, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271367935.868, "dur": 729.756, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271368724.778, "dur": 3926.795, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271372804.883, "dur": 122.267, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271366229.251, "dur": 6851.629, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "KVCacheManager.allocate_slots", "ph": "X", "ts": 271373209.868, "dur": 42.493, "pid": 110, "tid": 1, "args": {"note": "1 new token, 0 new blocks"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271373181.516, "dur": 200.453, "pid": 110, "tid": 1, "args": {"note": "1 token for 1 request"}}, {"name": "msgpack encode", "ph": "X", "ts": 271373774.793, "dur": 13.981, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271374188.329, "dur": 14.151, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271374315.697, "dur": 49.453, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "runner: prepare_inputs", "ph": "X", "ts": 271374263.354, "dur": 865.224, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: prepare_attn", "ph": "X", "ts": 271375194.452, "dur": 171.961, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "runner: replay full CUDA graph", "ph": "X", "ts": 271375890.246, "dur": 58.913, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271373516.817, "dur": 2516.767, "pid": 110, "tid": 1, "args": {"note": "1 token"}}, {"name": "GPUModelRunner.sample_tokens (V2)", "ph": "X", "ts": 271376152.802, "dur": 1106.589, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271377356.637, "dur": 1730.838, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271379174.561, "dur": 116.907, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271373176.176, "dur": 6203.828, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271379529.063, "dur": 66.924, "pid": 110, "tid": 1, "args": {"note": "0 tokens for 0 requests"}}, {"name": "msgpack encode", "ph": "X", "ts": 271379835.922, "dur": 11.791, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271379670.842, "dur": 519.342, "pid": 110, "tid": 1, "args": {"note": "0 tokens"}}, {"name": "msgpack decode", "ph": "X", "ts": 271380355.355, "dur": 27.561, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271380593.82, "dur": 91.135, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "AsyncOutput.get_output: wait for the GPU", "ph": "X", "ts": 271380276.13, "dur": 5372.746, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271385788.545, "dur": 205.852, "pid": 110, "tid": 1, "args": {"note": "outputs for 1 request"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271379524.423, "dur": 6620.284, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 271386261.004, "dur": 12.111, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271386732.824, "dur": 29.591, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "OutputProcessor.process_outputs: detokenize", "ph": "X", "ts": 271386935.976, "dur": 139.409, "pid": 31, "tid": 2, "args": {"note": "outputs for 1 request"}}, {"name": "Scheduler.schedule", "ph": "X", "ts": 271387600.618, "dur": 73.934, "pid": 110, "tid": 1, "args": {"note": "0 tokens for 0 requests"}}, {"name": "GPUModelRunner.execute_model (V2)", "ph": "X", "ts": 271387780.349, "dur": 331.811, "pid": 110, "tid": 1, "args": {"note": "0 tokens"}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271388232.907, "dur": 40.753, "pid": 110, "tid": 1, "args": {"note": "outputs for 0 requests"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271387595.038, "dur": 721.265, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "Scheduler.update_from_output", "ph": "X", "ts": 271388390.297, "dur": 19.461, "pid": 110, "tid": 1, "args": {"note": "outputs for 0 requests"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271388372.866, "dur": 108.607, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 271388533.016, "dur": 9.051, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack encode", "ph": "X", "ts": 271388700.656, "dur": 8.241, "pid": 110, "tid": 4, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271388833.445, "dur": 13.101, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "msgpack decode", "ph": "X", "ts": 271388941.632, "dur": 7.07, "pid": 31, "tid": 2, "args": {"note": ""}}, {"name": "core: _handle_client_request", "ph": "X", "ts": 271901790.439, "dur": 7.071, "pid": 110, "tid": 1, "args": {"note": "WAKEUP"}}, {"name": "EngineCore.step_with_batch_queue", "ph": "X", "ts": 271901913.857, "dur": 4.55, "pid": 110, "tid": 1, "args": {"note": ""}}, {"name": "thread_name", "ph": "M", "pid": 110, "tid": 1, "args": {"name": "MainThread"}}, {"name": "thread_name", "ph": "M", "pid": 31, "tid": 2, "args": {"name": "MainThread"}}, {"name": "thread_name", "ph": "M", "pid": 110, "tid": 3, "args": {"name": "Thread-2 (process_input_sockets)"}}, {"name": "thread_name", "ph": "M", "pid": 110, "tid": 4, "args": {"name": "Thread-3 (process_output_sockets)"}}, {"name": "request 1 sent", "ph": "i", "s": "g", "ts": 270501202.545, "pid": 0, "tid": 0}, {"name": "request 2 sent", "ph": "i", "s": "g", "ts": 271207386.409, "pid": 0, "tid": 0}]}